Characterization of carbohydrate complex of cereals and its nutritional implications.

Abstract

45 CHARACTERIZATION OF CARBOHYDRATE COMPLEX OF CEREALS AND ITS NUTRITIONAL IMPLICATIONS PRAN VOHRA AND GHIAS SHARIFF Summary Two varieties of triticales, 3 of soft wheats, 2 of hard wheats and 3 of durum wheats were anlayzed for moisture, crude protein, ash, lipid fraction, acid and neutral detergent fibers, .available and non-available carbohydrates, hemicellulose, lignin, reducing sugar and starch contents. Amylose /amylopectin ratios in starch were determined. The nutritional value.was determined using Tribolium castaneum larvae and Japanese quail. The growth of test organisms was ooorly correlated to the protein content of cereals. It has stronglylcorrelated to the available and nonII available carbohydrate contents. . I. INTRODUCTION The plant breeders manipulate the genetic make up of cereals to develop new varieties for increased yields and pest resistence. The ultimate use of these cereals in the diets ofhumans and animals is'of direct concern to the nutritionist. He evaluates the nutrient content of these cereals in terms of their protein content and amino acid profiles, nature of carbohydrates, lipids, vitamins and minerals, their availabilities, and the presence of any deleterious antinutrients. The 'evaluation of availability of nutrients has to be done by feeding trials : with mice, rats,, or poultry and the sample size has to be large enough G3 to conduct these experiments. It is desirable to develop a bio-assay for screening a large number of samples which may be available in less than 10 q quantities. .. I ,Insects, specially which infest cereals, deserve further study as test organisms. The rationale is that any cereal which does not support the optimal growth of pest insect larvae is either deficient in a nutrient or contains some deleterious antinutrient. The larvae of red flour beetle (Tribolium castaneum) have proved quite useful as test organi,sms in evaluation of nutritive value of some c,arbohydrates, starches. and cereals (Vohra et al. 1978, 1979 a,b). We are interested in the nutritional quality of cereals, particularly their ca.rbohydrate complex. The results of.a study on the chemical characterization of this complex and its. effect on the growth of Tribolium larvae as well as quail are presented in this report. . Department of Avian Sciences, Unjversity of Caljfornia, Davis, CA 9%X6 46 II . MATERIALS -AND METHODS Milled rice samples with different amylose contents were obtained from Western Regional Laboratory, Berkeley, California. The wheats and triticales were grown in the University of California Agronomy Research Farm in Davis. Moisture, ash, lipids, crude protein (CP), acid detergent fiber (ADF), and neutral detergent fiber (NDF) were determined by the methods given in AOAC (1975) and as described by Robertson (1978). A factor of 5.8 was used to convert .Kjeldahl nitrogen. to crude protein for wheats and triticales. Starch was isolated from test samples as described by Wolf (1964). Amylose in starch was measured according to the procedure of Williams et al. (1958). Amylopectin was estimated by subtracting the value of . amylose from that of total starch. The methods described by Southgate (1969 a,b) were used for the determination of unavailable carbohydrate and available carbohydrates. An'other enzymatic method suggested by Hellendoorn et al. (1975) was also used for determination of unavailable carbohydrates. Unavailable carbohydrates were further fractionated into water soluble pentosans, cellulose, pcetic substances, hemicellulose and lignin asoutlined by Southgate (1969 b). For bioassays, larval weights of Tribolium castaneum fed the test cereal diets (Table 1) from 6 days to 14 days of age were determined. The diets were ground finely 'to pass through a loo-mesh sieve. Diets containing 90% finely ground test cereal and 10% brewers yeast were also ' used in another study. Essentially, the procedure was the same as described previously by Vohra et al. (1979 a). Day-old Japanese quail (Coturnix coturnix japonica) were fed the s test cereal diets of the composition given .in Table 1. A combination of a test cereal and DL-glutamic acid with.an estimated protein equivalence of59.5%obtained by multiplying its nitrogen content by 6.25 was used to obtain a total of 15% crude pro'tein. The gain in body weight of Japanese quail was determined. over an experimental period of 2 weeks. 