Popcorn belongs to the botanical species Zea mays ssp. mays L.. There are several hypotheses described for its origin and evolution. Erwin (1949) suggests that popcorn recently originated as a mutation within the flint maize type, but this proposal was critiziced in view of archeologic evidence. Another author quoted by Mangelsdorsf (1974) suggests that popcorn originated from hybridization between the Euchlaena and Zea genera, due to the fact that crossbreeding the two results in an ear with sharp, small and hard kernels. Additionally, it was observed that "teosinte" may explode like "popcorn".

Popcorn was classified into two primary types of kernels: pearled and riced. Pearled kernels have a slight crown and riced are sharp. Both types are hard and small, and the endosperm contains a reduced proportion of farinaceous starch, because most of the starch-producing cells only produce corneous starch and a little farinaceous starch around the embryo.

Expansion capability, defined as expanded volume of 1 g. of grain, is the most important trait for selection in popcorn, because an enhanced expansion is associated with increased palatability. Another important factor is grain yield and type of grain (there are little white, little yellow and big yellow kernels; each type may be preferred more than the others). For quantifying grain size, the number of grains in 10 g. must be considered (large type has 52-67 grains, medium 58-75 and little 76-105). Paired maturity of crop also must be considered, because this has a big influence on expansion capability.

The problem situation focused on by plant breeding scientists is in breaking the negative correlation that exists between traits of cultivation interest and expansion volume, for getting a product which can satisfy producers' and consumers' demands simultaneously. Farmers desire one genotype with large yield, stability, disease resistance and increased expansion capability. Consumers demand tender, soft, tasty popcorn flakes, with an attractive color and without pericarp.

The aim of this work was to analyze this quest, in order to identify cultivars with the highest grain yield and increased expansion capability, resulting in recomendations to "Cuenca del Río Salado" (Buenos Aires Province, Argentina) farmers.

Trials were carried out in a region of Buenos Aires Province known as "Cuenca del Río Salado", where dairy farming is the major activity. It is located at 34 degrees, 38 minutes of South latitude and 58 degrees, 48 minutes of West longitude, and topographically is 23 m. above sea level. This region has heterogeneous soils and for trials parcels classified as Molisols were used, with a molic epipedon of 27 cm. of depth, 4.5% of organic matter and use of the soil capability types I and II.

Fourteen cultivars were sown (simple hybrids), named C1 to C14, in 5 locations, during 2 years (1998 and 1999). A statistical design was used of randomized complete blocks with a factorial arrangement: 14 treatments x 3 replications x 10 environments, according to the following model:

Y_ijk = m + a_i + b_j + (a x b)_ij + r_k(j) + e_ijk

Where :

Y_ijk = Observation corresponding to i treatment in j environment of k replication

m = General mean of evaluated trait

a_i = Fixed effect of i genotype

b_j = Random effect of j environment

(a x b) _ij = G x E interaction effect

r_k(j) = Replications nested within environmental effect

e_ijk = Random variable corresponding to experimental error

The experimental unit consisted of 2 furrows of 5 cm of longitude with a distance between rows of 0.70 m.

Evaluated variables were:

YIELD: Grain yield (kg) / experimental unit

EXVOL: Expansion volume (cc/g)

For selection of cultivars the method of Kang was applied (Agron. J. 85:754-757, 1993), who defines a stability-yield statistic Ysi that considers type II error for both components, yield and stability. The stability component of Ysi is substantiated in stability variance by Shukla s_i² (Heredity 29:237-245, 1972) and variation statistics by Lin and Binns (Can J. Plant. Sci. 68:193-198, 1988), that is a relative dependent measurement of genotypes included in the work and shows the contribution of a genotype to G x E interaction, which is assumed to all genotypes included in this experiment.

Results of the combined analysis are presented in Table 1, where significant effects for environments and genotypes are observed for the EXVOL variable, without detecting significant effects in G x E interaction; by contrast, for the YIELD variable, this interaction was significant.

Table 2 shows mean values of genotypes in all environments, Shukla´s variance of stability and selection index. The selection index for the EXVOL variable, will be based only in mean values ranking, because G x E interaction and therefore stability variances, are not significant.

Genotype C5 recorded the highest index (Y_si = 16), followed by C2 (Y_si = 14), C9 (Y_si = 13), C3 (Y_si = 12), and C11 (Y_si = 11).For the YIELD variable, genotypes which made the major contribution to G x E interaction sum of squares were observed. They are the most non-stable genotypes: C1 (s²_i = 0,44), C6 (s²_i = 0.41), C5 (s²_i = 0.29), then, with minor values C9 (s²_i = 0.23) and C10 (s²_i = 0.22). The highest selection index resulted from C7 (Y_si = 15), followed by C3 (Y_si = 14), C10 (Y_si = 12), C2 (Y_si = 11) and C13 (Y_si = 9).

This work allows us to establish which are the genotypes with the most promise, considering the ecological conditions of the region studied. If yield, as weight variable, is preferred, the best genotype is C7; if stability is preferred, C3· and C13 are the best genotypes; for both factors, C3 is the best genotype. For selecting by yield and expansion volume simultaneously, the highest index corresponding to both variables must be combined. In this case, C2 and C3 were selected. The C5 genotype showed high expansion capability, low yield and low adaptability. This phenomenon may be explained by the negative correlation that exists between yield vs. expansion volume, according to reports in most papers on popcorn.

Table 1. Mean squares from the combined anova for EXVOL and YIELD.
 

	Environment (E)	Genotype (G)	Interaction (G * E)	Heterogeneity	Residual	Pooled error
DF#	93	13	117	13	104	260
EXVOL	479.2**	25.3**	8.19 ns	8.48 ns	8.15 ns	7.92
YIELD	93.4**	0.82**	0.19**	0.21 ns	0.18**	0.098

#Degrees of freedom.
**p< 0.01; *p<0.05; ns: non-significant.

Table 2. Stability statistics corresponding to the two variables studied.
 

Gen	EXVOL			YIELD			Gen	EXVOL			YIELD
	mi	s2i	Ysi	mi	s2i	Ysi		mi	s2i	Ysi	mi	s2i	Ysi
C1	25.33	3.38	3	2.67	0.44**	5	C8	25.44	6.56	4	2.44	0.13	4
C2	26.77	8.87	14+	2.65	0.11	11+	C9	26.73	10.45	13+	2.57	0.23*	6
C3	26.57	7.23	12+	2.69	0.065	14+	C10	25.48	3.56	5	2.78	0.22*	12+
C4	24.04	5.36	-1	2.44	0.12	3	C11	26.07	8.16	11+	2.46	0.11	5
C5	27.07	11.47	16+	2.25	0.29**	-10	C12	25.62	14.38	4	2.26	0.14	-1
C6	25.10	10.13	2	2.49	0.41**	-2	C13	24.11	4.86	0	2.55	0.063	9+
C7	25.88	13.19	7	2.74	0.12	15+	C14	25.92	7.05	10	2.40	0.19*	-2

** p< 0.01; * p<0.05; Gen= genotype; m_i = general mean of each genotype; s²_i = stability variance by Shukla; Y_si = selection index by Kang; + = selected genotypes.
 
 

Return to the MNL 77 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage