GENETIC CONTROL OF WHEAT SEEDLING ROOT GROWTH

Wheat cultivars should have long primary roots to allow good crop establishment, considering the short crop establishment season (April) in the State of São Paulo, Brazil, where the occurrence of water stress is frequent. This paper demonstrates the control and type of inheritance of the primary root growth trait. Crosses were made between genotypes, BH-1146 and KAUZ “S”/IAC-24 M4 with strong and reduced primary root growth, respectively. F2 and F3 generation seeds from these crosses and F2 generation seeds from the backcrosses of both parents were also obtained. Seedlings from these genotypes plus the parentals were evaluated in relation to primary root growth in complete nutrient solutions containing 3.875 mg L phosphorus, at pH 4.0 and a temperature of 25 ± 1°C for 10 days. Control of the primary root growth trait was demonstrated to have quantitative inheritance. The degrees of dominance showed that the genes for strong root growth had a partially recessive behavior. Heterosis and heterobeltiosis values were negative. The estimated broad-sense heritability for root growth indicated that a great part of the observed variation was of genetic origin. The narrow-sense heritability indicated that a great part of the total genetic variability in relation to the trait under consideration is due to a small number of genes. Considering the estimated coefficient of determination, selection for strong root growth would be effective even when made in the early segregant generations after the cross.


INTRODUCTION
Wheat in the State of São Paulo, Brazil, is grown in acid soils under dryland conditions, in the period from April to August, in succession to soybean and corn.The occurrence of toxic levels of aluminum, especially in acid soils, has been reported by many researchers as very common in the Brazilian Cerrado region.This makes plants more sensitive to drought by preventing them from obtaining water from deeper soil layers (Foy et al., 1965;Campbell & Lafever, 1976;Camargo, 1985).
Mutant lines obtained by gamma irradiation of the cultivar Anahuac (sensitive to aluminum toxicity) present high aluminum toxicity tolerance, which is due to a single pair of dominant alleles.These alleles express the same tolerance as cultivars BH-1146 and IAC-24 (Camargo et al., 1997;2000).
In addition to the selection of lines that exhibit Al 3+ tolerance (Camargo, 1993), the selection of lines of plants that have long primary roots in the initial stages of development is of great importance, considering the short crop establishment period (April) when the occurrence of water stress is frequent, to allow good development of the crop (Camargo & Ferreira Filho, 2000).
The objective of this study was to obtain information on the type of inheritance involved in the expression of stronger root growth in wheat cultivar BH-1146 at the initial stages of development (7 and 15 days).

MATERIAL AND METHODS
Crosses were made between wheat genotypes BH-1146 (P6) and KAUZ "S"/IAC-24 M 4 (P25), tolerant to aluminum toxicity, which have strong and reduced primary root growth, respectively (Camargo et al., 2002;2004).Seedlings from the parentals, F 2 and F 3 generation seedlings from the cross between the parentals from individual F 1 and F 2 generation plants, in addition to F 2 generation seedlings from backcrosses for both parentals also from individual F 1 generation plants were submitted to a ten-day growth period in nutrient solutions using the following technique: Seeds from the populations were washed with a 10% sodium hypochlorite solution and placed to germinate in a refrigerator at a temperature of 12 o C for 72 hours.After this period, roots were beginning to emerge.
Fourteen nylon screens were used, adapted over 14 plastic containers (8.3 L capacity each), containing complete nutrient solutions with 3.875 mg L -1 phosphorus, at pH 4.0 and a temperature of 25 ± 1 o C. Four populations and both parentals were evaluated on each screen.The evaluated populations are listed in Table 1.About 100 seeds from each population and 25 seeds from each parental were selected and placed on each screen with tweezers.Seeds were maintained moist with the emerging roots touching the solutions, therefore obtaining a ready supply of water and nutrients.
Seedlings were allowed to develop under these conditions for ten days.After this they were removed from each nylon screen for evaluation.The root growth for each plantlet was obtained by measuring central primary root length.
During the entire experiment, the pH in the nutrient solutions was maintained as near as possible to 4.0 by daily adjustments.Means, variances, standard errors, and coefficients of variation were calculated for all populations.
The degrees of dominance in relation to root growth during ten days in complete nutrient solutions were calculated for each cross in the F 2 generation, according to the method proposed by Strickberger (1968).Thus, a degree of dominance equal to +1 would mean the complete dominance of genes that would condition for greater root growth during ten days in complete nutritive solutions, and -1 would mean the complete dominance of genes that would condition for smaller root growth.
Heterosis as well as the percentage of increase of F 2 over the parental mean, were calculated by means of the formula described by Matzinger et al. (1962).The superiority of F 2 over the parental mean with the highest value for the evaluated character was defined as heterobeltiosis, and was estimated using the formula proposed by Fonseca & Patterson (1968).
Broad-sense heritability (proportion of genetic variance and phenotypic variance) was calculated following the method cited by Briggs & Knowles (1977), and narrow-sense heritability (proportion of additive genetic variance and phenotypic variance) by regression of the mean for the F 2 's over the corresponding parental means; the coefficient of determination was calculated by the correlation between the mean for the F 2 's and the corresponding parental means, according to Falconer (1960).
The number of genes (N) contributing to a quantitative character was estimated by the formula proposed by Poehlman & Sleper (1995): where: P1 and P2 = P 1 and P 2 parental means; σ F1 and σ F2 = standard errors for the F 1 and F 2 generations.In this work, the standard error for the F 1 generation was replaced by the mean for the parental standard errors.

