Improving the quality of sorghum injera by decortication and compositing with tef

Transcription

1 Journal of the Science of Food and Agriculture J Sci Food Agric 85: (25) DOI: 1.12/jsfa.213 Improving the quality of sorghum injera by decortication and compositing with tef Senayit Yetneberk, 1,2 Lloyd W Rooney 3 and John RN Taylor 2 1 Food Science Research Division, Melkassa Agricultural Research Center, PO Box 436, Melkassa, Ethiopia 2 Department of Food Science, University of Pretoria, Pretoria 2, South Africa 3 Cereal Quality Laboratory, Department of Soil and Crop Sciences, Texas A&M University, College Station, TX , USA Abstract: Injera is an Ethiopian fermented pancake-like bread made from cereals, with tef being preferred. Decorticationand compositing with tef were evaluated as methods to improve the injera-making quality of red tannin-free and tannin-containing sorghums. Both decortication and compositing improved sorghum injera quality. Concerning decortication, mechanical abrasion was found to be more effective than hand pounding because acceptable injera was obtained with lower milling loss. Good quality injera was produced at an extraction rate of 54 g kg 1 for tannin-containing and 83 g kg 1 for tannin-free sorghum. With compositing, good quality injera was produced with a 5:5 (w/w) composite of whole tannin-containing sorghum and tef. Both processes reduced the tannin content of the flours, which appeared to relieve the inhibiting effects of tannins on the fermentation. Decortication also seemed to improve sorghum flour injera-making quality by improving flour pasting as a result of reducing the level of interfering substances such as lipids and proteins. In contrast, the improvement brought about by compositing with tef seemed to be due to inherent differences between tef and sorghum starch granules and an increase in the water solubility index of the flour. Compositing seems to be a more useful method of improving sorghum injera quality than decorticating as it avoids the grain loss associated with decortication. 25 Society of Chemical Industry Keywords: injera; sorghum; tef; tannin; decortication; water absorption index; water solubility index; composite flour; pasting INTRODUCTION Injera, a staple food in Ethiopia, is a large pancake-like bread prepared from cereals such as tef and sorghum. It is characterized by having eyes (honeycomb-like holes) in its top surface, which are produced due to the production and escape of carbon dioxide during fermentation and baking. Injera prepared from flour of tef, a tiny millet-like grain, is preferred because it is soft and can be rolled. Notwithstanding this, nearly 8 g kg 1 of the sorghum produced in Ethiopia is used for the production of injera. 1 Both white and ed tannin-free sorghums are used. However, the white tannin-free sorghums are preferred because of the light injera. 1 Tannin-containing sorghums are predominantly used for local beer production. Since ed tannin-free and tannin-containing sorghums are lower priced than white sorghum and tef, it would be most desirable to improve their injeramaking potential through the use of simple, practical technologies. Such technologies include decortication (often incorrectly referred to as dehulling since the sorghum grain does not have a hull) and the use of composite flours. In fact, compositing tef flour with sorghum flour has been found to improve injera texture. 1,2 However, these studies were very limited in scope. The objectives of the work reported here were to evaluate grain decortication and compositing with tef flour as methods to improve the quality of injera made from tannin-containing and tannin-free red sorghums. MATERIALS AND METHODS Grain A red pericarp tannin-containing cultivar (Seredo) and a red tannin-free cultivar (IS 2284) were obtained from the Ethiopian Sorghum Improvement Program (ESIP) at Melkassa Agricultural Research Center, Nazret, Ethiopia. White tef was purchased from a local market. Grain characterization Thousand kernel weight (TKW) was measured by weighing 1 randomly selected unbroken kernels. Correspondence to: John RN Taylor, Department of Food Science, University of Pretoria, Pretoria 2, South Africa jtaylor@postino.up.ac.za Contract/grant sponsor: The Agricultural Research and Training Project of the Ethiopian Agricultural Research Organization Contract/grant sponsor: International Sorghum and Millet Collaborative Research Support Program (Received 2 December 23; revised version received 11 August 24; accepted 17 August 24) Published online 1 February Society of Chemical Industry. J Sci Food Agric /25/$

2 Improving sorghum injera quality Glume was determined by examining the inside of the glume after removing the kernel. 3 The presence of a pigmented testa was determined using the bleach test method. 4 The vanillin HCl method of Burns 5 with sample blank subtraction was used to determine the level of condensed tannins. Pericarp thickness was determined by viewing longitudinally sectioned grains using a scanning electron microscope. Endosperm texture was viewed using a stereomicroscope and images were captured with a camera. Total endosperm and floury endosperm areas and pericarp thickness were measured using the UTHSCSA, Image Tool Software, version 2. (University of Texas Health Science Center, San Antonio, TX, USA). The vitreous area was expressed as a percentage of the total endosperm area. Grain was measured in L, a and b units using a Hunter Lab Color Quest 45/ (Hunter Associates, Reston, VA, USA). Decortication and milling Decortication was performed by hand pounding and mechanical abrasion. For hand pounding, sorghum kernels (1 kg) were washed, placed in a wooden mortar and pounded with a wooden pestle. Samples were removed after 3, 6 and 1 min of pounding. The number of pounding strokes was recorded for each time interval. Decorticated kernels were on a clean cloth and dried in the sun. The bran and decorticated grains were separated by winnowing. Mechanical abrasion was performed using a tangential abrasive dehulling device (TADD) according to Reichert et al. 6 Sorghum grain (1 g) was decorticated for 1, 2, 3, 4 and 5 min using a TADD fitted with sand paper of 6 grit (Norton type R284 metalite) (Saint-Gobain Abrasives, Isando, South Africa). Extraction rate as a proportion (g kg 1 ) of decorticated grain recovered was calculated for both methods. Samples were milled to flours using an Udy cyclone mill (Udy Corp, Fort Collins, CO, USA) fitted with a 2 µm opening screen size. Tef was not decorticated prior to milling. Composite flours Wholegrain flour of the tannin-containing sorghum (Seredo) was composited with tef flour at five levels: 833, 667, 5, 333 and 167 g kg 1 ; 1 g kg 1 Seredo flour and 1 g kg 1 tef flour were also included. Flour characterization The effects of compositing sorghum flour with tef flour on flour water absorption index (WAI) and water solubility index (WSI) were determined according to Anderson et al. 7 Pasting properties of flours from the sequentially decorticated tannin-containing sorghum (Seredo) and composite flours of tef and Seredo were determined using a rapid Visco analyzer (RVA) model 3D (Newport Scientific, Warriewood, Australia). Flour (4 g, 14 g kg 1 moisture basis) was mixed with 25 ml distilled water in the RVA sample canister. A programmed heating and cooling cycle was used, where the suspension was held at 5 C for 1 min, heated at a uniform rate to 93 C for 8 min and then held at 93 C for 5 min before cooling to 5 C within 8 min, and finally held at 5 C for 1 min. The RVA parameters measured were: peak viscosity (PV) the maximum hot paste viscosity at 93 C; hot paste viscosity (HPV) the trough at minimum hot paste viscosity; and cold paste viscosity (CPV) viscosity after cooling to 5 C and holding at this temperature. Preparation of injera Injera was prepared from flours of hand pounded and mechanically abraded sorghum and the composite flours of Seredo and tef using the method described by Yetneberk et al. 8 Sensory evaluation of injera A panel of six people, who were injera consumers, evaluated three replicate samples of injera. The attributes used, which were generated by the authors, were injera top and bottom surface, description of the eyes (honeycomb structure of the top surface of the injera), texture, taste, aftertaste and overall rating. Rolled pieces of injera (3 mm wide) were presented to the panelists on a tray at ambient temperature (about 25 C) within 2 h after baking. Statistical analysis Analysis of variance was performed on the data to establish significant (p <.5) differences between the samples. Linear regression analysis was done to establish relationships between extraction rate and flour pasting properties. Table 1. Grain characteristics of Seredo (tannin-containing) and IS 2284 (tannin-free) sorghum cultivars and tef (local landrace) Tannins (g kg 1 Pericarp Endosperm vitreousness Grain TKW Glume Pigmented catechin thickness (percentage of total Cultivar (g) Color testa equivalent) (µm) endosperm area) L a b Seredo 21.5 (.3)b Purple Yes 73. (2.7)a (26.5)b 36.3 (3.8)a 36.7 (.4)b 9.5 (.2)b 12.7 (.4)a IS (1.7)c Purple No 4.7 (.8)b 6. (1.)a 48. (5.7)b 33.2 (.5)a 16.4 (.4)c 13. (.2)a Tef.3 (.)a Tan No.5 (.)c ND ND 64.8 (.7)c 3.2 (.1)a 14.4 (.1)b TKW, thousand kernel weight. Values in parentheses are standard deviations of three replications. Values followed by the same letter in the same column are not significantly different (p >.5). ND, not determined. J Sci Food Agric 85: (25) 1253

3 S Yetneberk, LW Rooney, JRN Taylor (A) 1 A C VE E FE G B D FE G VE Extraction rate (g kg-1) Number of pounding strokes P (B) 1 1 P Figure 1. Microstructure of Seredo (tannin-containing) and IS 2284 (tannin-free) sorghum grains. (A) Seredo endosperm; (B) IS 2284 endosperm; (C) Seredo pericarp; (D) IS 2284 pericarp. P = pericarp; VE = vitreous endosperm; FE = floury endosperm; G = germ. RESULTS AND DISCUSSION Grain characterization Seredo had a lower TKW (21.5 g) compared to IS 2284 (35.9 g; Table 1). TKW of tef was only.3 g. Seredo and IS 2284 had purple glume, while tef had tan glume. The tef grain was much lighter (L value 64.8) compared with the sorghums; Seredo L value was 36.7 and IS 2284 L value was Seredo contained a pigmented testa and had a tannin content of 73 g kg 1 catechin equivalents, while IS 2284 and tef were tannin-free. Seredo had a much thicker and more loosely packed pericarp with starch granules in the mesocarp, while IS 2284 had a thin, tightly packed pericarp with practically no mesocarp (Fig 1). Seredo had a more floury endosperm compared with IS Decortication Two methods of decortication were evaluated, traditional hand pounding with a pestle and mortar and mechanical abrasion (simulating industrial mechanical decortication). The objective of decorticating the sorghums was to remove the bran, germ and testa (which contained tannins in Seredo) in order to improve the, taste and appearance of the injera. With hand pounding an increase in the number of pounding strokes was directly related to a decrease in extraction rate for both sorghum cultivars [Fig 2(A)]. The extraction rate of Seredo was significantly lower (643 g kg 1 ; p <.5) after 4 pounding strokes compared with IS 2284 (69 g kg 1 ). The softer endosperm and thick pericarp 6,9,1 explain the reduced milling yields from Seredo. Extraction rate (g kg-1) Time (min) Figure 2. Effect of decortication on sorghum grain extraction rate. (A) Hand pounding; (B) mechanical abrasion with a TADD. (ž) Seredo (tannin-containing); ( ) IS 2284 (tannin-free); ( ) tannin content, Seredo. Similar results were obtained using mechanical abrasion [Fig 2(B)]. After 4 min of abrasion, the extraction rate for Seredo was 668 g kg 1 compared with 712 g kg 1 for IS As stated, the extraction rate depends on the hardness of the grain. For Seredo, as the time of abrasion was increased to 5 min, decorticated grain yield declined to only 54 g kg 1. This is because soft floury grains have a tendency to disintegrate during decortication. 