SciELO - Scientific Electronic Library Online

vol.50 issue2Optimum experimental plot size in the castor beanMultivariate analysis and vigor tests to determine the quality of Brachiaria decumbens seeds author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Ciência Agronômica

Print version ISSN 0045-6888On-line version ISSN 1806-6690

Rev. Ciênc. Agron. vol.50 no.2 Fortaleza Apr./June 2019 

Crop Science

Diallel analysis and heterosis components in paprika peppers1

Análise dialélico e componentes da heterose em páprica

André Lasmar2

Wilson Roberto Maluf3

César Augusto Ticona-Benavente4  *

Douglas William Nogueira2

Danilo Gustavo Nogueira2

2Terras Gerais, Sitio Campo Lindo, Lavras-MG, Brasil, (ORCID ID 0000-0003-1733-5133)

3Departamento de Agronomia, Universidade Federal de Lavras, Lavras-MG, Brasil, (ORCID ID 0000-0002-3815-0405), (ORCID ID 0000-0002-2005-6533), (ORCID ID 0000-0002-1412-7569)

4Instituto Nacional de Pesquisas da Amazônia, Coordenação de Sociedade, Ambiente e Saúde, Manaus-AM, Brasil, (ORCID ID 0000-0002-3636-2324)


The aim of this study was to quantify and assess the components of heterosis in paprika hybrids related to yield, capsanten pigment content (estimated as ASTA degrees) and resistance to Phytophthora capsici, and to identify parental lines with high general combining ability (GCA) to be used in future breeding programmes and/or to obtain new hybrids. Fifteen hybrids were obtained through a complete diallel cross (reciprocal hybrids excluded) among six proprietary paprika breeding lines from HortiAgro Sementes S.A., four of which originally introduced from Peru (P1=PIM 032-03; P2=PIM 033-11; P3=PIM 034-19; P4=PIM 035-01) and two from the U.S.A. (P5=PIM 036-08; P6=PIM 037-18). Epistatic gene action was involved in the expression of heterosis for fresh and dry yields and carotenoid pigment contents, and heterosis was predominantly in the direction of higher yields. No significant heterosis effects were detected for resistance to P. capsici, and gene action was of incomplete dominance for the resistant phenotype. The parental lines P1 and P5 showed high GCA values for all characters, and may be used in breeding programmes to obtain new improved lines or for the production of higher yielding hybrids. The most promising hybrid was P3xP5, which outperformed the standard cultivar “Papri Queen” in fresh (68 t ha-1) and dry (10 t ha-1) fruit yields and ASTA degrees (157), in addition to being resistant to P. capsici. The parental line P5, as well as the hybrids in which it participated as a parent were assessed as resistant to P. capsici.

Key words: Capsicum annuum; General combining ability; Phytophthora capsici; Hybrid vigor; Gene action; Epistasis


O objetivo deste estudo foi quantificar e avaliar os componentes da heterose em híbridos de páprica para produtividade, teor de capsanteno (graus ASTA) e resistência a Phytophthora capsici, e identificar as linhagens parentais elevada capacidade geral de combinação (CGC) para serem utilizados em futuros programas de melhoramento e/ou para obter novos híbridos. Para tanto, 15 híbridos foram obtidos por cruzamentos dialélicos (excluindo os recíprocos) entre seis linhagens from Hortiagro Sementes S.A. Elas são oriundas do Peru (P1=PIM 032-03; P2=PIM 033-11; P3=PIM 034-19; P4=PIM 035-01) e dos EUA (P5=PIM 036-08; P6=PIM 037-18). Os resultados indicam que a ação génica epistática controla a heterose de produtividade de frutos frescos e secos, e o teor de capsantenos, sendo, em geral, a heterose no sentido de aumentar a produtividade. A resistência a P. capsici não teve efeito heterótico significativo, sendo explicado por dominância incompleta no sentido da Resistencia. As linhagens parentais P1 e P5 mostraram elevada CGC para os caracteres avaliados, e poderiam ser utilizadas em programas de melhoramento para obter novas linhagens melhoradas ou para a produção de híbridos com elevada produtividade.. O cruzamento P3xP5 superou à testemunha ‘Papri Queen’ em produtividade de frutos frescos (68 t ha-1), secos (10 t ha-1), graus ASTA (157) e resistência a P. capsici. A linhagem P5 assim como nos híbridos onde participou como parental foram avaliados como resistentes.

Palvras-chaves: Capsicum annuum; Capacidade geral de combinação; Phytophthora capsici; Vigor híbrido; Ação genica; Epistasis


The presence of heterosis in paprika set the possibility to create hybrid cultivars (REIFSCHNEIDER; RIBEIRO; CARVALHO, 2013; SOMOGYI, 2010; SOMOGYI et al., 2011; SURYA KUMARI et al., 2014). The knowledge of the F1 hybrids behavior comparing to their parent cultivars let the breeder choose the best genetic combinations between the paprika lines for the considering trait.

