Combining ability of white grain popcorn populations

CA Scapim et al. Crop Breeding and Applied Biotechnology 6:136-143, 2006 Brazilian Society of Plant Breeding. Printed in Brazil Combining ability of white grain popcorn populations Carlos Alberto Scapim1, Ronald José Barth Pinto1*, Antônio Teixeira do Amaral Júnior2, Freddy Mora3, and Thatiana Silva Dandolini3 Received 11 July 2005 Accepted 29 November 2005 ABSTRACT – The objectives of this study were to indicate the best improvement strategy and select parents to begin an improvement program of white grain popcorn based on the combining ability and heterosis of eight populations selected in experiments in the northwestern region of Paraná. The traits plant and ear height, grain yield and popping expansion were evaluated. The base populations, the F1 and five controls were evaluated in Maringá, state of Paraná, over the course of two years. Heterosis for popping expansion was very low and the best improvement strategy is to raise the values of popping expansion up to commercial levels through intrapopulation improvement of the populations BRS Angela and SC 002. Intense selection must be applied to reduce plant and ear height; interpopulation selection must not be initiated at this moment. Key words: Zea mays, yield, popping expansion. INTRODUCTION In 2004, according to estimates of the packaging industry, the annual demand for popcorn in Brazil was 65 to 70 thousand tons, of which 15 to 20 thousand tons were imported. Over the last five years this market has gone through several changes. The first was the introduction of a producer – distributor partnership system. The producers supply seed and warrant the sale of the product, as far as it meets quality standards (Scapim et al. 2006). From this year on, the release of hybrid IAC-112 developed by the Instituto Agronômico de Campinas (IAC), began to reduce the dependence of the country on imported seed notably, according to data obtained from packaging industries (Yoki and Hikari). Formerly, most farmers used to sow their own seed, derived from local varieties or advanced generations of American hybrids (Sawazaki 1995, Sawazaki et al. 2000). In spite of the advances, the scope of Brazilian research is restricted to the work of a few researchers of public institutions and some private seed producer companies. Nation-wide, only a few varieties and hybrid lines have been developed (Matta and Viana 2001). On this background, new improvement programs ought to be set up. One of the genetic-statistically most interesting techniques is the analysis of diallel crosses in view of the huge amount of genetic information it can provide breeders with (Cruz and Regazzi 1994). There are few reports on diallel analysis in popcorn. Zanette (1989) analyzed diallels of seven popcorn populations and inferred the existence of intermediate heterosis for popping expansion. Later, 1 Departamento de Agronomia, Universidade Estadual de Maringá (UEM), Avenida Colombo, 5790, 87.020-900, Maringá, PR, Brasil. *E-mail: [email protected] Departamento de Agronomia, Universidade Estadual do Norte Fluminense, Campos dos Goytacazes, RJ, Brasil 3 UEM 2 136 Crop Breeding and Applied Biotechnology 6: 136-143, 2006 Combining ability of white grain popcorn populations other diallel studies indicated low (Andrade et al. 2002) or even negative specific heteroses (Larish and Brewbaker 1999, Scapim et al. 2002) Scapim et al. (2002) evaluated nine yellow grain popcorn populations in the state of Paraná in diallel crosses and recommended the populations Beija-Flower and RS-20 for a reciprocal recurrent selection program. The two are complementary regarding grain yield and popping expansion, aside from the possibility of developing composites