Correlation between seed size , protein and oil contents , and fatty acid composition in soybean genotypes

Se analizaron 18 genotipos de soja {Glycine max (L.) Merrill) con grupos de madurez IV, V, VI o Vil, cultivados en 1995/96 en la Estación Experimental Agropecuaria (EEA-INTA) de Manfredi y Marcos Juárez, Argentina. El propósito de la investigación fue determinar posibles asociaciones entre el tamaño del grano, los contenidos de proteínas y aceite y la composición de ácidos grasos del mismo. El tamaño del grano varió entre 13.9-21.0 g/100 granos. Los porcentajes de proteínas y aceite estuvieron comprendidos entre 331-448 y entre 198-267 g kg"\ respectivamente, y no mostraron correlaciones significativas con el tamaño del grano. Se observaron correlaciones significativas entre el tamaño del grano y determinados ácidos grasos: positivas con esteárico y oleico y negativa con linoleico. Las asociaciones encontradas podrían ser de utilidad en programas de mejoramiento de soja.


INTRODUCTION
The soybean is the most important source of edible vegetable oil and high quality vegetable

Plant material.
Eighteen soybean genotypes {Glycine max (y.) Merrill) with maturity groups IV, V, VI or VII, were grown in 1995/96.In order to examine the feasibility of environmental variation in chemical composition, the genotypes were grown at the Estación Experimental Agropecuaria (EEA -INTA) from two different localities: Manfredi and Marcos Juarez, Córdoba province, Argentina.Manfredi (latitude 31° 49' 8) is 292 meters above mean sea level and had an average monthly temperature of 20.9° C during the growing season (november-abril).Marcos Juárez (32° 35 S) is 110 meters above mean sea level and had an average monthly temperature of 22.8° C. Summer rainfall is normally about the same for the two locations; Manfredi received 577.5 mm and Marcos Juárez had a total of 629.0 mm during the growing season.The experiment was replicated three times in a split-plot arrangement of a randomized complete block design.Seeds of each lot were harvested separately by hand at maturity when seed moisture was reduced to 10 % or less.

Oil extraction.
Crude oils were extracted with n-hexane in a Soxhiet apparatus for 12 hs.After drying the solution with anhydrous sodiunn sulphate, solvent was removed by vacuum distillation at 30°C.Oil percentages were determined by weight difference.

Fatty acid analysis.
conducted in triplicate.Statistical differences among cultivars were estimated from ANOVA test at the 5 % level (P=0.05) of significance for all parameters evaluated.Whenever ANOVA indicated significant difference, a pairwise comparison of mean by Least Significant Difference test (LSD) was carried out.Correlation analysis was performed employing Pearson's test.Correlation coefficients were calculated on cultivar values for each location.
The lipids were subjected to alkaline saponification (1 N KOH in methanol).Unsaponifiable matter was extracted with n-hexane.The fatty acid methyl esters (FAME) of total lipids were obtained using 1 N H2SO4 in methanol and analyzed by gas chromatography (GC) according to Maestri and Guzman (1993).The theoretical iodine number was calculated according to Carreras ei a/.(1989).

Protein content.
The nitrogen content was determined by the Kjeldahl method and it was converted to protein content by using the conversion factor 6.25 (AOAC, 1980).

Statistical analyses.
All chemical and physical determinations were

RESULTS AND DISCUSSION
Seed size presented a statistically significant variation among cultivars, ranging from 13.9 to 20.7 g/100 seeds; cultivars of maturity group IV showed the highest values (Tables I and II).
Statistically significant differences were found among genotypes for protein and oil contents.Protein content ranged from 331 to 448 g kg'\ whereas oil content constituted more than 198 g kg"^ in all cultivars.Protein and oil contents varied significantly between locations; cultivars grown at Manfredi showed the highest values (Tables I and II).There was no a negative correlation between protein and oil contents as indicated previously (Leffel and Rhodes, 1993;Liu etal., 1995).

Table I
Seed size (g/100 seeds), protein and oil contents (g kg"^), fatty acid composition (% of total fatty acids), percentage of unsaturated (% US), oleic to linolenic (0/Ln) ratios and iodine values (IVs) of seed oils from 18 soybean cultivars grown at Manfredi.Means (n=3) from each cultivar followed by the same letter within each column are not significantly different at P < 0.05 by LSD pairwise comparison of means  *tr, trace < 0.1%.
The fatty acid compositions were characterized by a high proportion of unsaturated fatty acids which varied between 83.0 and 85.2 %.The relative percentage of major fatty acids ranged from 9.9 to 12.7 for palmitic acid (16:0), 3.2 to 5.1 for stearic acid (18:0), 16.9 to 27.4 for oleic acid (18:1), 50.2 to 58.9 for linoleic acid (18:2) and 6.8 to 10.0 for linolenic acid (18:3).Genotypes of maturity group IV from two localities studied, presented the highest percentages of oleic acid and the lowest percentages of linoleic and linolenic acids.As a consequence, the highest 0/Ln ratios and the smallest IVs were found for genotypes of maturity group IV.
Seed size showed no significant correlation with both protein and oil contents.Similarly, there was no significant correlation between total percentage of unsaturated fatty acids and seed size.However, when individual fatty acids were considered, it was found a significant correlation (P<0.05) with seed size: positive with stearic and oleic acids, and negative with linoleic acid.In addition, there was a negative correlation between seed size and linolenic acid percentage but it was not significant at P<0.05 level (Tables III and IV).These findings are in general agreement with those previously published by Liu et a/., (1995).Regarding the relationships among the major individual unsaturated fatty acids, there was a negative correlation between oleic acid with each of two polyunsaturated acids (18:2 and 18:3), while a positive correlation between linoleic and linolenic acids was observed at P<0.05 level.These results agree with those obtained by Burton et al. (1983) and Liu et al. (1995).Also, the following pairs of relationships were found: positive between 18:0-18:1, and negative between 18:0-18:2 and 18:0-18:3.
Current biochemical evidence indicates that in soybeans and in several other plant species the polyunsaturated fatty acids (18:2 and 18:3) are produced by the consecutive desaturation of 18:1 (Wilson et al., 1980).Liu et al. (1995) have suggested that genes encoding enzymes responsible for desaturation of oleoyl-phosphatidylcholine (PC) to linoleoyl-PC and further to linolenoyl-PC are related not only to each other but also to genes controlling seed size.
In summary, the results obtained in this work suggest that, besides oil and protein contents and fatty acid composition, seed size and its relationship with individual fatty acids and the association among these fatty acids, must be considered in soybean breeding programs.