The efficiency in the obtaining process of lines or cultivars that could be used for hybrid seeds production demands the knowledge of the genetic effects involved in the traits determination. One difficulty faced in the selection process is the non-information about the inheritance of the quantitative traits. The achievement of this information may enable highest genetic gains, increasing the breeding programs efficiency. Diallel crosses are important ways to obtain this information, providing useful parameters estimative in the selection of parents for hybridization (CRUZ; REGAZZI; CARNEIRO, 2012). The analyzing method proposed by Gardner and Eberhart (1966) preview a detailed study of the heterosis and it’s components, providing a fast way to evaluate the parent lines potential for hybrids achievement.

Amongst the available diallel analyzing methods, the one proposed by Jinks and Hayman (1953) bases in the environmental nature and statistics genetic knowledge (averages, variances and covariances) obtained from a diallel table, providing information about the studying trait genetic control, the genetic values and selection limits of the studying traits (CRUZ; REGAZZI; CARNEIRO, 2012; VENCOVSKY; BARRIGA, 1992).

Somogyi (2010), tested experimental paprika hybrids in Hungry and related that for dry matter and ground color obtained by the powdering of dry fruits, the hybrids showed intermediary values related to the parents averages. However, the hybrids showed fruit production values five to six times higher than the obtained by the open population cultivars, defending the use of the hybrids for higher gains in yield for area unit.

The aim of this study was to evaluate the heterosis components in paprika hybrids for productivity, carotenoids pigments amount and resistance to Phytophthora capsici, and identify, between the used lines, parents with good general combining ability, that could be used in future breeding programs or to obtain new hybrids.



Six lines from Hortiagro Seeds S.A.’s Capsicum annuum L. breeding program with paprika characteristics were used to obtain the experimental F1 hybrids: four from Peru ((P1=PIM 032-03; P2=PIM 033-11; P3=PIM 034-19; P4=PIM 035-01) and two from the USA (P5=PIM 036-08; P6=PIM 037-18). The lines were previously selected by the tasting test through two self-generations and characterized as non-pungent (no capsaicin).

The crosses were made manually in green houses with all the lines, through flower buds emasculations followed by controlled pollination, by the diallel scheme (without the reciprocal), obtaining 15 experimental hybrids.

Agronomic test

Fifteen hybrids from the previous step were tested, along the six parental lines (P1=PIM 032-03; P2=PIM 033-11; P3=PIM 034-19; P4=PIM 035-01; P5=PIM 036-08; P6=PIM 037-18) and the open population cultivar “Papri Queen” used as control, totalizing 22 treatments.

The experiment was carried out at the Vegetable Research Station of Hortiagro Sementes S.A., Ijaci-MG-Brasil (21°14’16” S, 45°08’00’’ W, altitude 920 m). Seeds were sown on October 23rd of 2011 in 128 cells Styrofoam trays and, 30 days after, they were transplanted to plastic covered beds. A random complete block design was used with four replications and 11 plants in each parcel, totalizing 968 plants. Was used the spacing of 0.75 m between lines and 0.35 m between plants in the line, resulting in a density of 38 thousand plants per hectare.

In the end of four consecutive harvestings (03/01/2012, 03/28/2012, 05/02/2012 and 05/31/2012) per experimental plots, done when the fruits achieved maturity and were completely red, was estimated the fresh fruits total production.

Samples of approximately 800 g of fresh fruits from each parcel were taken for the determination of dry weight. After measured their initial weight (fresh weight), the fruits were dried in laboratory oven with forced air circulation for five days at 50 ˚C until reach constant weight. The sample’s weights were measured again and then estimated the percentages of dried matter that, multiplied by their fresh fruit yield, resulted in the dried fruit estimative.

The dried samples were stored in paper bags for 90 days at room temperature and later they were powdered in electric grinder (MARCONI MA 340).

The fruits capsantin content of each plot was determined by the ASTA 20.1 methodology (AMERICAN SPICE TRADE ASSOCIATION, 2004). The absorbances were measured, in spectrophotometer regulated to wave length of 460 nm, of each plot and the data were expressed in ASTA units.

Inoculation and resistance evaluation to Phytophthora capsici

The genotypes reactions to the pathogen P. capsici were determined in a side experiment, by inoculation with the pathogen. The plants were inoculated with solution containing a mix of the P. capsici isolates “Pc11” and “Pc31” when they achieved 10 cm high in a random complete block design with four replications and eight plants per plot. These isolates were obtained with the Sakata/Agroflora company, Bragança Paulista, SP, and were originally collected in the region of Bernardino de Campos, SP and Santa Cruz do Rio Pardo, SP, respectively. The isolates were kept, grown and cultivated following the methodology according to Urben (1980). After seven days, were deployed with Drigalsky handle. To release zoospores, the sporangia suspension was kept under incubation at room temperature for one hour. The suspension was then filtered into tissue and withdrawing a portion of filtrate for counting the zoospores amount in a Neubauer chamber. To stimulate zoospores encystment, the suspension was vortexed for one minute. After counting and established dilution at a concentration of 104 zoospores/5 ml the zoospores suspension was immediately used.

