UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Storage changes in natural and model lipid-protein systems El-Lakany, Safaa 1972

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1972_A1 E4.pdf [ 7.66MB ]
Metadata
JSON: 831-1.0101245.json
JSON-LD: 831-1.0101245-ld.json
RDF/XML (Pretty): 831-1.0101245-rdf.xml
RDF/JSON: 831-1.0101245-rdf.json
Turtle: 831-1.0101245-turtle.txt
N-Triples: 831-1.0101245-rdf-ntriples.txt
Original Record: 831-1.0101245-source.json
Full Text
831-1.0101245-fulltext.txt
Citation
831-1.0101245.ris

Full Text

STORAGE CHANGES IN NATURAL AND MODEL LIPID-PROTEIN SYSTEMS by SAFAA EL-LAKANY B . S c , University of Alexandria, Alexandria (Egypt), 1965 M . S c , University of Br i t i sh Columbia, Vancouver, (B.C.) 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of POULTRY SCIENCE We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1972 In p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e fo r reference and study . I f u r t h e r agree t h a t permiss ion f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood that copying o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l ga in s h a l l not be a l lowed wi thout my w r i t t e n p e r m i s s i o n . Depa rtment The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8 , Canada Abstract The primary objective of the study was to investigate the nutr i t ional significance of the interactions which occur between oxidized fat and protein in foods during storage at room temperature (21°C) and at -20°C. Herring meal, because of i t s high content of protein and of highly unsaturated fat , was selected as the material for investigation. Studies were also conducted on a considerably less complex system, con-s is t ing of egg albumin and herring o i l (a model system). The effects of antioxidant-treatment on the chemical and nutr i t ive properties of herring meal were determined as wel l . Since freeze-drying has become an impor-tant process for the preservation of foods, freeze-dried herring meal was also included as a test material. Oxidation of l ip ids in the herring meals and in the model systems was monitored through measurements of iodine value, peroxide value, thio-barbituric acid value and ultraviolet spectrophotometry. The changes in the amount of ether-extractable material were also followed during the storage period. The following conclusions were drawn from the results of the study: 1. The antioxidant, 6-ethoxy- l , 2 dihydro-2, 2, 4-tr imethyl-quinoline (ethoxyquin), added to the herring meal at a concentration of 0.025% considerably l imited the decrease in the ether-extractabil i ty of l ip ids and the decrease in iodine value of the ether-extract which i i i i i occurred in meal not treated with antioxidant. The effect of the ant i -oxidant was evident at both storage temperatures. 2. Peroxide value, thiobarbituric acid absorbance at 535 nm, and the ultraviolet absorption at 233 nm were much higher in the untreated herring meals than in the antioxidant-treated ones at both storage tem-peratures . 3. The available lysine in the herring meals declined during the storage period. The rate of decline was most pronounced in the untreated meal stored at room temperature. 4. Pepsin d igest ib i l i t y tests showed that antioxidant-treated meals and untreated herring meal stored at -20°C contained the highest amounts of digestible protein while the untreated herring meal stored at room temperature contained the lowest amounts of digestible protein throughout the storage period. 5. The solvent-extracted meals, whether stored at room temper-ature or at -20°C exhibited only a sl ight decrease in the available lysine content and in pepsin -d igest ib i l i ty . 6. The peptide maps showed no marked changes in the pattern of peptides released by the action of pepsin on the different herring meals during the storage period. 7. With one exception the herring meals were similar in nut r i -tive value as sources of supplementary protein or of energy, even after storage for 10 months. The unstabilized herring meal stored at room temperature, however, showed signif icant depressions of metabolizable energy value and supplementary protein quality. iv 8. In vivo d igest ib i l i t y tests showed that the antioxidant-treated herring meals were digested much faster in the small intestine of chickens compared to the untreated meals regardless of the storage temperature. It was also noted that antioxidant-treated and untreated meals which had been stored at -20°C for 11 months were digested at faster rates than those similar ly treated but stored at room temperature for the same period of time. 9. Solvent-extracted meal stored at -20°C gave higher metabo-l i zable energy values and showed better protein quality than the solvent-extracted meal stored at room temperature. Similar ly , in vivo d igest ib i l i t y tests indicated that solvent-extracted meal stored at -20°C was digested at a faster rate as compared to the meal stored at room temperature. 10. Ether-extractable fat from freeze-dried meals (whether stored under air or under nitrogen) and from presscake showed- considerable oxidative deterioration as measured by iodine value, thiobarbituric acid value, peroxide value and ultraviolet spectrophotometry. A decrease in the available lysine during the storage period was also evident. Pro-tein d iges t ib i l i t y , as measured by pepsin solubi l i zat ion, declined s l ight ly throughout the storage period. No changes in the peptide maps of these meals were observed. 