among populations of high general combining ability. Miranda et al. (2003) indicated the pair Beija-Flower and RS-20 as genetically most distant by means of the evaluation of the genetic diversity through multivariate analysis. In view of the lack of reports in Brazil on diallel analysis with the populations PR 038, PR 079, RR 046, SC 016, PR 017, RS Angela, SC 002, and PR 009 this study was conducted with the following objectives: to propose an indication of the best improvement strategy and the selection of parents for an improvement program, based on information of combining ability and heterosis for white popcorn populations. MATERIAL AND METHODS Eight popcorn populations were crossed in a diallel mating design (PR 038, PR 079, RR 046, SC 016, PR 017, BRS Angela, SC 002, and PR 009). These populations are advanced generations of American and Argentinean hybrids, with exception of BRS Angela, released by the EMBRAPA-CNPMS. The populations were chosen based on regional results of yield and popping expansion in the northwestern region of Paraná. For the establishment of the F1 plants, the populations were sown in four 10 m long rows in all possible pair combinations. In the flowering period, all possible crosses between the populations were performed manually with pollen mixture, with approximately 70 crosses; twenty-five to 30 healthy ears were harvested per intervarietal cross. The populations were multiplied on a separate area, using the same procedure with pollen mixture. The trials were developed in the growing seasons (2000/2001 and 2001/2002) on an experimental field of the Universidade Estadual de Maringá, in Iguatemi-PR, Crop Breeding and Applied Biotechnology 6: 136-143, 2006 in a randomized block design in three replicates, with 41 treatments that consisted in eight populations, the 28 intervarietal hybrids and five controls (Zélia, IAC 112, UEM J1, PR 023, and SE 013). Each plot consisted of four rows of 5.0 m, spaced 0.9 m apart. In the plots, two seeds were sown at every 0.2 m in a planting hole. After thinning out 20-25 days after emergence, one plant per spot was left over, so that in the useful area of each plot, 50 plants were left to grow. The evaluated traits were: plant height – measured from the ground level to the point of insertion of the flag leaf, of six healthy plants per plot after tasseling; ear height - measure from the ground level to the insertion of the highest ear in the stem, in the same six plants per plot; grain moisture - obtained in a sample of grains from each plot; grain yield – obtained by weighing the threshed grains in the plot and transformed into kg ha -1 , adjusted to a moisture of 15.5%; and popping expansion - obtained as the ratio of the popped popcorn volume and the raw grain mass. Two grain samples of 30g per plot, measured on a precision scale, were popped at a temperature of 280 oC for 2 min in an electrical popper with automatic temperature control developed by the EMBRAPA-CNPDIA. The volume of the popped popcorn was measured in a graduated cylinder of 1000 mL. The grain mass for popping was taken from the mid-basal part of the ears. Before evaluating the popping expansion, the grain samples and a pilot sample of 1.0kg for moisture monitoring were stored in a cold and dry chamber. The evaluated of the trait popping expansion only started from the day on when the pilot sample had attained a moisture content of about 12% (Hoseney et al. 1983). Firstly, the analysis of variance of each year was performed. The homogeneity of the residual variances of the years was tested by means of the F maximum test, at 5% probability. In the joint analysis, the effects of treatments and years were considered as fixed. The analyses of variance and the test of Scott and Knott (1974) were performed using SAEG software (Gomes and Braga 1992). Heterosis was evaluated by the methodology of Gardner and Eberhart (1966), analysis II (Vencovsky and Barriga 1992) together with the methodology of Griffing (1956); the Genes software (Cruz 1997) was used for the diallel analysis. 