On January 15, 2012, at 45 days after seeds sown, the seedlings in the Styrofoam trays were inoculated with 5ml of the suspension in each cell, close to the plants stem. Daily evaluations were made after the inoculation, using a scale of scores (1 = no symptoms, 2 = wilting and necrosis and 3 = plant death). The plant mortality percentage was also registered. The evaluations started at the third day after inoculation until January 30, 2012, the 15th day after inoculation, date that was registered the scores and plants mortality percentage.

Statistics analyses

Dried and fresh fruit yield, ASTA degrees, scores and mortality rate data were analyzed according to the diallel model of Gardner and Eberhart (1966) assuming the fixed model:



yij: average value of genotype ij.

µ : general average;

vi e vj: i-th or j-th variety effect, respectively;

h: average heterosis

hi e hj: i-th and j-th variety heterosis effect, respectively;

sij: specific heterosi’s effect

eij: average experimental error

The variety effects (vi), average heterosis (¯h) and varietal heterosis (hi) from Gardner and Eberhart (1966) were analyzed in equal to the concepts of general and specific combining ability from Sprague and Tatum (1942):

gi = 1/2vi + + hi


gi = parental i’s general combining ability

vi = variety’s effect

= average heterosis

hi = varietal heterosis

The data were also graphically analyzed according to the diallel analyzing model proposed by Jinks and Hayman (1953), wherein the regression coefficient estimative “β” of Wr (parental line progeny’s “r” covariance with the non-recurrent parental) in Vr (parental line’s “r” variance) other than 1 indicates presence of epistasis; and, if not, its absence. Duncan test was used for averages comparison, with 5% probability.


Fresh fruits yield

The treatments differed in fresh fruit total yield (Table 1). The average heterosis component, obtained by Gardner and Eberhart’s analysis, was significant, indicating that hybrids averages were in general significantly higher than the parent’s average, and that heterosis is predominantly unidirectional (in the way of higher yields) (Table 1). The varietal and specific heterosis components were not significant (Table 1), indicating that were not detected differences in the parents contribution to the heterosis and there were not differences in the parents allele frequency. The results indicate, therefore, that hybrids fresh fruit total yield may be estimated, in general, by the parent’s average added the corresponding value of the average heterosis estimated for all hybrids.

Table 1 Analyses of variance according Gardner and Eberhart (1966) for fruits yield and color, and Phytophthora capsici resistance 

Fonts of variation GL Mean Squares
Fresh fruit yield (t ha-1) Dried fruit yield (t ha-1) Color (ASTA degrees) P. capsici resistance (score) Mortality percentage caused by P. capsici
Treatments 20 276.48** 5.24* 1771.04** 0.4966** 1583.83**
Varieties 5 302.13* 4.47ns 1419.24ns 1.7231** 5382.87**
Heterosis 15 267.93* 5.49* 1888.31** 0.0877ns 317.49ns
Average Heterosis 1 918.11** 19.46* 235.04ns 0.0032ns 179.46ns
Varietal Heterosis 5 219.40ns 4.08ns 1368.49ns 0.1160ns 551.36ns
Specific Heterosis 9 222.66ns 4.72ns 2360.80** 0.0814ns 202.89ns
Error 60 120.06 2.87 748.12 0. 1359 403.05
Coefficient of Variation (%) 21.82 20.93 24.58 18.05 33.39

*, **(P<0.05) e (P<0.01) respectively;

nsno significant;

score:1 = no symptoms, 2 = wilting and necrosis and 3 = plant death

The high positive value and significance of the GCA estimatives, according with the equivalence of Sprague and Tatum (1942) concepts (Table 2), to a line’s fresh fruit total yield are important indicatives of their potentiality in generating good populations, because indicates high frequency of additive nature favorable alleles (CRUZ; REGAZZI; CARNEIRO, 2012). Thus, the parents P1 and P5, showing GCA of 11.3 and 16.1 t ha-1, respectively (Table 2), have good potential in being utilized in breeding programs, looking for new lines selection and/or to obtaining more productive hybrids.