11. In the model system, the presence of protein promoted oxida-tion in the o i l f ract ion, irrespective of the storage temperature. In addition, some alterations occurred in the protein fraction as a result of storage in mixture with o i l . There were no differences in the peptide V maps of the pepsin digests of the model system and of albumin stored at the two temperatures. There was, however, in both cases, a general decrease in the number and intensity of the peptide spots in the maps after storage of the samples. v i TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES . . . • . . v i i i LIST OF FIGURES x i i ACKNOWLEDGMENT xv i i Chapter 1. INTRODUCTION 1 2. REVIEW OF LITERATURE 5 A. LIPID OXIDATION 5 B. INTERACTION BETWEEN PROTEIN AND OXIDIZED LIPID . . . . . 16 C. HERRING MEAL 21 3. MATERIALS AND METHODS . • 30 A. PREPARATION OF THE MEALS 30 B. PREPARATION OF THE MODEL SYSTEM 31 C. CHEMICAL ANALYSES 31 D. BIOLOGICAL EVALUATION OF THE HERRING MEAL 33 Metabolizable Energy 33 Protein Quality 35 Digest ib i l i ty Test (In Vivo) 36 Fractionation on Sephadex Gel 37 4. RESULTS 43 A. CHEMICAL TESTS 43 v i i Chapter Page Herring Meal 43 Model System • 66 B. BIOLOGICAL ASSAYS 85 Metabolizable Energy Tests 85 Experiment 1 85 Experiment 2 ' 85 Protein Quality Tests 87 Experiment 1 87 Experiment 2 88 In Vivo Digest ib i l i ty 90 5. DISCUSSION . . 110 A. CHEMICAL TESTS . . . . . . . . 110 B. BIOLOGICAL TESTS 129 6. CONCLUSIONS 145 REFERENCES 146 APPENDIX . . . . 159 v i i i LIST OF TABLES Table Page 1. Herring Meals Employed in the Biological Tests 40 2. Composition of Basal Diet 41 3 . Composition of Control Diet and of Diets Containing Fish Meal A Stored for Five Months at 21°C 42 4. Metabolizable Energy Values of Herring Meals Stored for Four Months. Growth Response and Eff iciency of Feed Conversion of Chicks Fed Herring Meals Stored for Five Months 92 5. Metabolizable Energy Values of Herring Meals Stored for Nine Months. Growth Response and Eff iciency of Feed Conversion of Chicks Fed Herring Meals Stored for Ten Months 93 6. Supplementary Protein Values of Herring Meals Stored for Five Months 94 7. Supplementary Protein Values of Herring Meals Stored for Ten Months 95 8. Nitrogen and Dry Matter Content of the Lumen of the Small Intestine of Adult Chickens lh Hour After Ingesting the Different Herring Meals Stored for 11 Months 96 9. The Relative Amounts of Dissolved Proteins, Peptides, and Free Amino Acids in the Soluble Nitrogen Fraction of the Contents of the Small Intestine of Chickens, 1% Hour After Ingesting Different Herring Meals. The Amounts are Expressed as Area Units under the Graphs in Figures 34-45 . 97 APPENDIX 3(B) Composition of Diets Containing Fish Meal B Stored for Five Months at 21°C 160 3(C) Composition of Diets Containing Fish Meal C Stored for Five Months at -20°C 160 3(D) Composition of Diets Containing Fish Meal D Stored for Five Months at -20°C . . . 161 ix Table Page 3(E) Composition of Diets Containing Fish Meal E Stored for Five Months at 21°C 161 3(F) Composition of Diets Containing Fish Meal F Stored for Five Months at -20°C . . 162 3(G) Composition of Diets Containing Fish Meal A Stored for Ten Months at 21°C 162 3(H) Composition of Diets Containing Fish Meal B Stored for Ten Months at 21°C . . . 163 3(1) Composition of Diets Containing Fish Meal C Stored for Ten Months at -20°C 163 3(J) Composition of Diets Containing Fish Meal D Stored for Ten Months at -20°C 164 3(K) Composition of Diets Containing Fish Meal E Stored for Ten Months at 21°C 164 3(L) Composition of Diets Containing Fish Meal F Stored for Ten Months at -20°C . . 165 4(A) Analysis of Variance and Duncan's Multiple Range Test of the Metabolizable Energy Values of the F i rs t Experiment, Treatments A, B, C and D 166 4(B) Analysis of Variance of the Metabolizable Energy Values of the F i rst Experiment, Treatments E and F 166 4(C-a, b Analysis of Variance of the Protein Quality in the F i rs t and c) Experiment (Expressed as Gain in Weight) for Treatments A, B, C and D 167 4(D-a, b Analysis of Variance of the Protein Quality in the F i rs t and c) Experiment (Expressed as Gain in Weight) for Treatments E and F 168 4(E-a, b Analysis of Variance of the Protein Quality in the F i rs t and c) Experiment (Expressed as Diet Efficiency) for Treatments A, B, C and D 169-70 4(F-a, b Analysis of Variance of the Protein Quality in the F i rs t and c) Experiment (Expressed as Diet Eff iciency) for Treatments E and F 171 X Table Page 4(G) Analysis of Variance of the Protein Quality in the F i rs t Experiment (Including 3 Levels of Supplementary Protein), Expressed as Gain in Weight for Treatments A, B, C and D 172 4(H) Analysis of Variance of the Protein Quality in the F i rst Experiment (Including 3 Levels of Supplementary Protein), Expressed as Gain in Weight for Treatments E and F . . . . 172 4(1) Analysis of Variance of the Protein Quality in the F i rs t Experiment (Including 3 Levels of Supplementary Protein), Expressed as Diet Eff iciency for Treatments A, B, C and D 173 4(J) Analysis of Variance of the Protein Quality in the F i rs t Experiment (Including 3 Levels of Supplementary Protein), Expressed as Diet Eff ic iency for Treatments E and F . . . 