11 Notwithstanding this, abrasion of Seredo for 5 min reduced the tannin content of the grain from 73 to 13 g kg 1 catechin equivalents [Fig 2(B)], a reduction of 82%. Beta et al 12 obtained a similar reduction in tannin content (71 81%) with abrasion of Zimbabwean tannincontaining sorghums, but with very much lower decortication losses (decorticated grain recovered g kg 1 ), due to the grains being harder. Sensory responses for injera made from handpounded Seredo and IS 2284 flours are presented in Table 2. Injera made from whole Seredo flour had brown top and bottom surfaces, without eyes. It had a sticky texture and a bitter taste. Hence it was rated as poor. The tannins in Seredo were responsible for the brown and the bitter taste of the injera. Tannins are known to cause dark and astringent bitterness in foods prepared from whole sorghum. 13 Carbon Tannin (g kg-1 catechin equiv) 1254 J Sci Food Agric 85: (25)

4 Improving sorghum injera quality Table 2. Sensory panel responses of injera prepared from sorghum flours of Seredo and IS-2284 decorticated by hand pounding and with TADD at different extraction rates Injera quality attributes Cultivar Extraction Rate (g kg 1 ) Top surface Bottom surface Description of the eyes a Texture Taste Aftertaste Overall rating (A) Hand pounding Seredo (tannincontaining) 1 b Brown Brown None (flat injera) Sticky Bitter Bitter Poor 84 Brown Brown None (flat injera) Sticky Bitter Bitter Poor 776 Brown Brown Few, small, Sticky Bitter Bitter Poor 643 Light brown Light brown Few, small, Sticky Slightly bitter Slightly bitter Poor IS b Red Light brown Few, large, Crumbly Slightly sweet Slightly sweet Poor (tannin-free) 845 Red Red Few, large, Slightly soft Slightly sweet Slightly sweet Fair 767 Red Red Many, small, Slightly soft Sour Slightly sour Good 683 Light red Light red Many, small, Soft Sour Slightly sour Very good (B) Abrasion with a TADD Seredo (tannincontaining) 1 b Brown Brown None (flat injera) Sticky Bitter Bitter Poor 895 Brown Brown None (flat injera) Sticky Bitter Bitter Poor 83 Brown Brown Few, small, Sticky Bitter Slightly bitter Poor 753 Light brown Light brown Few, small, Sticky Slightly bitter Slightly bitter Poor 668 Red Red Few, small, Sticky Slightly bitter Slightly bitter Fair 54 Red Red Many, small, Soft Slightly sour Slightly sour Good IS b Light brown Light brown Few, large, Crumbly Slightly sweet Slightly sweet Poor (tannin-free) 918 Red Red Few, large, Crumbly Slightly sweet Slightly sweet Fair 833 Red Red Many, small, Soft Sour Slightly sour Good 767 Light red Light red Many, small, Soft Sour Slightly sour Good 712 Light red Light red Many, small, Soft Slightly sour Bland Very good a Eyes (cells) refer to the honeycomb-like structure of the top surface of injera formed due to escaping gas bubbles during baking. b Whole flour (undecorticated). dioxide produced during fermentation is known to play a fundamental role in the formation of cellular structure of leavened breads. 14 Thus the absence of eyes in whole flour Seredo injera is indicative of very little carbon dioxide being produced during fermentation. Since condensed tannins are associated with anti-microbial 15 and enzyme-inhibitory activity, 16 it can be assumed that the tannins inhibited the fermentation process. Injera prepared from Seredo flour of the lowest extraction rate examined for hand pounding (643 g kg 1 ) had a slightly bitter taste and was sticky with small, eyes. This was presumably because the flour still contained significant levels of tannins. The tannin-free cultivar IS 2284 made good injera from hand-pounded flour of higher extraction rate (767 g kg 1 ). At 683 g kg 1 extraction, a similar rate to the lowest examined for the tannincontaining Seredo, the injera was rated very good. The much better