Table 2 Estimates of average (µ), variety effect (vi), average (-h) and varietal heterosis (hi), general combination ability (gi) and specific heterosis (sij) for fruits yield, color and Phytophthora capsici resistance 

Mean components Fresh fruit yield (t ha-1) Dried fruit yield (t ha-1) Color (ASTA degrees) P. capsici resistance (score) Mortality percentage caused by P. capsici
µ 45.0 (± 2.23) 7.33 (± 0.34) 113.9 (± 5.67) 2.03 (± 0.08) 57.4 (± 4.73)
P1 2.01 (± 5.00) 0.40 (± 0.77) -6.31 (± 12.69) -2.80 (± 0.19) -11.6 (± 10.6)
P2 6.54 (± 5.00) 1.25 (± 0.77) 11.2 (± 12.69) 0.11 (± 0.19) 3.87 (± 10.6)
P3 -5.84 (± 5.00) -1.05 (± 0.77) 1.81 (± 12.69) 0.30 (± 0.19) 17.5 (± 10.6)
P4 -3.03 (± 5.00) -0.56 (± 0.77) -17.7 (± 12.69) 0.60 (± 0.19) 30.0 (± 10.6)
P5 1.80 (± 5.00) -0.15 (± 0.77) -10.4 (± 12.69) -0.70 (± 0.19) -36.6 (± 10.6)
P6 -1.48 (± 5.00) 0.12 (± 0.77) 21.4 (± 12.69) -0.03 (± 0.19) -3.27 (± 10.6)
7.32 (± 2.64) 1.07 (± 0.40) -3.71 (± 6.71) 0.02 (± 0.10) 3.74 (± 5.59)
P1 3.06 (± 3.53) 0.40 (± 0.54) 7.22 (± 8.97) 0.05 (± 0.13) -1.22 (± 7.48)
P2 -7.71 (± 3.53) -1.17 (± 0.54) -12.8 (± 8.97) 0.22 (± 0.13) 15.9 (± 7.48)
P3 -0.32 (± 3.53) -0.01 (± 0.54) -1.94 (± 8.97) 0.02 (± 0.13) 0.02 (± 7.48)
P4 -1.10 (± 3.53) 0.00 (± 0.54) -6.29 (± 8.97) -0.01 (± 0.13) 2.25 (± 7.48)
P5 7.97 (± 3.53) 0.99 (± 0.54) 22.7 (± 8.97) -0.21 (± 0.13) -13.7 (± 7.48)
P6 -1.90 (± 3.53) -0.20 (± 0.54) -8.92 (± 8.97) -0.07 (± 0.13) -3.26 (± 7.48)
gi = 1/2vi + h + hi (Sprague; Tatum, 1942)
P1 11.3 (± 8.67) 1.67 (± 1.32) 0.35 (± 22.02) -1.33 (± 0.33) -3.28 (± 18.4)
P2 2.88 (± 8.67) 0.52 (± 1.32) -10.8 (± 22.02) 0.29 (± 0.33) 21.6 (± 18.4)
P3 4.08 (± 8.67) 0.53 (± 1.32) -4.74 (± 22.02) 0.18 (± 0.33) 12.5 (± 18.4)
P4 4.70 (± 8.67) 0.79 (± 1.32) -18.8 (± 22.02) 0.30 (± 0.33) 21.0 (± 18.4)
P5 16.1 (± 8.67) 1.98 (± 1.32) 13.7 (± 22.02) -0.54 (± 0.33) -28.3 (± 18.4)
P6 4.68 (± 8.67) 0.93 (± 1.32) -1.90 (± 22.02) -0.06 (± 0.33) -1.15 (± 18.4)
1x2 3.64 (± 4.24) 0.46 (± 0.65) 1.54 (± 10.77) -0.06 (± 0.16) -1.19 (± 8.98)
1x3 -5.78 (± 4.24) -0.52 (± 0.65) -34.5 (± 10.77) 0.16 (± 0.16) 3.72 (± 8.98)
1x4 9.74 (± 4.24) 1.40 (± 0.65) 6.03 (± 10.77) -0.25 (± 0.16) -8.93 (± 8.98)
1x5 -9.63 (± 4.24) -1.47 (± 0.65) -2.01 (± 10.77) 0.10 (± 0.16) 9.82 (± 8.98)
1x6 2.03 (± 4.24) 0.13 (± 0.65) 28.9 (± 10.77) 0.04 (± 0.16) -3.43 (± 8.98)
2x3 -3.81 (± 4.24) -0.45 (± 0.65) -8.24 (± 10.77) 0.04 (± 0.16) 3.22 (± 8.98)
2x4 -5.74 (± 4.24) -0.77 (± 0.65) 17.4 (± 10.77) 0.15 (± 0.16) -0.50 (± 8.98)
2x5 5.88 (± 4.24) 0.99 (± 0.65) -21.4 (± 10.77) -0.13 (± 0.16) -9.52 (± 8.98)
2x6 0.03 (± 4.24) -0.23 (± 0.65) 10.7 (± 10.77) 0.00 (± 0.16) 7.98 (± 8.98)
3x4 1.06 (± 4.24) 0.17 (± 0.65) 12.6 (± 10.77) 0.13 (± 0.16) 8.57 (± 8.98)
3x5 11.2 (± 4.24) 1.42 (± 0.65) 30.2 (± 10.77) -0.16 (± 0.16) -4.62 (± 8.98)
3x6 -2.68 (± 4.24) -0.62 (± 0.65) -0.18 (± 10.77) -0.17 (± 0.16) -10.9 (± 8.98)
4x5 -6.58 (± 4.24) -1.23 (± 0.65) -1.70 (± 10.77) 0.01 (± 0.16) -0.60 (± 8.98)
4x6 1.51 (± 4.24) 0.43 (± 0.65) -34.4 (± 10.77) -0.04 (± 0.16) 1.44 (± 8.98)
5x6 -0.90 (± 4.24) 0.29 (± 0.65) -5.11 (± 10.77) 0.17 (± 0.16) 4.91 (± 8.98)