47 III . RESULTS AND DISCUSSION The results of a preliminary screening of a large number of wheat and triticale varieties (Table 2) suggested that Tribolium larvae developed significantly better on triticales and soft wheats than on durum Hard wheats were intermediate between the soft and wheats (P40.01). durum wheats but did not differ significantly from either of them. The .correlation between the crude protein content of the test cereals and the larval growth was poor, especially for the soft and hard wheats. There may be some constituentbesides the protein content, which may have,a better correlation with the larval growth. The preliminary data on the bioassay of some rice samples with .Tribolium larvae (Table 3) suggest a significant correlation between the, crude protein (Xl) or amylose (X2) contents and the larval growth (Y) with r values of -0.86 and-0.79, respectively. It implies that the carbohydrate complex is of equal importance in the nutritional evaluation of cereals. The 'proximate analyses of some wheats and'triticale s,amples which were further characterized for their carbohydrate complex are given in Table 4. Two of the soft wheat (Atlas 66 and NapH al/Atlas 66) were not acclimatized to Davis climate and the grains were shrivelled. This explains their significantly high protein content and low. total carbohydrate concentration. No significant differences in lipid or ash contents of. these samples were observed, 4 8 No significant differences in the amylose content of the starches from'any of'the test cereals were detected (Table 5). The ratio between amylopectin as determined by difference method and amylose was roughly between 3.2`to 4:5. As no significant differences were observed `in amylose and amylopectin contents, both of these constituents of starch were lumped together with reducing and non-reducing sugars in the 49 total avail,able carbohydrate fraction of cereals. . Enzymatic method of Hellendoorn et al. (1975) for unavailable carbohydrate, which he calls dietary fiber, also estimated ash content of test cereals (Table 6). This explains why the values for unavailable carbohydrate were higher by this method than of Southgate (1969b) in. which ash was characterized separately. ADF is always less than NDF as determined by the methods of Van Soest (1966) because hemicellulose is not measured in ADF; It measures cellulose and 1igni.n only. ADF and NDF were lower than unavailable carbohydrate. As pectins; gums and soluble pentosans werelost in, . solution in the detergent fiber methods, these methods are not very suitable for measuring total unavaila.ble carbohydrate of.cereals. , Significant differences in the unavailable carbohydrate contents were found for the two abnormal soft wheats and a hard wheat variety Nacozari 76 according to the method of Southgate (1969b). Results of further fractionation of unavailable carbohydrate complex by the procedure of Southgate (1969b) are given in Table 7. The abnormal ,soft wheats had significantly higher levels of-water soluble pentosans, pectic substance and lignin than triticales, the normal soft I wheat, and durum wheats. Hard wheats were not significantly different ,; from other testsamples, but had a higher content of pectic substance. 50 TABLE 6 - Unavailable carbohydrate content of test cereals as determined by 4 differe'nt methods \ TABLE 7 - Fractionation of unavailable carbohydrate comples. by'southgate method. 51 TABLE 8 - The effect of various constituents of cereals on the growth 'of T. castaneum larvae and Japanese quail. The data.on weight gain of Tribolium.larvae fed the test cereal ,diets with yeast indicate the .best growth on triticales followed by a softwheat (H-Ra2F2), durum wheats (Mexi chap and Mexi Fg), dnd hard. wheats (Nacozari 76 and Anza). The poorest growth was on the soft wheats with the hiqhest protein contents (Atlas 66 and NapHal/Atlas 66) and on the durum.wneat (Mexicali 75) (Table 8). The correlation coefficient between crude protein and larval growth (Ll) was only 0.36. But the correlation coefficient between unavailable carbohydrate and larval growth was highly significant (r=0.89) for Li.