173 4(A) Analysis of Variance and Duncan's Multiple Range Test of the Metabolizable Energy Values of the Second Experiment, Treatments A, B, C and D 174 5(B) Analysis of Variance of the Metabolizable Energy Values of the Second Experiment Treatments E and F 174 5(C) Analysis of Variance and Duncan's Multiple Range Test of the Metabolizable Energy Values of the F i rst and Second Experiments, Treatments A, B, C and D 175 5(D) Analysis of Variance and Duncan's Multiple Range Test of the Metabolizable Energy Values of the F i rs t and Second Experiments, Treatments E and F 175 5(E-a, b Analysis of Variance of the Protein Quality in the Second and c) Experiment (Expressed as Gain in Weight) for Treatments A, B, C and D 176 5(F-a, b Analysis of Variance of the Protein Quality in the Second and c) Experiment (Expressed as Gain in Weight) for Treatments E and F . 177 5(G-a, b Analysis of Variance of the Protein Quality in the Second and c) Experiment (Expressed as Diet Eff iciency) for Treatments A, B, C and D 178-79 5(H-a, b Analysis of Variance of the Protein Quality in the Second and c) Experiment (Expressed as Diet Eff iciency) for Treatments E and F 180 XX Table Page 5(1) Analysis of Variance of the Protein Quality in the Second Experiment (Including 3 Levels of Supplementary Protein), Expressed as Gain in Weight for Treatments A, B, C and D . 181 5(J) Analysis of Variance of the Protein Quality in the Second Experiment (Including 3 Levels of Supplementary Protein), Expressed as Gain in Weight for Treatments E and F . . . . 181 5(K) Analysis of Variance of the Protein Quality in the Second Experiment (including 3 Levels of Supplementary Protein), Expressed as Diet Eff iciency for Treatments A, B, C and D 182 5(L) Analysis of Variance of the Protein Quality in the Second Experiment (Including 3 Levels of Supplementary Protein), Expressed as Diet Eff ic iency for Treatments E and F . . . 182 x i i LIST OF FIGURES Figure Page 1. Ether-Extractable Fat from Presscake Stored at -20°C Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in Air 49 2. Ether-Extractable Fat from Herring Meal Treated with Antioxidant and Stored at 21°C, Herring Meal Treated with Antioxidant and Stored at -20 C, Herring Meal Stored at 21°C and Herring Meal Stored at -20°C 50 3. Iodine Values of Ether-Extractable Fat of Presscake Stored at -20 C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in Ai r 51 4. Iodine Values of Ether-Extractable Fat of Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20 C, Herring Meal Stored at 21°C and Herring Meal Stored at -20°C 52 5. Peroxide Values of Ether-Extractable Fat of Presscake Stored at -20°C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in Air 53 6. Peroxide Values of Ether-Extractable Fat of Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20 C, Herring Meal Stored at 21°C and Herring Meal Stored at -20°C . . . . 54 7. Storage Changes in TBA Absorbance at 535 nm Calculated Per Gram of O i l in Presscake 1 '. Stored at -20 C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in A i r 55 8. - Storage Changes in TBA Absorbance at 535 nm Calculated Per Gram of O i l in Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20°C, Herring Meal Stored at 21°C and Herring Meal Stored at -20°C . . . 56 9. Storage Changes in Absorbance at 233 nm of L ip id Extract of Presscake : Stored at -20 C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in Air 57 x i i i Figure Page 10. Storage Changes in Absorbance at 233 nm of L ip id Extract of Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20°C, Herring Meal Stored at 21 C and Herring Meal Stored at -20°C 58 11. Available Lysine Changes During Storage. Presscake Stored at -20 C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in A i r . Available Lysine Values Expressed as a Percentage of the Total Protein . . . 59 12. Available Lysine Changes During Storage. Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20 C, Herring Meal Stored at 21 C and Herring Meal Stored at -20 C. Available Lysine Values Expressed as a Per-centage of the Total Protein 60 13. Pepsin Digest ib i l i ty Changes During Storage. Presscake Y Stored at -20 C, Freeze-Dried Meal Stored in Nitrogen and Freeze-Dried Meal Stored in A i r . Pepsin Digest ib i l i t y Expressed as a Percentage of the Total Protein 61 14. Pepsin Digest ib i l i t y Changes During Storage. Herring Meal Treated with Antioxidant and Stored at 21 C, Herring Meal Treated with Antioxidant and Stored at -20 C, Herring Meal Stored at 21°C, and Herring Meal Stored at -20°C. Pepsin Digest ib i l i ty Expressed as a Percentage of the Total Protein 62 15. Available Lysine Changes During Storage. Solvent-Extracted Meal Stored at 21°C and Solvent-Extracted Meal Stored at -20 C. Available Lysine Values Expressed as a Percentage of the Total Protein 63 16. Pepsin Digest ib i l i t y Changes During Storage. Solvent-Extracted Meal Stored at 21°C and Solvent-Extracted Meal Stored at -20 C. Pepsin Digest ib i l i ty Expressed as a Percentage of the Total Protein 64 17. Tracing of a Typical Peptide Map of Pepsin Digests of Herring Meal 65 18. Ether-Extractable Fat from Model System Stored at 21°C and from Model System Stored at -20°C 69 xiv Figure Page 19. Iodine Values of Herring O i l Stored at 21°C, Herring O i l Stored at -20°C, Ether-Extractable Fat of Model System Stored at 21°C and Ether-Extractable Fat of Model System Stored at -20°C 70 20. Peroxide Values of Herring O i l Stored at 21°C, Herring O i l Stored at -20°C, Ether-Extractable Fat of Model System Stored at 21°C and Ether-Extractable Fat of Model System Stored at -20°C 71 21. Storage Changes in TBA Absorbance at 535 nm Calculated Per Gram of O i l in Herring O i l Stored at 21°C, Herring O i l Stored at -20°C, Model System Stored at 21°C and Model System Stored at -20°C . . . . . . . . . . . 