P1=PIM 032-03; P2=PIM 033-11; P3=PIM 034-19; P4=PIM 035-01; P5=PIM 036-08; P6=PIM 037-18;

score:1 = no symptoms, 2 = wilting and necrosis and 3 = plant death

In the other hand, SCA estimates may have important genetic meaning, both their signal and their relative magnitude. The SCA (sij) average components estimates ranged from -9.63 to 11.2 t ha-1 (amplitude of 20.8 t ha-1) (Table 2), which is fairly representative of the average (µ=45.0 t ha-1) and indicative that, beyond the additive genetic effects, the non-additives also may be important on traits expression of some hybrids. Indeed, however for the hybrid’s majority, the sij estimates were close to the pattern-error (and, so, non significant), some cases the sij value were significantly positive (s14 = 9.74; s35 = 11.2) or negatives (s15 = -9.63), which would indicates that, in these cases, the “per se” parental’s average is not a good indicative for the hybrids average performance to fresh fruit total yield.

The highest SCA’s positive estimates belong to the P3xP5 and P1xP4 hybrids with values of 11.2 e 9.74 t ha-1, respectively (Table 2). These same hybrids showed, beyond a high SCA value, at least one parental with high GCA value, which is desirable. The SCA is related, in your major parts, to the non-additive allele frequency differences, presenting an important variation source to the fresh fruit total yield in these hybrids.

The significance of the heterosis effects (Table 1) highlights the importance of non-additives effects, emphasizing the importance of the non-additive allele interactions in this trait control. Similarly, Shapturenko et al. (2014), Shrestha, Luitel and Kang (2011), Nascimento et al. (2010) and, Prasath and Ponnuswami (2008) also related significant heterosis for total fruit yield in Capsicum annuum, suggesting that non-additive effects are more important.

In addition, the non-additive effects nature means to be epistatic, once the β regression coefficient between Wr and Vr, measured by Jinks and Hayman (1953) diallel analyses, was significantly other than 1 (Table 3).

Table 3 Regression coefficients (β) for association of Wr and Vr from Jinks and Hayman (1953) diallel analyses for fruit yield and color and Phytophthora capsici resistance 

Characteristic β H0: β=0 H0: β=1 Epistasis
Fresh fruit yield (t.ha-1) 0.064403 ns ** sim
Dried fruit yield (t.ha-1) 0.123105 ns ** sim
Color (ASTA degrees) 0.042529 ns ** sim
P. capsici resistance (scores) 0.807188 ** ns não
Mortality by P. capsici (%) 1.075527 ** ns não

***: (P<0.01) and (P<0.05) respectively by t test.

nsno significant

Dried fruits yield

Differences were detected between the treatments for dried fruits total yield (Table1). The non-significant varietal heterosis indicates that were not detected differences in the parents contribution to the heterosis. The non-significant specific heterosis indicates that there not to be, in general, differences in the allele frequencies between the parents (Table 1).

The average heterosis’component was significant and positive, indicative that the hybrids yield was higher than the parent’s average (Table 1), whilst the varietal and specific heterosis’components were non-significant. The results indicate, however, that the hybrid’s dried fruit total yield may, in general, be estimated as the parent’s average plus the correspondent value estimated for the average heterosis ( ).

The parents P1 and P5, with GCA estimates of 1.67 and 1.98 t ha-1, respectively, may be used in breeding programs, looking for selection of new lines and/or to obtain more productive hybrids, according to the equivalence of Sprague and Tatum’s (1942) concepts mentioned before.

The estimates of the SCA average’s components (sij) ranged from -1.47 to 1.42 t ha-1 (amplitude of 2.89 t ha-1) (Table 2), which equates to about 40% of average (µ=7.33 t.ha-1) and indicative that, further the additive genetic effects, the non-additives may also be important on the trait’s expression. Similarly to the exposed for fresh fruit total yield, also for dried fruit total yield the majority of the sij estimates didn’t differ from zero. Just in the hybrids P1xP4 (s14 = +1.40) and P3xP5 (s35 = +1.42) the sij values were significantly higher than zero, whilst for P1xP5 (s15 = -1.47) and P4xP5 (s45 = -1.23) the sij values may be consider smaller than zero. Thus, only in these few cases the “per se” parent’s average is not a good indicative of the hybrid’s average performance for dried fruit total yield.