137 CA Scapim et al. RESULTS AND DISCUSSION Since the F test at 5% probability did not show heterogeneity in the mean squares of the residue of the individual analyses in any trait, it was possible to perform the joint analyses of variance. According to the classification proposed by Scapim et al. (1995), the coefficients of variation are considered low for plant and ear height and mean for grain yield, demonstrating good local control (Table 1). The trait popping expansion (PE) did not present classification of its own, but was classified as within the acceptable limits for agricultural experimentation (c.v=13%) compared with some other authors (Pacheco et al. 1998, Simon et al. 2004, Vendruscolo et al. 2001) and high compared with Sawazaki et al. (2000). The mean squares of years were significant at 5% probability for plant height and grain yield, indicating the differences between means of years for these traits. Only the trait grain yield did not present significance regarding the populations x years interaction, which shows that the populations did not perform differently in the two years (Table 2). The formation of four groups and five groups was observed for plant and ear height, respectively (Table 1). In the overall mean, the treatments presented high values of plant height (2.45m) and ear height (1.57m). Such tall populations can cause problems in the harvests. The population RR 046 (2.59m and 1.62m) is somewhat sensitive to lodging. Improvement programs must focus on a reduction of plant and ear heights. Several intervarietal hybrids presented a productivity of over of 4000 kg ha-1 (Table 1). These values outmatch, statistically, all controls, including the triple hybrid Zélia, the only one available on the market for those farmers that do not belong to the partnership system with the packagers. This hybrid is susceptible to corn stunt, attack of fall armyworm on the ears and ear rots (Sawazaki 2001). For popping expansion, seven groups were formed in 2001 and six in 2002. The hybrids IAC 112 (32mL g-1) and Zélia (33mL g -1) presented the best expansion means. The population BRS Angela (28mL g-1) ranked as second group in expansion. Among the intervarietal hybrids, PR 079 x BRS Angela (23mL g-1) and PR 009 x BRS Angela (26mL g -1) stood out, forming the third group in expansion. The population means of popping 138 expansion varied from 8.2mL g-1 of RR 046 to 28mL g-1 of BRS-Angela. For the intervarietal hybrids the means of popping expansion varied from 6.8mL g-1 for hybrid PR 038 x RR 046 to 26mL g-1 for hybrid BRS-Angela x PR 009. Popping expansion is a limiting factor for commercial planting of the intervarietal hybrids tested here. This seems to be a common limitation for Brazilian popcorn genotypes, with rare exceptions, such as Zélia which is a triple hybrid; IAC-112, a modified simple hybrid, and the population BRS Angela that was improved with CMS-43 (Pacheco et al. 1998). Regarding the quality of traded grain in the country, the mean popping expansion of popcorn of the trademarks Yoki and Hikari reached 39 mL g-1 and 34 mL g-1, respectively, in a hot air popper (Matta and Viana 2001). Based on these results the conclusion can be drawn that hybrids and varieties in Brazil must reach values of popping expansion over 30mL g-1. Even with low values of popping expansion, the studies of improvement indicate a promising outlook on intrapopulation improvement regarding