injera-making quality of IS 2284 at similar extraction rates to Seredo was presumably due to the better milling characteristics of IS 2244 as a result of it being harder and the fact that it was tannin-free. In contrast to hand pounding, mechanical abrasion of Seredo resulted in injera of good quality at an extraction rate of 668 g kg 1 [Table 2(B)]. This is presumably because at this extraction rate the tannin content of the flour had been reduced to only 2 g kg 1 [Fig 2(B)]. Thus, mechanical abrasion, which works by polishing the grains against the abrasive surface J Sci Food Agric 85: (25) 1255

5 S Yetneberk, LW Rooney, JRN Taylor and each other, 17 removed the tannin-containing layer more effectively compared with hand pounding. In hand pounding the pestle causes a mechanical shock, generating strong interactive forces between the grains and between the grain and the equipment, which is said to result in the breaking off of large pieces of hull. 17 However, in the case of soft, tannin-containing grains such as Seredo, considerable grain breakage appears to occur as a result of these strong interactive forces. For the tannin-free red sorghum IS 2284, good quality injera was obtained by mechanical abrasion at an extraction rate of 833 g kg 1, which was very much higher than for the tannin-containing Seredo and also higher than for hand-pounded IS To help explain the improvement in injera-making quality with decortication, flours from sequentially TADD abraded tannin-containing sorghum (Seredo) were tested for their pasting properties. Whole Seredo flour had the lowest PV, HPV and CPV (Fig 3). With successive abrasion, PV, HPV and CPV increased markedly. Linear regression analyses of the relationships between extraction rate and PV, HPV and CPV gave r-values of.94,.98 and.98, respectively, indicating that the extent of decortication and these pasting parameters were closely related in an inverse manner. PV in particular indicates the water-holding capacity of starch and can be used as a measure of the resistance of starch granules to swelling. 18,19 The relationship between PV and extent of decortication suggests that one or more of the grain components capable of inhibiting starch swelling were progressively removed with the bran. Some of the increase in viscosity with decortication could also have been due to the small increase in flour starch content which occurs as a result of bran removal. 2 Proteins and lipids are involved in resistance to starch swelling. 21 Reductions in starch swelling and gelatinization rates in the presence of protein in sorghum have been associated with the tendency of grain protein to act as a physical barrier to starch swelling. 22 It is probably of significance that the highest concentration of endosperm protein is Viscosity (RVU) Temperature profile 1 15 Time (min) Figure 3. Effect of sequential decortication of a tannin-containing sorghum (Seredo) on pasting properties of the flours. Thin solid line = 54 g kg 1 ; dashed line = 668 g kg 1 ; dotted line = 753 g kg 1 ; dashed and dotted line = 83 g kg 1 ; thick solid line = 895 g kg 1 ; dashed and double dotted line = 1 g kg 1 (undecorticated) Temperature ( C) sorghum is in the peripheral endosperm, 23 and much of the lipid is in the germ, 24 both of which would be removed by decortication. Decortication would also enrich the starch content of the decorticated grain through the removal of grain components such as non-starch polysaccharides, proteins and lipids which are concentrated predominantly in the outer endosperm and the germ. 24 It is therefore additionally possible that the successive removal of those grain components in the bran may have led to a progressive elimination of their diluting effects on the starch concentration, leading to increasingly higher PV, HPV and CPV values of the paste. Flours containing higher amounts of starches have been associated with generally higher paste viscosities. 