The hybrids P3xP5 and P1xP4 showed the highest positive estimates of SCA with values of 1.42 and 1.40 t ha-1, respectively (Table 2). These hybrids also showed, besides the high value of SCA, at least one parent with high value of GCA, which is desirable.

The significance of the heterosis (Table 1), by the Gardner and Eberhart (1966) method, indicates the non-additive effects are responsible for the dry matter amount increase. Therefore, the use of hybrids may contribute to increase significantly the paprika cultivars yield. Prasath and Ponnuswami (2008) found heterosis over the best parent ranged from -40.35 to 126.32 percent, being non additive effects more important for this character.

The regression coefficient β between Wr and Vr, measured by Jinks and Hayman (1953) analysis, was significantly different of 1 (Table 3), pointing the non-additive effects are, at least in part, epistatic nature.

Parent 2 (PIM 033-11) showed the highest variety effect with 1.25 t ha-1 (Table 2), becoming the most promising genitor for be used as “per se” variety, in the other hand it showed low GCA value (0.52 t ha-1) (Table 2), what decreased its value as parental line for hybrids development. Nevertheless, parent 5 (PIM 036-08), with the highest GCA (g5 = +1,98) has good potential for be used as genitor line for hybrids production.

Fruits color

Capsaten is the main pigment in paprika (KEVRESAN et al., 2009) for this reason ASTA method focus in quantify it. Parent lines PIM 033-11 (P2) and PIM 037-18 (P6) showed higher capsanten content (ASTA>125) than control line “Papri Queen” (ASTA=90.4) (Table 4). The observed heterosis for this trait can be mainly explained by specific heterosis (Table 1), indicating that heterosis is due to the difference in allele frequency amongst individual parents.

Table 4 Averages of fruit yield and color, ASTA degrees and Phytophthora capsici resistance 

Genotype Fresh fruit yield (t ha-1) Dried fruit yield (t ha-1) ASTA degrees P. capsici resistance (score) Mortality percentage caused by P. capsici
P1xP2 55.4 abc1 8.92 abcd 108.6 bcd 2.17 abcde 70.8 abcde
P1xP3 47.2 bc 8.01 abcd 78.7 de 2.29 abcd 66.6 abcde
P1xP4 63.3 ab 10.10 ab 105.1 bcde 2.00 bcdef 62.5 abcdef
P1xP5 55.4 abc 8.58 abcd 129.7 abc 1.50 ef 31.9 efg
P1xP6 55.6 abc 9.08 abcd 145.0 ab 1.92 bcdef 45.8 cdefg
P2xP3 40.7 c 6.88 cd 93.7 cde 2.52 ab 91.0 ab
P2xP4 39.4 c 6.81 cd 105.2 bcde 2.75 a 95.8 a
P2xP5 62.4 ab 9.68 abc 99.0 cde 1.63 cdef 37.5 defg
P2xP6 45.1 bc 7.44 abcd 115.5 abcd 2.24 abcd 82.1 abc
P3xP4 47.4 bc 7.84 abcd 106.6 bcde 2.63 ab 95.8 a
P3xP5 67.8 a 10.20 a 156.9 a 1.50 ef 33.3 efg
P3xP6 43.6 c 7.09 bcd 110.7 bcd 1.96 bcdef 54.1 bcdefg
P4xP5 51.8 abc 7.84 abcd 110.8 bcd 1.79 cdef 45.8 cdefg
P4xP6 48.4 bc 8.34 abcd 62.4 e 2.21 abcde 75.0 abcd
P5xP6 57.2 abc 9.38 abc 124.3 abcd 1.58 def 29.1 fg
P1=(PIM-032-03) 46.9 bc 7.64 abcd 107.6 bcde 1.75 cdef 45.8 cdefg
P2=(PIM-033-11) 51.4 abc 8.50 abcd 125.1 abcd 2.14 abcde 61.3 abcdef
P3=(PIM-034-19) 39.0 c 6.31 d 115.7 abcd 2.33 abc 75.0 abcd
P4=(PIM-035-01) 41.8 c 6.78 cd 96.1 cde 2.63 ab 87.5 ab
P5=(PIM-036-08) 46.6 bc 7.10 abcd 103.4 bcde 1.33 f 20.8 g
P6=(PIM-037-18) 43.1 c 7.34 abcd 135.3 abc 2.00 bcdef 54.1 bcdefg
Papri Queen 48.5 bc 7.57 abcd 90.4 cde 2.13 abcde 65.5 abcde

1Same letters at the columns indicate there not to be difference among genotypes by Duncan test (P<0,05);

score:1 = no symptoms, 2 = wilting and necrosis and 3 = plant death

P5 was the only parent line that showed high value of GCA (13.7 ASTA degrees) that however, cannot be considered as different from zero, due to the magnitude of the standard deviation of its estimate (Table 2).