this trait, by using different populations and improvement methods, as those described by Pacheco et al. (1998), Larish and Brewbaker (1999), Pereira and Amaral Júnior (2001), Scapim et al. (2002), Simon et al. (2004), and Scapim et al. (2006). These articles indicate the predominance of additive effects and high narrow-sense heritability for this trait. In the present study, in agreement with the literature, the relation between mean square of the general and specific combining ability was twelve-fold. This ratio was used by Vencovsky and Barriga (1992). All most productive intervarietal hybrids did not perform particularly well regarding expansion. The intervarietal hybrids BRS Angela x PR 079 (22mL g-1 and 3948kg ha -1) and BRS Angela x PR 009 (26mL g-1 and 3315kg ha-1) are the only ones with a considerable popping expansion (> 22mL g-1) and good yield (> 3000kg ha-1) requiring new evaluations to verify the stability of these two hybrids in relation to these traits. The mean squares of the general combining ability (GCA) were significant (P<0.05) for both traits (Table 2), evidencing the genetic variability among the evaluated populations and the manifestation of additive gene action, a favorable situation for improvement. The estimates of the GCA ( g$ i) population effects are shown in Table 3. A lowi, value indicates that the hybrid means Crop Breeding and Applied Biotechnology 6: 136-143, 2006 Combining ability of white grain popcorn populations Table 1. Means of the traits plant height (PH), ear height (EH), grain yield (GY) and popping expansion (PE) in popcorn parent populations of the hybrids F 1 and controls, in Maringá and Iguatemi-PR, in the growing seasons 2000/01 and 2001/02 Populations PR 038 PR 079 RR 046 SC 016 PR 017 BRS Angela SC 002 PR 009 Hybrid population means PR 038 x PR 079 PR 038 x RR 046 PR 038 x SC 016 PR 038 x PR 017 PR 038 x BRS Angela PR 038 x SC 002 PR 038 x PR 009 PR 079 x RR 046 PR 079 x SC 016 PR 079 x PR 017 PR 079 x BRS Angela PR 079 x SC 002 PR 079 x PR 009 RR 046 x SC 016 RR 046 x PR 017 RR 046 x BRS Angela RR 046 x SC 002 RR 046 x PR 009 SC 016 x PR 017 SC 016 x BRS Angela SC 016 x SC 002 SC 016 x PR 009 PR 017 x BRS Angela PR 017 x SC 002 PR 017 x PR 009 BRS Angela x SC 002 BRS Angela x PR 009 SC 002 x PR 009 Hybrid means Controls Zélia IAC 112 UEM J1 PR 023 SE 013 Control mean CV (%) PH (m) PH (m) EH (m) EH (m) GY (kg ha-1) PE (mL g-1) PE (mL g-1) 2000/01 2.15 D 2.10 D 2.68 A 2.30 C 2.35 C 2.56 B 2.40 C 2.18 D 2.34 2.60 B 2.35 C 2.36 C 2.46 B 2.23 D 2.21 D 2.60 B 2.80 A 2.60 B 2.75 B 2.68 A 2.41 C 2.73 A 2.48 B 2.68 A 2.68 A 2.60 B 2.46 B 2.66 A 2.60 B 2.25 D 2.73 A 2.56 B 2.35 C 2.80 A 2.56 B 2.78 A 2.60 B 2.55 2001/02 2.33 C 2.61 B 2.50 C 2.28 C 2.83 A 2.60 B 2.23 C 2.30 C 2.46 2.66 A 2.73 A 2.60 B 2.46 C 2.23 C 2.63 A 2.51 B 2.71 A 2.70 A 2.68 A 2.76 A 2.71 A 2.68 A 2.73 A 2.65 A 2.76 A 2.60 B 2.78 A 2.66 A 2.63 A 2.48 B 2.55 B 2.51 B 2.65 A 2.31 C 2.48 B 2.75 A 2.61 B 2.61 2000/01 1.40 D 1.38 D 1.83 A 1.53 C 1.60 B 1.38 D 1.33 D 1.58 B 1.50 1.61 B 1.65 B 1.76 A 1.86 A 1.36 D 1.30 D 1.86 A 1.68 B 1.73 B 1.80 A 1.80 A 1.35 D 1.91 A 1.63 B 1.91 A 1.76 A 1.66 B 1.85 A 1.91 A 1.60 B 1.63 B 2.01 A 1.71 B 1.56 B 1.88 A 1.65 C 1.85 A 1.80 A 1.72 2001/02 1.58 B 1.71 A 1.41 C 1.71 A 1.83 A 1.71 A 1.35 D 1.73 A 1.63 1.60 B 1.76 A 1.66 B 1.86 A 1.36 D 1.68 B 1.76 A 1.63 B 1.78 A 1.81 A 1.58 B 1.70 A 1.71 A 1.81 A 1.61 B 1.73 A 1.60 B 1.90 A 1.65 B 1.66 B 1.73 A 1.75 A 1.60 B 1.76 A 1.53 C 1.55 C 1.73 A 1.53 C 1.68 Mean 2442 D 2123 D 2361 D 3234 C 2859 C 3556 B 2138 D 1300 E 2501 3752 B 4482 A 3496 B 3306 C 3518 B 3584 B 3420 B 3881 B 4263 A 3591 B 3948 B 3343 C 4459 A 4259 A 3718 B 4362 A 2972 C 2822 C 4455 A 4424 A 3046 C 2714 C 4290 A 3299 C 3259 C 4143 A 