72 22 Storage Changes in Absorbance at 233 nm of Herring O i l Stored at 21°C, Herring O i l Stored at -20°C, L ip id Extract of Model System Stored at 21 C and Lipid Extract of Model System Stored at -20°C . . . . . . 73 23. Available Lysine Changes During Storage. Albumin Stored at 21°C, Albumin Stored at -20°C, Model System Stored at o o 21 C and Model System Stored at -20 C. Available Lysine Values Expressed as a Percentage of the Total Protein . . . 74 24. Pepsin Digest ib i l i t y Changes During Storage. Albumin Stored at 21°C, Albumin Stored at -20°C, Model System Stored at 21°C and Model System Stored at -20°C. Pepsin Digest ib i l i t y Expressed as a Percentage of the Total Protein 75 25. Tracing of the Peptide Map of a Pepsin Digest of Fresh Albumin 76 26. Tracing of the Peptide Map of Pepsin Digest of Albumin Stored at 21°C for 6 Months 77 27. Tracing of the Peptide Map of Pepsin Digest of Albumin Stored at -20°C for 6 Months . 78 28. Tracing of the Peptide Map of Pepsin Digest of the Model System Stored at 21°C for 6 Months 79 29. Tracing of the Peptide Map of Pepsin Digest of the Model System Stored at -20°C for 6 Months 80 30. Tracing of the Peptide Map of Pepsin Digest of the Albumin Stored at 21°C for 11 Months 81 XV Figure Page 31. Tracing of the Peptide Map of Pepsin Digest of the Albumin Stored at -20°C for 11 Months 82 32. Tracing of the Peptide Map of Pepsin Digest of the Model System Stored at 21°C for 11 Months 83 33. Tracing of the Peptide Map of Pepsin Digest of the Model System Stored at -20°C for 11 Months 84 34. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens 1% Hour After Ingesting Herring Meal Treated with Antioxidant and Stored at 21 C and Herring Meal Stored at 21°C 98 35. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens lh Hour After Ingesting Herring Meal Treated with Antioxidant and Stored at -20 C and Herring Meal Stored at -20°C 99 36. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens lh Hour After Ingegting Herring Meal Treated with Antioxidant and Stored at 21 C and Herring Meal Treated with Antioxidant and Stored at -20°C 100 37. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens lh Hour After Ingesting Herring Meal Stored at 21°C and Herring Meal Stored at -20 C 101 38. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens lh Hour After Ingesting Solvent-Extracted Meal Stored at 21 C and Solvent-Extracted Meal Stored at -20°C . . . . . . . . 102 39. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chicken Starved for 18 Hours . . . . 103 40. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens 1% Hour After Ingesting Herring Meal Treated with Antioxidant and Stored at 21 C and Herring Meal Stored at 21°C 104 41. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens lh Hour After Ingesting Herring Meal Treated with Antioxidant and Stored at -20 C and Herring Meal Stored at -20°C 105 xvi Figure Page 42. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens 1% Hour After Ingesting Herring Meal Treated with Antioxidant and Stored at 21 C and Herring Meal Treated with Antioxidant and Stored at -20°C 106 43. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens 1% Hour After Ingesting Herring Meal Stored at 21 C and Herring Meal Stored at -20°C 107 44. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chickens 1% Hour; After Ingesting Solvent-Extracted Meal Stored at 21 C and Solvent-Extracted Meal Stored at -20°C . 108 45. Fractionation by Sephadex G-25 of the Soluble Nitrogen of the Intestinal Contents of Chicken Starved for 18 Hours . . . . 109 Acknowledgments I wish to thank Professor B. E. March, Department of Poultry Science, The University of Br i t ish Columbia for her guidance, interest and constructive cr i t ic ism throughout this study. Sincere appreciation is also extended to Professor J . B ie ly , Department of Poultry Science for his enthusiasm and encouragement and to the other members of my thesis committee: Dr. W. D. K i t t s , Chairman, Department of Animal Science and Poultry Science, Professor E. L. Watson, Department of Agricultural Engineering, and Dr. W. J . Polglase, Professor of Biochemistry, a l l of The University of Br i t i sh Columbia. The f inancial assistance provided by The University of Br i t i sh Columbia and the National Research Council of Canada is gratefully acknowledged. 