The estimates of the SCA average components (sij) ranged from -34.5 to 30.2 ASTA degrees (64.7 ASTA degrees of amplitude) (Table 2), what equate to about 57% of the average (µ=113.9 ASTA degrees) and indicates that, besides the additive gene effects, the non-additives are also important for the trait expression. Thus, just the parent’s “per se” average is not a good indicative for the hybrid’s performance on capsanten content.

The highest positive estimates for SCA (sij) belong to the hybrids P1xP6 and P3xP5 with values of 28.9 and 30.2 ASTA degrees, respectively (Table 2). The hybrid P3xP5 has the parent P5, which has the highest value of GCA, what makes it the hybrid with the best coloration amongst the hybrids tested (Table 4). On the other hand, in the hybrid P1xP6, the parents show low or negatives GCA values, presuming the non additive effects were more important for the superiority of this hybrid combination (Table 2), however, it was lower than the hybrid P3xP5.

The significance of just the heterosis indicates the non-additive effects are the most important on the expression of the trait. These findings are different from found by Surya Kurami et al. (2014), they could not detect significance of any genetic effect for capsanten content.

The epistatic gene action contributes to the non-additive effects, once the β regression coefficient between Wr and Vr, measured by Jinks and Hayman (1953) diallel analyses, was significantly other than 1 (Table 3). This finding complement the findings of Prasath and Ponnuswami (2008), who reported importance of additive and non-additive effect for oleoresins content, which is a character related to capsanten content.

Resistance to P. capsici

According to Gardner and Eberhart (1966) analyses, heterosis was not detected to P. capsici resistance. The differences on the hybrids mortality rates may be explained basically to the differences on the variety’s effect (vi, vj).

The GCA estimate values to the trait are of little magnitude. However, when positives, indicates a tendency of the parents to originates plants with higher scores and higher mortality rates on the evaluation to P. capsici resistance. On the other hand, negative estimates indicates the parent contributes for a reduced mortality rate, what is desirable. The GCA estimate values for score ranged from -1.33 to 0.30 (1.63 amplitude) and for mortality rates the values ranged from -28.3% to 21.6% (49.9% amplitude) (Table 2). The parent P5 stands out with the highest negative value of mortality rate (-28.3%), contributing favorably to the reduction of the trait expression (Table 2).

SCA estimates (sij) for scores ranged from -0.25 to 0.17 (0.42 amplitude), representing 20% of the average (µ=2.03) (Table 2). For mortality rate, the SCA estimates ranged from -10.9% to 9.82% (20.7% amplitude), representing 36% of the average (µ=57.4%).

On Jinks and Hayman (1953) analyses, the regression coefficient between Wr and Vr, was estimated β = 0.807 for scores and β = 1.075 for mortality rate (Table 3), values that are not statistically different from 1 (α = 1%), but are statistically different from 0 (α = 1%), showing the additive-dominant model is adequate and there is no evidence of epistatic gene action. Line P5, showed the lowest scores and mortality rates (Table 4) and is located in the lower part of the regression lines (Figure 1), indicating to be the parent with more dominant alleles proportion.

Figure 1 Regression between Wr and Vr and its limitant parabola for Phytophthora capsici scores (a) and mortality percent (b). (Parental lines: 1= PIM-032-03; 2= PIM-033-11; 3= PIM-034-19; 4= PIM-035-01; 5= PIM-036-08; 6= PIM-037-18.) 

The graphical analyses (Figure 1) show the regression lines Wr in Vr intersect the y-axis in a negative value close to the source, which would indicate complete dominance gene action or mild over dominance of genes that control resistance. However, the most resistance parent (P5) shows scores and plant mortality percentage slightly lower than the hybrids where it is a parent, indicating incomplete dominance, but with average dominance degree close to 1. The alleles responsible to increase scores and mortality percentage are predominantly recessives, which concludes the correlations (Wr + Vr) and Yr, r = +0.895 (scores) and r = +0.848 (mortality percentage). The most plausible hypothesis to the interpretation of the results is that P. capsici resistance found in P5 is controlled by dominant allele(s), but with incomplete dominance.


  1. In paprika, there is heterosis, in general, to increase fresh and dried fruit and capsanten content. The non-additive gene effects were at least in part epistatic;

  2. Was not detected significant heterosis for P. capsici resistance, which is explained by incomplete dominance gene action;

  3. The parent lines PIM 032-03 (P1) and PIM 036-08 (P5) showed high GCA, what makes them potential new lines or parents for being used in breeding programs for new hybrids;

  4. The most promising hybrid for the traits evaluated was PIM 034-19 x PIM 036-08 (P3 x P5), outperforming the standard cultivar “Papri Queen” in fresh (68 t ha-1) and dry (10 t ha-1) fruit yields and ASTA degrees (157), in addition to being resistant to Phytophthora capsici;

