Frequency of heterotic hybrids in relation to general combining ability of parents in sweet corn

Experimental material and site

Seven inbred lines, MRCSC9, KH1831, SC Sel 1, SC Sel 2, SC Sel 3, SC Syn and SC Ind (Table 1), were obtained from the Winter Nursery Centre, Hyderabad, ICAR-Indian Institute of Maize Research. During rabi 2018–19, these lines were crossed in full diallel fashion at the ‘F’ block of the Main Agricultural Research Station, Dharwad, India, based on flowering synchrony. During kharif 2019, the seeds of forty-two hybrids, seven parental lines and checks hybrids Madhuri, Central Maize VL Sweet corn 1 and Misti were seeded in a randomized full block design with three replications at the ‘F’ block of the Main Agricultural Research Station, Dharwad which is located at an altitude of 750 m above Mean Sea Level (MSL) and at 150 49′N latitude and 740 99′E longitude. Parental lines and F₁-hybrids were raised in independent blocks and randomized separately. Each entry was raised in a two-row plot with a length of 4 m and a row spacing of 0.6 m. To ensure an optimal plant population in the experimental field, two seeds were dibbled at 0.2 m intervals at each hill, and seedlings were thinned to only one seedling per hill 15 days after sowing. All the standard cultural activities were carried out to produce a satisfactory crop.

Weather condition during the experimental period

The average maximum and minimum temperatures during the cropping period were 30.9 °C and 18.9 °C, respectively with a monthly mean of 26.9 °C and 20.3 °C (Table 2). It is important to note that 588.8 mm rainfall was received during experimental period against the long term average rainfall of 344.7 mm. Although, excess rainfall was received, growth stages of the crop were not affected due to even distribution of rainfall in the cropping period and draining of excess water on heavy rainy days to avoid crop stress due to excess soil moisture.

Recording of data

To record findings, five competitive plants were labelled at random in each entry throughout all replications. Flowering traits like days to 50% tasseling (DFT) and silking (DFS); growth parameters like plant height (PH) and ear height (EH); yield characters such as kernel rows per ear (KRN), resistance to Turcicum leaf blight disease (TLB), kernels per row (KPR), ear length (EL), ear girth (EG), green ear yield (GEY) and de-husked ear weight (DEW), brix content of freshly harvested selfed seeds (TSS), and green fodder weight (GFW) were recorded according to standard procedures (Vanipraveena et al., 2021).

Statistical analysis

The data collected from seven inbred lines and forty-two hybrids in each entry was first used to perform analysis of variance (Panse & Sukhatme, 1985). The total variation of crosses was divided into distinct sources like parents, hybrids, and parents versus hybrids (Griffing, 1956). For each character, the general combining ability (GCA) impacts of parental lines and the specific combining ability (SCA) of crosses were calculated (Kempthorne, 1957). Heterosis of crosses over better-parent and standard check was estimated by following well established methodologies (Turner, 1953; Hayes, Immer & Smith, 1955).

Percent heterosis over mid-parent (%), MPH = $\frac{\overline{F}1-\overline{M}P}{\overline{M}P}\times 100$

Percent heterosis over better-parent (%), BPH = $\frac{\overline{F}1-\overline{B}P}{\overline{B}P}\times 100$

Percent heterosis over the standard check (%), SH = $\frac{\overline{F}1-\overline{S}C}{\overline{S}C}\times 100$

Error mean sum of square (Me) was used to compute standard error (SE) for better parent and mid-parent heterosis = (2 Me/r)^1/2 and SE for standard heterosis = (2 Me/r)^1/2. Further significant deviation of F₁ from better parent (BP) and standard check (SC) was worked out by calculating ‘t‘ value using standard error by following standard method (Arunachalam, 1974).