1. INTRODUCTION One of the primary problems in the storage of food is the oxida-tion of l i p ids . In general, autoxidation may be defined as the reaction of any material with molecular oxygen. A free radical chain mechanism has been proposed for the autoxidation of l i p i d s . Unsaturated fatty acids in their different forms and combinations are susceptible in varying degree to autoxidation. Other food consti -tuents, such as those responsible for natural aroma, flavour:arid colouring along with some nutr i t ional ly essential vitamins, are also among the oxidizable food substances. The development of oxidative processes leads to the formation of compounds such as peroxides, aldehydes, ketones, low molecular weight acids and epoxy compounds as well as free radicals in fat and fat -con-taining products. These compounds are highly reactive and can react with other molecules including protein. As a consequence, a variety of complex interactions may be expected to occur between proteins and the degradation products of autoxidizing l ip ids under appropriate reaction conditions. These reactions frequently lead to a reduction in the nutr i t ive quality of foods. One of the major problems of the food industry is the preserva-tion of the quality of food during storage for long periods. Refriger-ation i s commonly used to minimize the changes in food quality. More posit ive methods such as the retardation and inhibi t ion of oxidative 1 2 reactions of fats are much needed however. L ip id oxidation, l ike most free radical reactions, i s very sensitive to even the smallest concentration of compounds able to react with the free radicals and to end the chain reactions. Antioxidants are known to retard autoxidation of l i p i d through the reaction with either the or iginal free radicals or with ones formed in the early stages of oxidation to give intermediates which block the chain reactions, at least temporarily. This study was designed to investigate the interactions between fat and proteins during prolonged storage at 21°C and at -20°C. A simplif ied model system approach has been employed with albumin and herring o i l . The assumption was made that oxidation of o i l proceeds at a faster rate in the presence of protein, and on the other hand that protein alteration occurs in the presence of oxidizing l i p i d . The oxidation of o i l was followed by means of measuring the iodine value, peroxide value, thiobarbituric acid (TBA) absorbance and ult raviolet spectrum during storage. Protein alteration was measured by the change in the available lysine and pepsin d iges t ib i l i t y . Because the pepsin d igest ib i l i t y method does not y ie ld information concerning the nature of the peptides released, separation of the peptides by two-dimensional chromatography and electrophoresis was employed. Further studies were carried out to determine the extent to which similar changes may occur during the storage of an actual feeding product. Herring meal was used as the test material because i t i s con-sidered a source of high quality protein and also because i t contains 3 re lat ively large amounts of unsaturated fat . The role of antioxidants in delaying oxidation of the o i l in the stored herring meals was studied. Freeze-drying has become an important process for the preserva-tion of foods because of the high quality of the rehydrated product compared to those prepared by other dehydration techniques. Despite the success of the freeze-drying process i t s e l f undesirable changes take place in the product during i t s storage which l imit s h e l f - l i f e . L ipid oxidation is among the most important reactions which occur in freeze-dried products. Oxidation i s fac i l i ta ted because the sponge-like structure of the product makes the l i p i d more accessible to oxygen, and because of the low moisture content which is known to promote oxidation. Interactions between l i p i d oxidation products and proteins are known to take place during the processing and storage of dried food. In the present work a study was made of the rate of oxidative deterioration in freeze-dried herring meal during storage at 21°C in air and in nitrogen. Measurements of the interaction between oxidized fat and protein in herring meal were made using bio logical procedures which are less open to question than the usual chemical determinations. Feeding tests in the present work were designed to f ind out whether the nutr i t ional value of the protein content as a supplement for a cereal diet was affected during storage and to determine the avai lab i l i ty of energy content for chicks by direct measurement of metabolizable energy values. An in vivo study of protein digestion was made on the various herring meals after 11 months storage. The composition of the unabsorbed soluble nitrogen fraction of the contents of the small intestine of 4 chicken 1-1/2 hr. after being given the test meals were compared. The fractionation into soluble protein, peptides and free amino acids was accomplished using a calibrated column of Sephadex G-25. 2. REVIEW OF LITERATURE Most food products contain a certain amount of l i p i d . Lipids present in edible fishery products tend to be relat ively unsaturated. As a result of high unsaturation f i sh l ip ids undergo extensive and rapid oxidation, (Lundberg 1967, Roubal 1967). Since oxidative deterioration in fishery products is the main reason for these undesirable alterat ions, some information concerning l i p i d oxidation in general w i l l be reviewed. This review w i l l be followed by a discussion of the interaction between protein and oxidized l ip ids with particular emphasis on the model system approach. F ina l l y , the effects of the above mentioned reactions on the nutr i t ive value of herring meal w i l l be surveyed. A. LIPID OXIDATION Food l i p i d s , primarily tr iglycerides and phospholipids, are susceptible in various degrees to oxidation, which renders the l i p i d rancid. In addition to the oxidation of the l ip ids themselves, other food constituents such as pigments, vitamins, ^and flavour constituents can be oxidized. The appearance, palatabi l i ty and nutr i t ive value of food are usually damaged as a result of l i p i d oxidation. Oxidation is primarily associated with unsaturated l i p i d s . The rate of oxidation markedly increases with an increase in the degree of unsaturation of the fatty acid components, (Lundberg 1962). 