21 Also, Parker et al 25 showed that starch plays a major functional role in the formation of the continuous amorphous structural matrix of injera during baking. Further, Subramanian and Jambunathan 26 reported a positive correlation between sorghum injera eye quality and the starch content of the flour. Compositing Sensory responses of injera made from composite flours of whole tannin-containing sorghum (Seredo) and tef are presented in Table 3. As stated, injera made from 1 g kg 1 Seredo flour had brown top and bottom surfaces, was flat (without eyes) and had sticky texture and bitter taste, and was rated as poor. In contrast, injera from 1 g kg 1 tef flour had white top and white bottom surfaces, many small eyes, very soft texture and a bland aftertaste, and was rated excellent. As mentioned, the quality defects of injera from Seredo are probably due to a considerable extent to tannins inhibiting the fermentation process. At a low level of compositing with tef (167 g kg 1 ), there were only a few small and eyes on the surface of the injera. Also, other attributes of injera such as, taste and texture were not improved. This can be attributed to minimal dilution of the tannin in the whole Seredo flour. Injera of good quality was produced with a 5:5 blend of whole Seredo and tef flours. At this level of substitution, all the quality attributes of injera (, eye number and distribution, texture, taste and aftertaste) were improved. It appears that, as substantial tannin dilution had taken place, this enabled the yeast and lactic acid bacteria to ferment the dough. Additionally, tef appeared to impart its intrinsic flour quality to positively affect injera quality. The improvement in injera quality continued as the proportion of tef flour in the composite was increased beyond 5 g kg 1. The changes in WAI and WSI when whole flour of Seredo and tef were composited are shown in Fig 4. As the proportion of tef flour in the composite increased, WSI increased progressively, while WAI declined somewhat. The increase in WSI agrees with the observation that, during mixing, tef dough tended to be more sticky compared with sorghum. Water-soluble 1256 J Sci Food Agric 85: (25)

6 Improving sorghum injera quality Table 3. Sensory panel responses of injera prepared from composite flours of whole tannin-containing sorghum (Seredo) and tef flours composited at different proportions Injera quality attributes Compositing proportion Top surface Bottom surface Description of the eyes c Texture Taste Aftertaste Overall rating 1 g kg 1 S a Brown Brown None (flat injera) Sticky Bitter Slightly bitter Poor 833 g kg 1 S + Brown Brown Few, small, Sticky Bitter Slightly bitter Poor 167gkg 1 T b 667 g kg 1 S + Brown Brown Few, small, Sticky Slightly bitter Slightly bitter Fair 333gkg 1 T 5 g kg 1 S + Light brown Light brown Many, small, evenly Soft Slightly sour Slightly sour Good 5gkg 1 T 333 g kg 1 S + Light red Light red Many, small, evenly Soft Slightly sour Slightly sour Very good 667gkg 1 T 167 g kg 1 S + Light red Light red Many, small, evenly Soft Slightly sour Bland Very good 833gkg 1 T 1 g kg 1 T White White Many, small, evenly Very soft Slightly sour Bland Excellent a S = sorghum (Seredo) whole flour. b T = tef whole flour. c Eyes (gas cells) refer to the honeycomb-like structure of the top surface of injera formed due to escaping gas bubbles during baking. WAI (g g-1) Tef flour (g kg-1) Figure 4. Changes in water solubility index and water absorption index of composite whole flours of a tannin-containing sorghum (Seredo) and tef composited in different proportions. (ž) WSI;( ) WAI. WSI (g kg-1) Viscosity (RVU) Temperature profile Time (min) Temperature ( C) components in the tef flour could have modified the dough rheology and the texture of injera positively. This was manifested by the softer texture of injera as the proportion of tef was increased. The decrease in WAI can be attributed to a reduction in damaged starch as the proportion of tef flour increased. Tef starch exists as compound starch granules made up of many tiny granules of 2 6 µm diameter, 27,28 whereas sorghum starch exists as large single starch granules of about 2 µm diameter. 