3315 C 3211 C 3744 2000/2001 17.0 D 13.1 F 8.2 G 17.4 D 12.3 F 28.0 B 23.5 C 19.0 D 17.3 14.7 E 8.3 G 12.1 F 14.5 E 17.6 D 14.6 E 18.7 D 12.6 F 10.4 G 13.1 F 22.1 C 14.0 E 22.7 C 13.7 F 9.6 G 18.4 D 16.6 D 19.4 C 14.8 E 20.8 C 23.0 C 22.9 C 20.4 C 10.5 G 17.6 D 22.9 C 25.0 C 21.0 C 17.0 2001/2002 13.7 E 11.3 F 9.2 F 15.0 E 10.4 F 28.0 B 26.5 C 26.1 C 17.5 15.5 E 6.8 F 10.6 F 14.5 E 17.6 E 14.2 E 14.0 E 13.2 E 15.0 E 11.5 F 24.1 C 16.7 E 20.8 D 15.5 E 9.5 F 16.1 E 14.8 E 14.3 E 13.4 E 20.0 D 23.4 D 18.4 E 21.3 D 12.7 F 14.4 E 22.7 D 26.0 C 23.1 D 16.4 2.05 D 2.34 C 2.70 A 2.81 A 2.50 B 2.48 3.81 1.97 D 2.26 C 2.38 C 2.48 B 2.37 C 2.29 3.23 1.18 E 1.13 E 1.78 A 1.90 A 1.55 C 1.51 6.86 1.08 E 1.28 D 1.46 C 1.55 C 1.55 C 1.38 5.93 2881 C 2695 C 3823 B 3828 B 3017 C 3249 18.7 31.5 A 32.0 A 15.0 E 18.5 D 14.5 E 22.3 13.2 34.5 A 31.5 A 14.5 E 17.5 E 12.5 F 22.1 12.8 Means followed by the same letter in the same column belong to the same group, according to the clustering of Scott and Knott (1974) Crop Breeding and Applied Biotechnology 6: 136-143, 2006 139 CA Scapim et al. Table 2. Analysis of variance for the traits grain yield (GY) and popping expansion (PE) in eight populations and their intervarietal F1 popcorn hybrids, according to the diallel analysis proposed by Gardner and Eberhart (1966), associated to that of Griffing (1956) Sources of variation df Populations Parentals Heterosis Mean heterosis Parental heterosis Specific heterosis Years Populations x Years Parentais x Years Heterosis x Years Mean heterosis x Years Parental heterosis x Years Specific heterosis x Years GCA GCA x Years Residue 35 75 (28) 1 77 20 17 35 73 28 11 73 20 73 76 160 MS GY (kg ha-1) 3645399.7 * 826048.0 * 3100237.7 * 57703104.0 * 28612.5 1200163.2 * 5924165.3 * 414811.8 42879.5 432794.9 545787.4 80017.2 395617.5 538066.2 * 26634.2 408194.8 PE (mL g-1) 156.1155* 648.4708* 33.0267* 22.4130* 35.4687* 32.7026* 4.2000 9.8494* 8.7865 10.1152* 4.1141 23.6222* 5.6878 391.1270* 14.5161* 4.85.245 * significant. at 5% probability by the F test Table 3. Estimates of effects of general combining ability ( ) of Griffing (1956) of the traits grain yield (GY in kg ha-1) and popping expansion (PE in mL g -1 ) in popcorn, in the growing seasons of 2000/01 and 2001/02 Populations PR 038 PR 079 RR 046 SC 016 PR 017 BRS Angela SC 002 PR 009 PE 2000/01 -2.917 -1.400 -3.233 -0.050 -2.917 4.867 0.767 4.883 in which population i participates do not differ much from the general mean of the diallel. A high positive or negative value indicates that population i performs better or worse than the other populations considered in the diallel, regarding the hybrid means. The population that has higher frequencies of favorable alleles attains a higher i (Cruz and Vencovsky 1989). The GCA x years interaction was significant (P<0.05) for popping expansion, indicating the need of selecting different parents to develop hybrids in specific years. These interactions are reported in the literature by different authors (Andrade et al. 2002, Pereira and 140 PE 2001/02 -3.636 0.297 -4.126 0.198 -2.943 5.448 2.098 2.664 PE mean -3.276 -0.551 -3.680 0.074 -2.930 5.157 1.432 3.774 GY mean 144.354 170.604 47.104 3.854 75.271 427.021 -435.813 -502.396 Amaral Júnior 2001, Scapim et al. 2002). The effects of the general combining ability of the parent populations for grain yield were estimated in the mean of the two years (Table 3). Population BRS Angela presented a positivei of high magnitude for grain yield and popping expansion. For popcorn improvement programs it matters to find populations that have favorable genes for both maximum yield and expansion. This population can therefore be identified as the most promising in methods of intrapopulation improvement. When