Okra (Abelmoschus esculentus (L.) Moench) belonging to the family Malvaceae is an important vegetable crop of the tropics and subtropics. Okra is specially valued for its tender, delicious green fruits which are cooked, canned and consumed in various forms in different parts of the country and for about 60% of the export of fresh vegetables. India is the largest producer of okra covering an area of 3.91 lakh ha with an annual production of 39.7 lakh tonnes. It has good nutriational value, particularly vitamin C (30 mg/100 g), calcium (90 mg/100 g) and iron (1.5 mg/100 g) in the edible fruit. It is a potential exporter earner accounting for 13% of export of fresh vegetables. All forms of plant improvement activities through breeding contemplate an eventual boost in genetic potential for yield. Since, yield is polygenically controlled and highly influenced by environment, selection based on yield alone is not effective. The breeder hence develops into proposition of selecting for high yield indirectly through yield associated and highly heritable characters after eliminating environmental components of phenotypic variance (Prakash et al., 2013).

Hybridization has been the most successful approach in increasing the productivity in vegetable crops. Selection of genetically superior and suitable genotypes is the most important stage from the standpoint of hybridization of vegetable crops in order to develop new genotypes having desirable characters. To break the yield barriers in existing open-pollinated varieties of okra, a hybridization-based breeding strategy would be desirable. Heterosis breeding has been the most successful approach in increasing the productivity in cross-pollinated vegetable crops (Medagam et al., 2012). The ease in emasculation, very high per cent of fruit set and large number of seeds per fruit makes commercial exploitation of hybrid vigour easy in bhindi. Being an often cross-pollinated crop, out crossing to an extent of 5%~9% by insects is reported which renders considerable genetic diversity (Duggi et al., 2013). Hence, the first step in okra improvement should involve evaluation of the germplasm for genetic variability. As a second step, it is required to generate crosses employing a suitable mating design to know the extent of heterosis for various economic traits and inheritance pattern of desired characters, which in turn, would help in deciding the breeding strategies as well as identifying potential parents and crosses for further use in breeding programme (Singh and Singh, 2012).

Results and Discussion
In the present investigation, fifteen F₁ hybrids derived from five lines and three testers were evaluated using Line×Tester analysis with one check variety (Varsha Upahar). Significant variation was observed among treatments for all the characters. Line×Tester interaction mean squares was significant for all the characters (Table 1). Lines varied significantly for days to first flowering and fruit girth while testers exhibited non-significant variation for all ten characters.

Table1 ANOVA for Line×Tester analysis in okra

Heterosis was estimated as percent increase or decrease of F₁ values over mid-parent (Relative Heterosis), better parent (Heterobeltiosis) and standard check variety (Standard Heterosis); Varsha Upahar. The nature and magnitude of heterosis are presented in Table 2. Ranking of three desirable hybrids on the basis of per se performance, standard heterosis and heterobeltiosis for various yield and yield contributing characters in okra are given in Table 3.

Table 2 Heterosis (%) over mid parent, standard variety and better parent for various yield and yield contributing characters in okra (RH- Relative heterosis; SH- Standard heterosis and HB- Heterobeltiosis)

Table 3 Ranking of three desirable hybrids on the basis of per se performance, standard heterosis and heterobeltiosis for various yield and yield contributing characters in okra

For leaf axils bearing first fruit, two hybrids exhibited desirable significant heterosis over mid parent and the highest percentage of significant positive relative, standard and better parent heterosis was noticed for Halu Bhendi×Punjab Phalgani (43.34%, 22.16% and 35.48%) closely followed by Holavanalli Local× Mallapalli Local (20.93%, 21.00% and 15.55% ). Rewale et al (2003) and Hosamani et al (2008) reported positive heterosis for this trait.

Relative heterosis was significant and positive for number of primary branches in one hybrid, Holavanalli Local×Mallapalli Local (27.39%). Among the fifteen hybrids, six hybrids exhibited significant and positive standard heterosis over standard variety and the hybrid Holavanalli Local×Mallapalli Local had the maximum standard heterosis (102.61%). Saha and Kabir (2001), Rewale et al (2003), Hosamani et al (2008) and Kumar (2011) reported positive heterosis for this trait.

The hybrid Holavanalli Local×Mallapalli Local recorded significant and positive relative and standard heterosis (5.19%) for plant height. Rewale et al (2003), Singh et al (2004), Hosamani et al (2008) and Kumar (2011) revealed similar result for this trait.

The short duration varieties are of breeder’s interest to fit them into rainfed cultivation, therefore, heterosis in negative direction is desirable for days to maturity. All the hybrids showed negative and significant heterosis over mid parent, s tandard variety and better parent. These findings are in agreement with those of Rewale et al (2003) Hosamani et al (2008) and Kumar (2011).