=3.6.9-0.07F. . Due to a shortage of supply, the abnormally shrivelled-softwheats i (Atlas 66 and NapHaljAtlas 66) were not tested with quail&. 'With a crude protein level of 15% i.n test diets, the triticales aqain were significantly superior than durum wheats or hard wheats in supporting the growth of either the Tribolium larvae or of Japanese quail. Soft wheat was jntermediate between one hard wheat (Anza) and 52 the triticales for larval growth. It was also intermeidate between triticales and the rest of'wheat samples (PfO.O1) The level of crude protein in test cereals was kept at 15% which : is not op,timal for the growth of Japanese quail., In our preliminary, : study, if a crude protein level of 30% was used by incorporating a ,large ' amount of isolated soybean protein', the differences in the growth of' : quai 1 due to difference in the carbohydrate, complex of cereals were .' masked. Even for the.estimation of protein quality of cereals, Fernandez et al. (1974) found it desirable to keep'the protein level of their diets at 14%. Our diets incorporated glutamic acid which was not ' deleterious for the larvae or for the Japanese quail, to bring the protein level to about 15%. On these diets, the growth of both the Tribolium larvae and of auail was poorly correlated to the protein content of the test cereals with correlation coefficients value of.023 and ' 0.26, respectively. The. growth was significantly correlated to the available and unavailable carbohydrate contents. The respective correlation coefficients for Tribalium larvae were 0.83 and 0.68.; and for : quail 0.90 and.0.70. Under our conditions, the differences in growth of test organisms due to the carbohydrate complex of cereals are demon' strated more clearly with a low protein .diet. The overall data sugqest that the'growth of Tribolium larvae and of quail was better, correlated to the available carbohydrate content'of . these cereals (r=0.83, 0.90) followed by unavailable carbohydrate con. te.nt (r=0.68. and 0.70), but poorly to the crude protein content (r= I IO.23 and 0.26). A multiple regression using both the carbohydrates and ,; protein did not imporve the correlation coefficient any further. Tribolium larval growth is better on samples higher in amylopectin/ amylose ratio (Vohra et al. 1979). Waxy maize and waxy rice are such samples, No`such significant differences in amylopectin/amylose ratio have been observed in wheats and triticales. . The deleterious effects of high amylose starches from some legumes and tubers on the NPU of casein for rats have been also reported (Rao and Rae,. 1978). This confirmed the earlier findings as reviewed by 'Vohra et al. (1979a) which suggested the importance of amylose/amyloI pectin ratios. . The agreement between the biological evaluation between .Tribolium larval growth and the growth of Japanese auail confirms the usefulness . of bioassay.with insects as a useful tool *for nutritional screening of cereals. , IV. .REFERENCES A.O.A.C. (1975). ''I Official Methods of Analysis' 12th Ed. (Assoc. of Official Analytical Chemists, Washinqton, D.C.) c , ii' FERNANDEZ, R.9 LLJCAS,,'E. and EcGFNNIS, J. '(1974). Poultry Sci. 53:39. ' HELLENDOORN, E.W., NOORDHOFF, M.G. and SLAGMAN, J. (1975). . J:`Sci. 'Fd. Agric.,26:1461. . RAO, C.N. an.d 'RAO, B.S.N. (1978). Br. J. Nutr. 4O:l. zzz 53 ROBERTSON, J.B. (1978). In Topics in Dietary Fiber Research. Spiller, G.A. and Aman, R.J., Eds. (Plenum Press, New York). SOUTHGATE, D.A.T. (1969.a). J. Sci. Fd. Agric. 20:225. . \ SOUTHG'ATE, i. D.A.T. (1969b). J. Sci. Fd. Agric. 20:331. VOHRA, P., SHARIFF, G., ROBINSON, D.W., QUALSETC.0. and GALL, G.A.E.. 18:289. (1978). Nutr. Reports Internat. = VOHRA, P., SHARIFF, G., RO,BINSON, D.W. and QUALSET, C.O. (1979a). Nutr. Reports Internat. G '19:lOl. : SHARIFF, G. and KRATZER, F.H. (1979b). Nutr. Reports hternat- . VOHRA, P., 19:463. z=s `I! WILLIAMS, D.C.., KUZINA, F.@. and HLYNKA, I. (1970). Cereal Chem. 48:411. . WOLF, M.J. (194 ). In Methods in Carbohydrate Chemistry Vol. IV, Whistler, R.L., Ed. (Academic Press, New York).