analyzing only the popping expansion, the populations SC 002 and PR 009 Crop Breeding and Applied Biotechnology 6: 136-143, 2006 Combining ability of white grain popcorn populations $ i ) for grain presented positive i and deserve attention within an Table 4. Estimates of the effects of populations ( V yield (GY) and popping expansion (PE), in popcorn, in the growing improvement program. Another favorable aspect is that seasons of 2000/01 and 2001/02 BRS Angela and SC 002 presented negativei for plant and ear height indicating that they contribute to a PE (mL g-1) GY (kg ha -1) Populations reduced height in the hybrids they take part in, which is mean mean quite satisfactory for the improvement program. PR 038 -59.25 -2.068 The mean squares of the effects of populations PR 079 -378.75 -5.218 and heterosis were significant, at 5% probability, for RR 046 -141.25 -8.718 the two traits, indicating that the populations do not SC 016 732.25 -1.218 form a homogeneous group (they differ in GCA) and PR 017 357.25 -6.068 that the hybrids express heterosis. BRS Angela 1054.75 10.581 In the partitioning of the sum of squares of SC 002 -363.25 7.581 treatments (parents and intervarietal hybrids) in the PR 009 -1201.75 5.131 analysis of variance, for the trait grain yield, the Var ( V̂i ) 119056.83 1.529 population effect represented 32% and, consequently, 78% are due to the effect of heterosis. It was therefore Var ( V̂ − V̂ ) 272130 3.496 i j concluded that the populations are contrasting for the variable grain yield. For popping expansion, the Table 5. Estimates of heteroses of populations ( $ ) and mean hi population effect represented 83% of the sum of squares ( hˆ ) for grain yield (GY) and popping expansion (PE) in popcorn for treatments and 27% are due to the effect of heterosis, GY (kg ha-1) PE (mL g-1) opposite to the result for yield. This means that the Populations PR 038 173.979 -2.241 populations are not very contrasting for popping PR 079 359.979 2.058 expansion and that jointly they are not favorable for RR 046 117.729 0.679 exploiting the heterosis effect in hybrids, in other words, ˆ 016 -292.270 0.683 (hSC i) the effect of dominance on this trait is lower than on PR 017 -103.354 0.104 grain yield. The low heterosis indicates that the BRS Ângela -100.354 -0.133 prediction of means is a good criterion for choosing SC 002 -254.187 -2.358 parents to be used in crosses. A similar result was PR 009 98.479 1.208 obtained by Scapim et al. (2002). 49607.013 0.637 Var It was possible to find with positive values for ˆ ˆ Var (hi − h j ) 113387.458 1.456 yield and popping expansion in the population BRS ˆ 1243 -0.77 (h ) Angela (Table 4). When analyzing only the popping ˆ 21867.581 0.280 expansion, as it was the case with , the populations Var (h ) SC 002 and PR 009 are auspicious for an improvement program. therefore superior to PR 079, regarding the performance It is known that the GCA effects of two populations per se, while PR 079 determines a higher heterotic value estimated by the methodology of Griffing (1956) are in the hybrids it takes part in. partly due to the contrast between the population effects For popping expansion, the heterosis mean was and partly to the contrast between the effects of varietal significant with a negative estimate of -0.774mL g-1 heterosis, estimated by the methodology of Gardner and (Table 5). This result demonstrates that for popping Eberhart (1966) (Cruz and Vencosvky 1989) and complemented by Pacheco et al. (2002). The populations expansion these populations jointly are not favorable that presented the highest GCA effects in grain yield to exploit the effect of heterosis in hybrids. This ratifies were BRS Angela and PR 079. Population BRS Angela the above-mentioned