29 Tef starch granules, because they are so much smaller, are presumably much less prone to damage during milling than sorghum granules and hence absorb less water. Figure 5 shows that whole Seredo sorghum flour exhibited significantly higher PV, HPV and CPV than whole tef flour. As the proportion of tef flour increased, the PV, HPV and CPV progressively decreased, in contrast to the effect of decorticating Seredo. For the sake of clarity only the 5:5 sorghum:tef curve is shown. The difference in pasting properties of sorghum and tef flours could also be related to inherent morphological differences in their starches. Working with potato, wheat and maize starches, Fortuna et al 18 concluded that the larger starch granules in a population have higher swelling capabilities, while Figure 5. Effect of compositing a tannin-containing sorghum (Seredo) with tef on pasting properties of the flours. Solid line = 1 g kg 1 sorghum; dashed line = 5 g kg 1 sorghum:5 g kg 1 tef; dotted line = 1 g kg 1 tef. smaller granules have higher resistance to swelling. This may apply to tef starch vs sorghum starch. CPV is related to the ability of the starch paste to form a gel after cooling. 21 With high setback, more syneresis is likely to take place. 3 For tef starch, Bultosa et al 28 reported a low setback viscosity and slow syneresis. This is probably related to the softer texture of tef injera compared with sorghum injera. CONCLUSIONS Decortication and compositing with tef are both effective ways of improving the injera-making quality of tannin-containing and tannin-free red sorghum. Decortication improves the and other quality attributes of injera through reduction in the level of non-starch components of the grain. In the case of tannin-containing sorghum, decortication also removes some of the tannins, improving injera fermentation. Mechanical abrasion is a more effective J Sci Food Agric 85: (25) 1257

7 S Yetneberk, LW Rooney, JRN Taylor method compared with hand pounding because acceptable injera can be obtained at higher flour extraction levels. Compositing of whole tannincontaining sorghum flour with tef flour improves injera quality, primarily by diluting the tannins. Also, differences in starch granule characteristics and the higher WSI of tef flour appear to positively influence the quality of injera. Compositing seems to be a more useful method of improving sorghum injera quality than decortication as it avoids the grain loss associated with decortication. ACKNOWLEDGEMENTS The Agricultural Research and Training Project (ARTP) of The Ethiopian Agricultural Research Organization (EARO) supported this research. Senayit Yetneberk gratefully acknowledges partial financial assistance from the International Sorghum and Millet Collaborative Research Support Program (INTSORMIL). REFERENCES 1 Gebrekidan B and GebreHiwot B, Sorghum injera preparation and quality parameters, in Proceedings of the International Symposium on Sorghum Grain Quality, Ed by Rooney LW and Murty DS. ICRISAT, Patancheru, India, pp (1982). 2 Yetneberk S and Haile T, Studies on methods to improve injera making qualities of two brown sorghum cultivars, in Proceedings of the Eighth EARSAM Regional Workshop on Sorghum and Millet, Ed by Mukuru SZ and King SB. ICRISAT, Patancheru, India, pp (1992). 3 Rooney LW and Miller FR, Variation in the structure and kernel characteristics of sorghum, in Proceedings of the International Symposium on Sorghum Grain Quality, Ed by Rooney LW and Murty DS. ICRISAT, Patancheru, India, pp (1982). 4 Waniska RD, Hugo LF and Rooney LW, Practical methods to determine the presence of tannins. J Appl Poultry Res 1: (1992). 5 Burns RE, Method for estimation of tannin in grain sorghum. Agron J 63: (1971). 