  5. The parental line PIM 036-08 (P5), as the hybrids in which it was a parent were resistant to P. capsici.

1Parte da Tese de Doutorado do primeiro autor e do pós-doutorado do terceiro autor


The authors wish to thank the following Brazilian institutions that provided support for this research project: FAPEMIG- Fundação de Amparo à Pesquisa do Estado de Minas Gerais, CNPq-Conselho Nacional de Desenvolvimento Científico e Tecnológico, FINEP/ MCT- Ministério de Ciência e Tecnologia/ FNDCT, CAPES/MEC-Brazilian Ministry of Education, UFLA- Universidade Federal de Lavras, HortiAgro Sementes Ltda, FUNDECC- Fundação para o Desenvolvimento Científico e Tecnológico, FAEPE-Fundação de Apoio ao Ensino, Pesquisa e Extensão.


AMERICAN SPICE TRADE ASSOCIATION. Extractable color in capsicums and their oleoresins: official analytical method of the American Spice Trade Association. 4 th ed. Englewood Cliffs, New Jersey, 2004. [ Links ]

CRUZ, C. D.; REGAZZI, A. J.; CARNEIRO, P. C. S. Modelos biométricos aplicados ao melhoramento genético - Volume 1. 4 th. ed. Viçosa, MG: UFV, 2012. 514 p. [ Links ]

GARDNER, C. O.; EBERHART, A. S. Analysis and interpretation of the variety cross diallel and related populations. Biometrics, v. 22, p. 439-52, 1966. [ Links ]

JINKS, J. L.; HAYMAN, B. I. The analysis of diallel crosses. Maize Genetics Cooperation Newsletter, v. 27, p. 48-54, 1953. [ Links ]

KEVRESAN, Z. S. et al. Carotenoid content in fresh and dry pepper (Capsicum annuum L.) fruits for paprika production. Food Processing, Quality and Safety, v. 1/2, p. 21-27, 2009. [ Links ]

NASCIMENTO, I. R. et al. Capacidade combinatória de linhagens de pimentão a partir de análise dialélica multivariada. Acta Scientiarum. Agronomy, v. 32, n. 2, p. 235-240, 2010. [ Links ]

PRASATH, D.; PONNUSWAMI, V. Heterosis and combining ability for morphological, yield and quality characters in paprika type chilli hybrids. Indian Journal of Horticulture, v. 65, n. 4, p. 441-445, 2008. [ Links ]

REIFSCHNEIDER, F. J. B.; RIBEIRO, C. S. C.; CARVALHO, S. I. C. Development of new capsicum cultivars at EMBRAPA (Brazil). In: EUCARPIA MEETING ON GENETICS AND BREEDING OF CAPSICUM AND EGGPLANT, 15, 2013, Torino. Proceedings... p. 71-78, 2013. [ Links ]

SHAPTURENKO, M. N. et al. DNA divergence as a criterion for the choice of the initial material of the sweet pepper (Capsicum annum L.) in selection for heterosis. Genetika, v. 50, n. 2, p. 138-46, 2014. [ Links ]

SHRESTHA, S.; LUITEL, B.; KANG, W. Heterosis and heterobeltiosis studies in sweet pepper (Capsicum annuum L.). Horticulture, Environment, and Biotechnology, v. 52, n. 3, p. 278-283, 2011. [ Links ]

SOMOGYI, N. et al. Hybrid condiment paprika breeding and adaptation of production system in Hungary. Acta Horticulturae, v. 925, p. 37-42, 2011. [ Links ]

SOMOGYI, N. Hybrid condiment paprika breeding and protected cultivation technologies. 2010. 16 p. Booklet of the doctoral thesis (Doctoral School in Plant Production and Horticultural Sciences) - University of Pannonia Georgikon Faculty, Keszthely, Hungria, 2010. [ Links ]

SPRAGUE, G. F.; TATUM, L. A. General vs. specific combining ability in single crosses of corn. Journal of the American Society of Agronomy, v. 34, n. 10, p. 923-932, 1942. [ Links ]

SURYA KUMARI, S. et al. Genetic divergence and combining ability studies for exploitation of heterosis in paprika (Capsicum annuum L). International Journal of Agricultural Science and Research, v. 4, n. 2, p. 59-66, 2014. [ Links ]

URBEN, A. F. Phytophthora capsici Leonian, agente etiológico da murcha de Capsicum annuumL. em Minas Gerais. 1980. 89 p. Dissertação (Mestrado em Genética e Melhoramento de Plantas) - Universidade Federal de Viçosa, Viçosa, MG, 1980. [ Links ]

VENCOVSKY, R.; BARRIGA, P. Genética biométrica no fitomelhoramento. Ribeirão Preto: Sociedade Brasileira de Genética, 1992. 486 p. [ Links ]

Received: November 03, 2014; Accepted: May 07, 2018

*Author for correspondence

Creative Commons License This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.