5 6 Lipid oxidation may be defined as a reaction of unsaturated fatty acids with molecular oxygen to produce mainly hydroperoxides. The hydroperoxides formed can decompose to a variety of degradation products such as vo lat i le aldehydes, ketones, and acids. These degradation pro-ducts usually cause the undesirable off - f lavours and off-odours character-i s t i c of oxidized l i p i d s , (Lea 1958) . L ip id oxidation is frequently accompanied by browning reaction. This reaction is known to be associated with the polymerization of the secondary breakdown products. The over -a l l oxidation mechanism i s shown below, (Labuza, £t_ a l . , 1968). breakdown products off-odours e. g. , hexanal off - f lavours Unsaturated fatty acids catalysts free radicals °2 hydroperoxides polymerization (darkening) oxidation of pigments flavours, vitamins When the hydroperoxide theory of autoxidation was postulated (Farmer and Sundralingam 1942, Farmer et a l . , 1942, Farmer 1946), i t was proposed that the in i t ia t ion of the chain reaction occurred either by a direct abstraction of a hydrogen atom from a methylene group adjacent to a double bond or by a two-step reaction involving addition of oxygen to the double bond followed by reaction of the intermediate with hydrogen on 7 the adjacent carbon atom, (Bolland and Gee 1946, Farmer 1946). Both reactions appeared to be equally possible thermodynamically, (Bolland and Gee 1946). Uri (1958, 1956) questioned the probability of molecular oxygen to react direct ly with a fat molecule on the basis of the thermo-dynamics of the reaction. Later in 1961 Uri and Heaton proposed that the i n i t i a l attack by molecular oxygen required traces of heavy metals (M) as catalysts in a series of electron transfer reactions of the following type: hydroperoxides in the i n i t i a l stages of oxidation involved heavy metal catalysts. Autocatalysis refers to a reaction whose rate increases with time as a result of the formation of products which themselves catalyze the reaction, (Lundberg 1962). L ip id oxidation i s therefore considered to be an autocatalytic process. In the early stages of l i p i d oxidation, most of the oxygen is contained in hydroperoxides, which are known to possess catalyt ic act iv i ty . Two hydroperoxides are formed during the early stage of the autoxidation of methyl l inoleate, (Frankel 1962). M n + 0 2 M They also proposed that in i t ia t ion through the decomposition of C C H 1 1 - CH = CH - CH = CH - CH - C 0 H n . - C0OCH, 5 11 i 8 16 00H CH - CH = CH - CH = CH - C„H. ^ C00CH 3 00H 8 Lundberg (1962) indicated that a typical curve for the rate of hydroperoxide build-up in fatty materials is characterized by an induction period during which the rate of oxidation is slow and then by a period during which the rate of oxidation accelerates and during which hydro-peroxides develop. Lundberg (1962) formulated a mechanism for the autocatalytic autoxidation based on observations made by several investigators as follows: 1. In many respects the reaction appeared to be analogous to certain autoxidations in inorganic systems that were known to involve chain reactions. 2. By spectrophotometric methods, i t was observed that shifts of double bonds occurred during autoxidation. This, together with the observation that hydroperoxides were formed, suggested that a free radical mechanism might be involved. 3. The autoxidizing l inoleate system contained hydroperoxides which decomposed; i t was known that the decomposition of other organic peroxides in unsaturated systems in i t ia tes polymerization reaction chains. Bolland and Gee (1946) re-ported that they had investigated the l inoleate hydroperoxides and found the decomposition reaction to be bimolecular. Based on the foregoing kinet ic and some chemical observations, a free radical chain mechanism was proposed. It has been modified on the basis of subsequent work, in the generalized form shown below, (Lundberg 1962): In i t iat ion: free radicals + free radicals (e .g . , R*, RO", RO* HO", etc.) RH + 0 2 ROOH | (ROOH)2j Propagation: R ' + 0 r -*• R O ; RO* + R H Termination: R* + R* R" + ROOH R' + RO^ Stable (non-radical) end products This scheme showed that hydroperoxides are formed and decomposed during the autocatalytic oxidation. The rate of hydroperoxide formation exceeds the rate of decomposition during the early stages of oxidation. Also during these stages the hydroperoxide concentration affects the rate of oxidation. In the later stages of oxidation, however, hydroperoxide dismutation predominates and hydroperoxide level decreases. Keeney (1962) provided some insight into the mechanism of hydro-peroxide dismutation as follows: R - C H (00H) - R R* + RCHO Aldehydes ->• R - CH - R + OH I 0* R" RH R - CHOH - R + R" Alcohols R - C - R + R H 1 0 Ketones ROOH + - CH = C H - CH - CH + ROH 0 Epoxides 10 This scheme shows how some of the secondary products of l i p i d oxidation are formed as a result of hydroperoxide