6 Reichert RD, Youngs CG and Oomah BD, Measurement of grain hardness and dehulling quality with multiple, tangential abrasive dehulling device (TADD), in Proceedings of the International Symposium on Sorghum Grain Quality, Ed by Rooney LW and Murty DS. ICRISAT, Patancheru, India, pp (1982). 7 Anderson RA, Conway HR, Pfeifer VF and Griffin EL, Gelatinization of corn grits by roll and extrusion cooking. Cereal Sci Today 14:4 7 (1969). 8 Yetneberk S, De Kock HL, Rooney LW and Taylor JRN, Effects of sorghum cultivar on injera quality. Cereal Chem 81: (24). 9 Sheperd AD, Laboratory abrasive decorticating mill for small grains. Cereal Chem 56: (1979). 1 Lawton JW and Faubion JM, Measuring kernel hardness using the tangential abrasive dehulling device. Cereal Chem 66: (1989). 11 Rooney LW, Waniska RD and Subramanian R, Overcoming constraints to utilization of sorghum and millet, in Proceedings of the International Conference on Genetic Improvement of Sorghum and Millet. INTSORMIL and ICRISAT, Lubbock, TX, pp (1997). 12 Beta T, Rooney LW and Taylor JRN, Effect of chemical conditioning on the milling of high-tannin sorghum. JSci Food Agric 8: (2). 13 Bacon JR and Rhodes MJC, Binding affinity of hydrolysable tannins to parotid saliva and to proline-rich protein derived from it. J Agric Food Chem 48: (2). 14 Bloksma AH, Rheology of the breadmaking process. Cereal Foods World 35: (199). 15 Ebi GC, Ifeanacho CJ and Kamalu TN, Antimicrobial properties of Uvaria chamae stem bark. Fitoterapia 7: (1999). 16 Scalbert A, Antimicrobial properties of tannins. Phytochemistry 3: (1991). 17 Munck L, Bach Knudsen KE and Axtell JD, Milling processes and products as related to kernel morphology, in Proceedings of the International Symposium on Sorghum Grain Quality, Ed by Rooney LW and Murty DS. ICRISAT, Patancheru, India, pp 2 21 (1982). 18 Fortuna T, Januszewska R, Juszczak L, Kielski A and Palasiński M, The influence of starch pore characteristics on pasting behavior. Intl J Food Sci Technol 35: (2). 19 Li JY and Yeh AI, Relationships between thermal and rheological characteristics and swelling power for various starches. J Food Eng 5: (21). 2 Serna-Saldivar S and Rooney LW, Structure and chemistry of sorghum and millets, in Sorghum and Millets: Chemistry and Technology, Ed by Dendy DAV. American Association of Cereal Chemists, St Paul, MN, pp (1995). 21 Whistler RL and BeMiller JN, Carbohydrate Chemistry for Food Scientists. American Association of Cereal Chemists, St Paul, MN (1999). 22 Chandrashekar A and Kirleis AW, Influence of protein on starch gelatinization in sorghum. Cereal Chem 65: (1988). 23 Shull JM, Chandrashekar A, Kirleis AW and Ejeta G, Development of sorghum (Sorghum bi (L) Moench) endosperm in cultivars of varying hardness. Food Struct 9: (199). 24 Wang S, Sosulski K, Sosulski F and Ingledew M, Effects of sequential abrasion on starch composition of five cereals for ethanol fermentation. Food Res Int 3:63 69 (1997). 25 Parker ML, Umeta M and Faulks RM, The contribution of flour components to the structure of Injera, Ethiopian fermented bread made from tef (Eragrostis tef ). J Cereal Sci 1:93 14 (1989). 26 Subramanian V and Jambunathan R, Sorghum grain characters in relation to traditional foods of Ethiopia, Kenya, India and Sudan, in Proceedings of the Seventh EARSAM Regional Workshop on Sorghum and Millet. OUA/SAFGRAD/ICRISAT, Nairobi, Kenya, pp (199). 27 Umeta M and Parker M, Microscopic studies of the major macro-components of seeds, dough and injera from tef (Eragrostis tef ). SINET Ethiop J Sci 19: (1996). 28 Bultosa G, Hall AN and Taylor JRN, Physico-chemical characterization of grain tef [Eragrostis tef (Zucc) Trotter] starch. Starch/Stärke 54: (22). 29 Hoseney RC, Principles of Cereal Science and Technology, 2nd edn. American Association of Cereal Chemists, St Paul, MN (1994). 3 Newport Scientific, Operation Manual for the Series 3 Rapid Visco-Analyser. Newport Scientific, Warriewood, Australia, pp (1995) J Sci Food Agric 85: (25)