Relative and standard heterosis for number of fruits was positive and significant for four hybrids. Heterobeltiosis was positive and significant for only one hybrid Holavanalli Local×Mallapalli Local (5.24 %). The hybrid, Holavanalli Local×Mallapalli Local, expressed positive significant heterosis over mid parent, standard variety and better parent for this character. Positive heterosis for number of fruits per plant is reported by Rewale et al (2003), Singh et al (2004), Hosamani et al (2008) and Kumar (2011).

The fruit weight had a positive and significant heterosis over the mid parent for four hybrids. Among fifteen hybrids evaluated, thirteen hybrids exhibited positively significant standard heterosis for fruit weight. The crosses Thirumala Local×Kattakada Local and Holavanalli Local×Mallapalli Local had positive significance for fruit weight over mid parent, standard variety and better parent. Singh et al (2004), Hosamani et al (2008), Senthil Kumar and Sreepar- vathy (2010) and Kumar (2011) revealed similar result for this trait.

For the fruit length eleven hybrids showed positively significant heterosis over mid parent. Positively significant heterosis over check variety was exhibited by all the fifteen hybrids. Thirteen hybrids exhibited significant positive heterobeltiosis for this trait. Similar results were presented by Singh and Singh (1979), Poshiya and Shukla (1986), Metwally and Etsamy (1990), Saha and Kabir (2001), Rewale et al (2003), Hosamani et al (2008) and Senthil Kumar and Sreeparvathy (2010).

The three hybrids had positive significant heterosis for fruit girth over mid parent. Positively significant standard heterosis over standard variety for fruit girth was observed for six hybrids. Heterobeltiosis was positively significant for only one hybrid with regard to fruit girth. Similar Positive heterosis for this trait is reported by Saha and Kabir (2001), Rewale et al (2003), Singh et al (2004) and Hosamani et al (2008).

Fruit yield is a complex trait. It is the end product of several basic yield components. The standard heterosis is more useful from practical point of view. All the fifteen hybrids exhibited significant relative heterosis and of which seven hybrids displayed significant positive standard heterosis for yield per plant. The cross Holavanalli Local×Mallapalli Local had the maximum significant positive value (260.03%) followed by Thirumala Local×Kattakada Local (140.32%), Kunnapuzha Local×Punjab Phalgani (109.84%) and Thirumala Local×Punjab Phalgani (67.15%). Heterobeltiosis was significant and positive for Holavanalli Local×Mallapalli Local (23.82%) and Thirumala Local×Kattakada Local (21.27%)_. The hybrids which showed positive and significant estimates for all three types of heterosis were Holavanalli Local×Mallapalli Local and Thirumala Local×Kattakada Local. These findings are in agreement with those of Rewale et al (2003), Singh et al (2004), Hosamani et al (2008), Senthil Kumar and Sreeparvathy (2010) and Kumar (2011).

In the present study, it was found that among the fifteen hybrids, the two hybrids Holavanalli Local×Mallapalli Local and Thirumala Local× Kattakada Local exhibited significant standard heterosis for all the characters and found promising for fruit yield and other traits. Hence, these hybrids need to be tested on larger scale for their commercial values and can be released for commercial cultivation.

Materials and Methods
The experimental material consists of fifteen F₁ hybrids developed from five lines (IC 1012-1, Thirumala Local, Halu Bhendi, Kunnapuzha Local and Holavanalli Local) and three testers (Punjab Phalgani, Kattakada Local and Mallapalli Local) in a Line×Tester fashion. The fifteen F₁ hybrids and their eight parents were evaluated along with one commercial check (Varsha Upahar) in a randomized block design with three replications at spacing of 60×45 cm and ten plants per treatment per replication during summer 2012.

Observations on yield and yield attributes viz., days to first flowering, leaf axil bearing first fruit, number of primary branches, plant height (cm), days to maturity, number of fruits per plant, fruit weight (g), fruit  length (cm), fruit girth (cm) and yield per plant (g) were recorded from eight parents and fifteen F₁ hybrids. The data was subjected to Line×Tester analysis and magnitude of heterosis was calculated as per standard procedure and significance of heterosis was worked out using the formula suggested by Wynne et al (1970).

Authors’ Contributions
DSK participated in design of study and collection and analysis of data. KA involved in conception and design of the study and helped in manuscript preparation. SD involved in collection of data, performed statistical analysis and helped in drafting the manuscript. SM participated in analysis and interpretation of data and drafting the manuscript. NRR helped in drafting the manuscript. CV participated in layout of the experiment and management of the crop. PSR involved in data collection and helped in statistical analysis. All authors read and approved the final manuscript.