dismutation. Polymerization i s a major reaction in autoxidizing l i p i d s . It may result from direct association of alkoxy and alkyl free radicals or through the following reactions. (Keeney 1962): + - CH = CH + ROOH • - CH - CH - 0 - OR > CH-CHOH-• I I H 0 R It i s believed that aldol condensation and the formation of epoxy crosslinks contribute to the polymerization process. L ip id oxidation, l ike most free radical reactions, i s very sensi -t ive to catalysts and inhibitors . Heat and l ight are considered to be among the catalysts. Besides, various pro-oxidant substances are known to greatly accelerate the rate of l i p i d autoxidation even i f present in small concentration. Among those that are important in relation to food fats are various metals, metal salts and organic compounds of metals; oxidative enzymes such as the lipoxidases; other b io logical catalysts such as hemoglobin and other hematin compounds. In some cases these accelerating factors affect the rate of one or other of the reactions involved in the mechanisms summarized by Lundberg (1962) and mentioned before. In other cases, however, fundamental differences in the peroxi-dation mechanisms are involved. On the other hand, many substances, some in very small amount, exert an antioxidant effect . Phenolic substances, for example, are known 11 to delay oxidation. Antioxidants act by interrupting the chain reactions in the autoxidative mechanism, as was proposed by Backstrom (1927). After the development of a more refined concept of autoxidation process, Bolland and tenHave (1947) advanced a simple mechanism in which the antioxidants acted as hydrogen donors or free radical acceptors. From the kinetics of the reaction Lundberg (1962) suggested that the free radical acceptors react primarily with a hydroperoxy radical (ROp and not with a hydro-carbon radical (R*) as follows: RO* + XH2 • ROOH + XH* XH* + XH* > X + XH2 The observations of Cooper and Melv i l le (1951) confirmed these findings. Boozer et a l . (1955), however, proposed another mechanism which involves the formation of a complex as follows: RO' + XH2 w (R02XH2)* (R02XH2)* + RO* • stable products. Later in 1956 Bickel and Kooyman suggested the following consecutive re -actions : XH2 + RO* > XH* + ROOH XH* + RO* > stable products. There appears to be kinet ic argument against both of these lat ter mechanism. Harle and Thomas (1957) suggested that both may be operative, i f the kinet ic 12 observations do not satisfy either mechanism. Recently, Roubal (1971 a,b) explained the inhibit ing effect of antioxidant in l i p i d oxidation reactions as inhibi t ion of trapped free radicals . He suggested the following mech-anism for the interaction between radical (R*, R*) and cel lu lar an t i -oxidant (D), as follows: >-Hydroperoxide R* R" Decay • ' Non radicals + D* y slow decay where D is shown here as hydroquinone OH H + 0' Davies et a l . (1956) and Uri (1961) examined and discussed the role of structure and the relationship of antioxidant ef f ic iencies to oxidation reduction potentials. Moisture content is also a factor which affects the rate of l i p i d oxidation. The presence of small amounts of water markedly inhibits 13 rancidity , as was shown in several food systems stored under various moisture levels , (Marshall et a l . 1945, Martin 1958 and Matz et a l . 1955). The inhibitory effect of water has been studied by several i n -vestigators. Maloney et a l . (1966), Labuza et a l . (1966) and Karel et a l . (1967) studied the oxidation of methyl l inoleate in freeze-dried model systems. In some of these experiments metal catalysts were added. Water was found to have (1) an inhibitory effect on the oxidation reaction, varying with water act iv i ty up to values of 0 .5 , (2) water even below the monolayer coverage has an inhibitory effect on l i p i d oxidation, (3) water was found to have an inhibitory effect on the metal-catalyzed oxidation. This inhibi t ion of the reaction by water is explained as due to hydration of the trace metals, and (4) the changeover from hydroperoxide mono-molecular decomposition kinetics to bimolecular decomposition occurs at higher oxidation levels as the amount of water present increases. Many hypotheses have been advanced to explain the protective effect of water in retarding l i p i d oxidation. Maloney et a l . (1966) cited five hypothesis as the most important ones: 1. water has a protective effect due to retardation of oxygen diffusion (Halton and Fish 1937), 2. water lowers the effectiveness of metal catalysts such as copper and iron (Uri 1956), 3. water promotes non-enzymatic browning, which can result in the formation of antioxidant compounds (Lea 1958), 4. water forms hydrogen bonds with hydroperoxides and retards hydroperoxide decomposition, 14 5. water is attached to sites on the surface of the system, thereby excluding oxygen from these sites (Salwin 1959). More recently, the water sorption characteristics of food have been reviewed by Labuza (1968). He suggested that above the monolayer value of water adsorption, other reactions such as non-enzymatic browning may occur which lead to deterioration of foods. Labuza et^  a l . (1969) concluded that water inhibits oxidation by hydrogen-bonding of some of the hydroperoxides formed during the propaga-t ion step of the reaction. As a result these hydroperoxides are removed from further reaction. Later on