‘t’ value for mid-parent heterosis = $\frac{\overline{F}1-\overline{M}P}{SE\left(\overline{M}P\right)}$

‘t’ value for better-parent heterosis = $\frac{\overline{F}1-\overline{B}P}{SE\left(\overline{B}P\right)}$

‘t’ value for standard heterosis = $\frac{\overline{F}1-\overline{S}C}{SE\left(\overline{S}C\right)}$

Estimation of variance components for combining ability

Variance due to GCA = $\frac{1}{P-1}\sum g_i^2=\frac{Mg-M{e}^{{}^{\prime }}}{2p}$

Variance due to SCA = $\frac{1}{p\left(P-1\right)}\sum \sum s_{ij}^2=Ms-M{e}^{{}^{\prime }}$

Estimation of combining ability effects

GCA effect of parents (g_i) = $\frac{1}{2p}\left(Yi.+Y.j\right)-\frac{1}{p^2}$ Y.SCA effect of hybrids (sij) = $\frac{1}{2p}\left(Yij.+Yji\right)-\frac{1}{2p}\left(Yi.+Y.j+Y.i+Yj.\right)+\frac{1}{p^2}Y..$

where, Y.. = Grand total of all hybrids, Yi. = Row total of i^th parent, Y.i = Column total of i^th parent, Y.j = Column total of j^th parent, Yj. = Row total of j^th parent, Yij = Mean value of ij^th hybrid, Yji = Reciprocal value of ij^th hybrid, P = number of parents involved

Significance of combining ability effects

The following formulas were used to determine the standard errors for testing the significance of GCA and SCA effects.

SE for GCA effects of parents = $\sqrt{\frac{\left(p-1\right)}{2p^2}}M{e}^{\prime }$

SE for SCA effect of hybrids = $\sqrt{\frac{p^2-2p+2}{p^2}}M{e}^{\prime }$

To calculate the relevant critical difference (CD) values, the SE value was multiplied by (2)^1/2 and the table ‘t’ value by 5% and 1%, respectively.

CD = (2)^1/2 (SE) (table ‘t’ value for error degrees of df) at 5% and 1%, respectively.

Estimation of overall GCA status of parents, SCA status and heterotic status of crosses

It is crucial to comprehend the overall state of the parents and hybrids while taking into account the GCA and SCA impacts for all characters at once because yield is related to many additional variables, some beneficially and some negative. The overall performance of a parental line or crossing in terms of GCA or SCA impacts was determined (Arunachalam & Bandyopadhyay, 1979) using established techniques with minor modifications (Mohan Rao, 2001). Based on the comprehensive overall GCA status of the parents, the crosses were classified as H × H, H × L, L × H, and L × L. To draw inferences, the overall SCA status and heterosis status of crosses were counted and mentioned under each group. Since the number of crosses in each cross group varied, the ratio of crosses with high overall heterotic and SCA status in each crossing group to the whole number of crosses with high overall SCA and heterotic status was calculated, expressed as the conditional probability of a cross with high overall SCA and heterotic status belonging to a specific cross group.

Per se performance of parental lines

The inbred line SC Sel 2 (61.0 days) taken minimum days to flowering followed by SC Sel 3 and KH1831, whereas MRCSC9 taken maximum days (Table 3). Green ear yield in parents ranged from 2.89 tons/hectare to 5.15 tons/hectare with an average yield of 4.32.The line SC Ind (5.15 tons /hectare) recorded highest green ear yield followed by SC Sel 3 (5.13 tons/hectare), MRCSC 9 (4.82 tons/hectare) and SC Sel 1 (4.65 tons/hectare). The inbred line SC Sel 3 (4.11 tons/hectare) manifested highest de-husked ear weight followed by SC Sel 1 (3.86 tons/hectare). These two lines also displayed maximum brix content (12.70% and 12.53% respectively) with minimum disease incidence for Turcicum leaf blight.

Variation among genotypes for quantitative traits

There were significant differences in total genotypes and hybrids (p ≤ 0.01) for all the variables studied (Table 4). Significant differences were observed among the parents for all the traits with exception of plant height, kernel rows per ear, number of kernels per row, de-husked ear weight and green ear yield. The mean sum of squares of parents versus hybrids was significant for ear height and number of kernel rows.