result in which 83% of the sum of (1054.75) and negative (- squares of treatments was population effect. presented a positive For specific heteroses ( ŝ ij ), the hybrids PR 079 x 100.354), while PR 079 presented a negativei (-378.75) and positive (359.979). Population BRS Angela is PR 009 (1045.30), SC 016 x PR 017 (561.39), BRS Angela Crop Breeding and Applied Biotechnology 6: 136-143, 2006 141 CA Scapim et al. x SC 002 (407.30), and RR 046 x SC 016 (393.55) in which one of the parent populations presented the highest general combining ability, were outstanding. Moreover, these are the hybrids of the first group in grain yield by the clustering criterion of Scott and Knott (1974) (Table 1). For specific heteroses, hybrid SC 016 x SC 002 (5.050) was auspicious. In this study it was not possible to identify any hybrid that had a high grain yield (over 4000kg ha-1) as well as good popping expansion (over 30mL g-1). A strategy for the establishment of popcorn populations of high yield and with good popping expansion would be the synthesis of composites for intrapopulation improvement programs. The choice of populations for the formation of these composites must be based on the V̂i and ĝ i , values, which depend on additive effects. Population BRS Angela was the only one that presented positive V̂i and ĝ i , estimates for grain yield and popping expansion, which makes the formation of composites in this phase of the program with these evaluated populations difficult. Analyzing only popping expansion, the composites formed by the populations that presented positive V̂i and ĝ i , such as BRS Angela, SC 002 and PR 009, would have a predicted value of 24 mL g-1. This estimate is inferior to the popping expansion of the three populations per se, considering the experiment of 2002. Based on this explanation, the best strategy is the intrapopulation improvement of the populations BRS Ângela and SC 002. In this context, as heterosis for popping expansion is very low, the best improvement strategy is to raise the values of popping expansion through intrapopulation improvement, until reaching commercial values and then exploit heterosis for grain yield, by means of the formation of composites or even the development of inbred lines and hybrids. CONCLUSIONS 1. Intrapopulation improvement of the populations BRS Angela and SC 002 is recommended. 2. Strong selection is required to reduce the plant and ear height of these populations. 3. No interpopulation selection should be initiated at this point. Capacidade combinatória de populações de milhopipoca de grãos brancos RESUMO – O objetivos deste trabalho foram de selecionar genitores e indicar a melhor estratégia de melhoramento para iniciar um programa de melhoramento de milho-pipoca de grãos brancos por meio das informações da capacidade combinatória e heterose de oito populações selecionadas em experimentos na região Noroeste do Paraná. As características analisadas foram altura de plantas e espigas, rendimento de grãos e capacidade de expansão. As populações, os F1 e cinco testemunhas foram avaliadas em Maringá-PR em dois anos. A heterose para capacidade de expansão foi muito baixa e a melhor estratégia de melhoramento é a elevação dos valores de capacidade de expansão, até os níveis comerciais, por meio do melhoramento intrapopulacional das populações BRS Angela e SC 002. Forte seleção deve ser feita para reduzir altura de plantas e de espigas. A seleção interpopulacional não deve ser iniciada nesse momento. Palavras-chave: Zea mays, rendimento, capacidade de expansão. REFERENCES Andrade RA, Cruz CD, Scapim CA, Silvério L, Pinto RJB and Tonet A (2002) Análise dialélica da capacidade combinatória de variedades de milho-pipoca. Acta Scientiarum 24: 