RESULTS AND DISCUSSION
The number of plants (n), mean, variance, standard error, and coefficient of variation (C.V.) for the populations in relation to primary root growth during ten days in complete nutrient solution are presented in Table 1.
P6 presented a mean root growth ranging between 180.6 and 253.2 mm, while P25 under the same conditions had a variation between 112.1 and 159.8 mm (Table 1).Parentals were markedly different in relation to root growth under similar cultivation conditions (nutrient solutions).
Although parentals were self-fertilized for many generations (thus having neared complete homozygosis), differences were observed within each of the parental groups.This variability probably resulted from experimental errors.
Taking into account the distribution frequencies (data not presented), standard errors and coefficients of variation (Table 1), the F 2 generation for crosses P6/P25 or P25/P6 was more variable than the parentals.Despite the fact that the F 2 generation is also affected by the environment, the segregation and gene recombination effects were responsible for a large part of the phenotypic variability that existed in this generation.
Individuals from various parts of the distribution curve of an F 2 population produced F 3 progenies that were markedly different in mean size (varying from 112.5 to 188.7 mm).Different F 2 individuals produced F 3 progenies with different variances (variabilities) (Table 1).Twentynine F 3 progenies had smaller variance than the F 2 progeny (729.2) which presented the highest value.Only progeny (P25/P6)-5 F 3 had a higher observed variance (896.4).This was expected, because an increase in the frequency of homozygous individuals and a consequent reduction in the frequency of heterozygous individuals occur at every self-fertilization generation.
Individuals from various parts of the distribution curve of an F 1 population from the (RC 1 ) P6/P25//P6 or P25/P6//P6 backcrosses produced F 2 progenies that were markedly different in mean size (varying from 164.9 to 221.1 mm) (Table 1).Taking into consideration the individuals from various parts of the distribution curve for an F 1 population of the (RC 2 ) P6/P25//P25 or P25/P6//P25 backcrosses, these produced F 2 progenies that were markedly different in mean size (varying from 124.3 to 167.4 mm) (Table 1).The stronger primary root growth trait exhibited by genotype 6 or reduced root growth exhibited by genotype 25 is controlled genetically and is explained by an inheritance of quantitative characters (Allard, 1960).These results indicate that plant selections performed in the early segregating generations for strong or reduced primary root growth would have a great opportunity of being successful.
The degrees of dominance for root growth are presented in Table 2. Genes for stronger root growth found in genotype 6 had a partially recessive behavior in the F 2 generation of the cross of this genotype with genotype 25, with smaller root growth, i.e., the mean for primary root growth in the F 2 populations was smaller than the parental mean, but higher than the mean for the parental that had reduced primary root growth (Table 1).
Negative heterosis and heterobeltiosis were observed in the F 2 generations, demonstrating that the hybrid means were lower than the mean root growth for the parents herein used, and lower than the mean root growth for the parent that exhibited longer roots, respectively (Table 2).The negative heterosis values of the F 2 generation indicated that root growth for this cross was lower than the mean root growth for the parents used, thus confirming the results obtained for degrees of dominance, i.e. partial dominance of the genes that condition for plants with smaller root growth.
The estimated broad-sense heritability for root growth ranged from 50.4 to 72.6%, indicating that a large part of the variation observed in the populations comes from a genetic source (Table 2).The estimated narrowsense heritability for root growth was 75%, showing that a large part of the total genetic variability for this trait is due to a small number of genes, therefore having little environmental influence in its expression.Since a coefficient of determination of 0.73, significant at 5% was also obtained for primary root growth, the data suggest that selection for this trait would be effective if made in early segregating generations, thus confirming the results previously discussed.
The number of genes contributing toward the quantitative primary root growth character was estimated by the formula proposed by Poehlman & Sleper (1995), and varied from 2 to 5, according to the F 2 population under consideration.These results confirmed the data previously discussed, related to the estimates for degrees of dominance and broad-and narrow-sense heritabilities, which suggested that a small number of genes acted on the root growth character.In crosses aimed at breeding, attempts to estimate the number of genes that contribute to a given quantitative character are impracticable (Poehlman & Sleper, 1995).Frequently, the best a breeder can do is to estimate whether the quantitative character is governed by a relatively large or a relatively small number of genes.
The practical significance of the results is the potential for making successful selections in the initial segregating generations after the cross, for plants that show long primary roots in the early stages of development, for a good establishment of the crop, especially when water deficits occur after the emergence of the seedlings in the field.

CONCLUSIONS
The primary root growth trait evaluated in nutrient solutions is controlled genetically and is explained by an inheritance of quantitative characters.The genes for stronger root growth found in genotype 6 have a partially recessive behavior when this genotype is crossed with genotype 25, with smaller root growth.The heterosis and heterobeltiosis values were negative for the primary root growth trait.The observed variation in the populations is, in large part, of additive genetic source and attributed to a small number of genes.Selection for primary root growth can be effectively performed since the initial segregating generations.), and estimates for number of genes involved in primary root growth, during ten days, in plastic container 8, 9, and 10 containing complete nutrient solution, derived from data obtained in parental and F 2 generations of crosses involving wheat genotypes (P6=BH-1146 and P25=Kauz"S"/ IAC-24 M 4 ). r

Table 1 -
Number of plants (n), mean, variance, standard error, and coefficient of variation (C.V.) for primary root growth during ten days, in 14 plastic containers containing complete nutrient solution, of populations from crosses involving