Variance resulting from combining ability effects

It is widely assumed that general combining ability is the result of additive gene effects and additive epistatic variance components. On the contrary, non-additive gene effects and the residual epistatic variation lead to specific combining abilities (Matzinger, Sprague & Cockerham, 1959). In relation to all thirteen traits, Table 5 shows the estimates of variations attributable to GCA, SCA, and the ratio of GCA and SCA. Sufficient variation for combining ability was observed for the traits indicating the presence of both additive and non-additive gene action for the inheritance of the concerned characters. GCA variance ranged from 0.01 (for ear girth) to 22.41 (plant height) and SCA variance ranged from 0.08 (for ear girth) to 90.09 (plant height). It is evident that the variance due to SCA outweighed the GCA variation for all of the characteristics that were investigated. SCA and GCA variance ratios ranged from 0.07 (for kernel row number) to 0.76 (for kernels per ear). For kernel rows per ear, green fodder weight, ear girth, brix content, ear-height, resistance to Turcicum leaf blight, green-ear yield and plant height, the SCA:GCA ratios were wider. Estimates of additive and dominance variation revealed that dominance variance was prominent for these parameters, with dominance variance being greatest for plant height, resistance to Turcicum leaf blight, ear-height, kernel rows per ear, and green fodder weight (Fig. 1). However, both additive and dominant variances were discovered to be significant for the rest of the characteristics.

Overall GCA status of parents

The magnitude of GCA effects of the seven parents differed substantially for each parameter, and none of the seven parents were demonstrated to be an efficient generalized combiner for all thirteen variables examined Table S1. Nevertheless, results of overall GCA status of parents showed that three lines (SC Sel 2, SC Sel 1 and SC Sel 3) manifested higher total score compared to final mean norm of inbred parental lines (52.0) suggesting that these three lines (Fig. 2) as good general combiners across the traits as evident from their maximum total score and high (H) overall GCA status among seven parents.

Overall SCA status of crosses

Based on the overall GCA status of their respective parental lines, the resulting hybrids were classified as ‘high × high’, ‘high × low’, ‘low × high’ and ‘low × low’. In order to derive the conclusion, the overall SCA status of crosses combinations, namely ‘High’ or ‘Low’, was also indicated under each group. Out of forty two crosses, eighteen manifested high overall specific combining ability with more than 279.1 final norm score (Table 6). Among these eighteen crosses, five crosses, SC Sel 1 × MRCSC9 (378.0), SC Sel 3 × SC Sel 1 (345.0), SC Sel 3 × SC Syn (331.5), SC Sel 1 × SC Ind (298.0) and SC Sel 2 × SC Sel 3 (289.0) had ‘high’ overall general combiner (SC Sel 1 or SC Sel 2 or SC Sel 3) as female parents. Furthermore, fifteen out of eighteen high specific combining crosses, SC Ind × SC Sel 1 (436.5), SC Ind × SC Sel 2 (434.0), SC Ind × SC Sel 3 (405.0), KH 1831 × SC Sel 2 (384.5), SC Sel 1 × MRCSC9 (378.0), SC Sel 3 × SC Sel 1 (345.0), MRCSC 9 × SC Sel 1 (341.0), SC Syn ×SC Sel 2 (337.5), KH 1831 × SC Sel 3 (333.3), SC Sel 3  × SC Syn (331.5), KH 1831  × SC Sel 1 (305.5), SC Syn   × SC Sel 3 (303.5), SC Sel 1  × SC Ind (298.0), SC Sel 2 × SC Sel 3 (289.0) and MRCSC 9 × SC Sel 2 (286.0) had either one or both of the parents in the cross-combination as ‘high’ overall general combiner. The cross SC Ind × KH1831 (L × L) showed highest overall specific combining ability status followed by SC Ind × MRCSC9 (L × L), SC Ind × SC Sel 1 (L × H), SC Ind × SC Sel 2 (L × H) and SC Ind × SC Sel 3 (L × H). Among the crosses, SC Ind × KH1831 (L × L) was identified as a best specific combiner for (nine traits) plant height, ear-height, ear length, ear girth, kernel rows per ear, kernels per row, green fodder weight, TSS and green-ear yield (Table S2). The cross SC Ind × MRCSC 9 (L × L) was the second best specific combiner (4 traits) for DFT, DFS, KPR and DEW. For traits DFT, DFS and GFW the cross, SC Sel 3  × SC Sel 1 (H × H) was found to be good specific combiner.