11971204. Cruz CD (1997) Programa Genes: Aplicativo computacional em genética e estatística. Editora UFV, Viçosa, 443p. Cruz CD and Regazzi AJ (1994) Modelos biométricos aplicados ao melhoramento genético. Editora UFV, Viçosa, 390p. 142 Cruz CD and Vencovsky R (1989) Comparação de alguns métodos de análise dialélica. Revista Brasileira de Genética 12: 425-438. Gardner CD and Eberhart SA (1966) Analysis and interpretation of variety cross diallel and retaled populations. Biometrics 22: 439-452. Gomes JM and Braga JMF (1992) SAEG. Editora UFV, Viçosa, 100p. Griffing B (1956) Concept of general and specific combining Crop Breeding and Applied Biotechnology 6: 136-143, 2006 Combining ability of white grain popcorn populations ability in relation to diallel crossing systems. Australian Journal of Biological Science 9: 463-493. Hoseney RC, Zeleznak K and Abdelrahman A (1983) Mechanism of popcorn popping. Journal Cereal Chemistry 1: 4352. Larish LLB and Brewbaker JL (1999) Diallel analyses of temperate and tropical popcorns. Maydica 44: 279-284. Matta FP and Viana JMS (2001) Testes de capacidade de expansão em programas de melhoramento de milho-pipoca. Scientia Agricola 58: 845-851. Miranda, GV, Coimbra RR, Godoy CL, Souza LV, Guimarães, LJM and Melo, AV (2003) Potencial de melhoramento e divergência genética de cultivares de milho-pipoca. Pesquisa Agropecuária Brasileira 38: 681-688. Pacheco CAP, Gama EEGE, Guimarães PEO, Santos MXD and Ferreira AS (1998) Estimativas de parâmetros genéticos nas populações CMS-42 e CMS-43 de milho-pipoca. Pesquisa Agropecuária Brasileira 33: 1995-2001. Scapim CA, Pacheco CAP, Tonet A, Braccini AL and Pinto RJB (2002) Análise dialélica e heterose de populações de milhopipoca. Bragantia 61: 219-230. Sawazaki E (1995) Melhoramento do milho-pipoca. Editora Instituto Agronômico de Campinas, Campinas, 53: 21p. Sawazaki E, Paterniani MEAGZ, Castro JL, Gallo PB, Galvão JCC and Saes LA (2000) Potencial de linhagens de populações locais de milho-pipoca par síntese de híbridos. Bragantia 59:143-151. Sawazaki EA (2001) Cultura do milho-pipoca no Brasil. O Agronômico 53: 11-13. Simon GA, Scapim CA, Pacheco CAP, Pinto RJB, Braccini AL and Tonet A (2004) Depressão por endogamia em populações de milho-pipoca. Bragantia 63: 55-62. Scott AJ and Knott MA (1974) Cluster analysis methods for grouping means in the analysis of variance. Biometrics 30: 507-512. Pacheco CAP, Santos MX, Cruz CD, Guimaraes PEO, Parentoni SN and Gama EEG (2002) Topics on the genetics of maize yield stability. Crop Breeding and Applied Biotechnology 2: 345-354. Vendruscolo EC, Scapim CA, Pacheco CAP, Oliveira VR, Braccini AL, and Gonçalves-Vidigal MC (2001) Adaptabilidade e estabilidade de cultivares de milho-pipoca na região centro sul do Brasil. Pesquisa Agropecuária Brasileira 36:123-130. Pereira MG and Amaral Júnior AT (2001) Estimation of genetic components in popcorn based on the nested design. Crop breeding and Applied biotechnology 1: 3-10. Vencovsky R and Barriga P (1992) Genética biométrica no fitomelhoramento. Revista Brasileira de Genética, Ribeirão Preto, 486p. Scapim CA, Braccini AL, Amaral Júnior AT, Pinto RJB, Silva RM, Moterle LM and Rodovalho MA (2006) Componentes genéticos de médias e depressão por endogamia em populações de milho-pipoca. Ciência Rural 36: 36-41. Zanette VA (1989) Análise da variabilidade genética em variedades de milho-pipoca (Zea mays L.) I. Heterose da capacidade de expansão do grão. Agropecuária Sulriograndense 25: 173-181. Scapim CA, Carvalho CGP and Cruz CD (1995) Uma proposta de classificação dos coeficientes de variação para a cultura do milho. Pesquisa Agropecuária Brasileira 30: 683-686. Crop Breeding and Applied Biotechnology 6: 136-143, 2006 View publication stats 143