Overall heterotic status of hybrids

Among forty-two experimental hybrid combinations, twenty-two crosses displayed high overall mid-parent heterotic status (Table 7) with more than 279.5 final norm of crosses. Among these twenty-two crosses, ten crosses had ‘high’ overall general combiner (SC Sel 1 or SC Sel 2 or SC Sel 3) as female parent in the cross combination. Furthermore, fifteen out of twenty-two high overall mid-parent heterotic crosses had either one or both of the parental lines with ‘high’ overall general combiner in the cross-combination. Five crosses SC Syn × KH1831 (L × L), MRCSC 9 × SC Sel 2 (L × H), KH 1831 × SC Sel 3 (L × H), KH 1831 × MRCSC9 (L × L) and MRCSC 9 × KH 1831 (L × L) manifested highest overall mid-parent heterotic status (Table 7). With respect to overall better-parent heterotic status (OBPHS) of hybrids, twenty-four crosses manifested high OBPHS. Furthermore, thirteen out of these twenty-four crosses had ‘high’ overall general combiners (SC Sel 1 or SC Sel 2 or SC Sel 3) as female parent and eighteen crosses had either one or both of the parents in the cross-combination as ‘high’ overall general combiner (Table 8). The crosses SC Sel 2 × SC Syn (H × L) followed by SC Sel 2 × KH1831 (H × L), MRCSC 9 × SC Sel 2 (L × H), KH 1831 × SC Syn (L × L) and SC Sel 2 × SC Sel 3 (H × H) showed high overall better-parent heterotic status (Table S3).

Mid-parent heterosis was significant for the majority of the crossings for all attributes investigated except DFS, DFT, TLB, and TSS, whereas substantial economic heterosis was noted in the lowest percentage of crosses. Determining overall heterotic status for standard heterosis was felt important besides estimating overall GCA and SCA status. Hence by following same method used to compute overall mid parent and better parent heterotic status across the traits, overall standard heterotic status was also calculated (Table S4). Twenty out of forty two crosses documented high overall standard heterotic status (OSHS). Among these twenty crosses, ten had ‘high’ overall general combiners (SC Sel 1 or SC Sel 2 or SC Sel 3) as female parent and sixteen crosses had either one or both of the parents in the cross-combination as ‘high’ overall general combiner (Table 9) Among the crosses, MRCSC 9 × SC Sel 2 (L × H) showed high standard heterosis for DFT, DFS, EG and GFW. For traits KRN and GEY, the cross SC Sel 2 × SC Sel 3 (H × H) manifested high economic heterosis. Among forty-two novel hybrid combinations, five crosses SC Sel 2 × SC Syn (H × L) followed by SC Syn × KH1831 (L × L), SC Sel 2 × SC Sel 3 (H × H), SC Sel 3 × SC Sel 1 (H × H) and SC Sel 2 × SC Sel 1 (H × H) showed high overall standard heterotic status.

The top five best cross combinations of higher specific combining ability and higher standard heterosis were compared for all thirteen characters studied (Table 10). Only in DFT, DFS and GFW traits, 2–3 crosses exhibited higher SCA where one or either of the parents involved had ‘high’ overall GCA status, while only cross combination revealed higher SCA for EH, EL, TLB, TSS and GEY traits. Interestingly, 3–5 crosses displayed higher standard heterosis for all the traits except DFT. The prevalence of heterotic crossings has been assessed in connection with the overall GCA and SCA status of parental lines and hybrids. Study revealed that the maximum probability of obtaining high overall SCA crosses in our investigation was 0.55 in ‘L × H’ parental GCA combination, where as it was only 0.16 in ‘H × L’ and ‘L × L’ and 0.11 in ‘H × H’ parental GCA combinations (Table 11). In contrast to this, the probability of obtaining high overall heterotic hybrids was ranged between 0.83 to 1.00 in ‘H × H’ parental GCA combination followed by 0.41 to 0.58 in ‘H × L’, 0.41 to 0.50 in ‘L × H’ and 0.33 to 0.50 in ‘L × L’ combination.