when no more hydroperoxides can be removed from the reaction, the rate of oxidation increases rapidly. The overal l rate constant of the reaction i s reduced as a result of hydration of catalysts, during both periods of oxidation. Labuza et a l . proposed the following overal l scheme of water inhib i t ion: In i t iat ion: (ROOH) HO 4-M -> RO" + *0H R' Propagation: R' + 0, Ko + R00' R00' + RH — ^ 2 — y (ROOH). + (ROOH),, + R* A iJ Termination: 2 R00* Kt •*• non-radical products. Inhibit ion: (ROOH) B water - l ip id interface H H H - 00R 15 Heidelbaugh et a l . (1971) performed experiments to determine the effects of system composition on oxidation at various water ac t i v i t i es , expecial -ly those associated with intermediate-moisture foods. They found that in a l l systems studied, addition of water i s antioxidant up to a c r i t i c a l water act iv i ty above which further increases promote oxidation. Addition of cobalt and glycerol decreased the numerical value of the c r i t i c a l water content and of water act iv i ty . The rate of oxidation of methyl l inoleate in protein and cellulose systems was investigated by Labuza et a l . (1971). They found that at low water content, water hydrated metals and hydrogen bonds with peroxides, and an overall decrease in the rate of l i p i d oxidation resulted. An i n -crease in oxidation rate was found to occur at higher water content (water act iv i ty of 0.6 to 0.7) and was explained on the basis that water acted as a solvent to dissolve and mobilize previously unavailable trace metals. They also reported that the use of chelating agents such as ethylenediaminetetraacetic acid and c i t r i c acid reduced oxidation s i g n i f i -cantly. Dehydration of food gives i t a sponge-like structure. As a consequence diffusion of oxygen to the inter ior of the products is greatly enhanced. In addition, dehydrated food has a large contact area between l i p i d and non- l ip id components, (Lea 1958). An increase in both the rate of oxidation, and the damage caused to non- l ipid components as a result of l i p i d oxidation are promoted by these two characteristics of dehydrated food mentioned above. The changes in texture and rehydrability of de-hydrated food led Koch (1962) to the conclusion that interaction between 16 l i p i d oxidation products and other food components especially proteins, takes place during processing and storage of the dried food. B. INTERACTION BETWEEN PROTEIN AND OXIDIZED LIPID Proteins are known to form complexes with a wide variety of l i p i d materials such as phospholipids (Olcott and Mecham 1947) and sterols (Detorin et a l . 1953). Also proteins can interact with other substances such as alkyl benzene sulfonates (Lundgren at a l . 1943) and gossypol (Baliga and Lyman 1957). Most of these complexes, however, are lab i le and may easily give back the constituent molecules in their or iginal form by the use of simple procedures, such as extraction with a suitable solvent system. The linkage between the l i p i d and the protein in the case of l ipoproteins probably involves van der Waals forces and weak electrostat ic attractions (Chargaff; 1944;^Dervichian-1949, and Oncley 1954). These lipoproteins are lab i le and the l i p i d can be separated by suitable extraction methods. On the other hand, Tappel (1955) observed that unsaturated fatty acids and esters could react with proteins, form-ing rather stable complexes which could not be ruptured by solvent treatments. He attributed the s tab i l i t y of these complexes to a chemical union between the aldehydes produced during autoxidation of the l i p i d and the amino groups of the proteins. The model system approach has been used in several research studies concerning the interaction between l i p i d and protein. Some studies have been conducted to determine the alterations in protein fractions which follow the oxidation of prote in - l ip id model systems. Tappel (1955) studied formation of yellow-brown copolymers in aqueous emulsions 17 containing l ino le ic acid, cod- l iver o i l and proteins. He suggested that the mechanism of copolymerization involves active aldehyde browning con-current with the oxidation of unsaturated fat . He also reported that aldehydes originating from l i p i d oxidation react readily with amino acids to form aldimines which would subsequently lead to the formation of brown nitrogenous polymers and copolymers, i . e . , melanoidins. The formation of brown polymers in highly reactive aqueous emulsion systems containing menhaden o i l and aqueous egg albumin at pH 7 was studied by Venolia and Tappel (1958). They concluded that active carbonyl-amine browning was not a dominant mechanism and suggested that oxypolymerization of unsaturated l i p i d led to the production of intensely colored polymers. The oxidation of a relat ively simple model system such as oxidized l i no le i c acid and egg albumin was studied by Narayan and Kummerow (1958). The results indicated absence of covalent bonding between the oxidized l i p i d and the protein. It was further suggested that hydrogen bonds were apparently responsible for complex formation. An attempt was made by Narayan et a l . (1964) to study the nature of the linkage between the oxidized l i p i d and the protein. The data obtained confirmed previous findings and further emphasized the role of keto and epoxy groups in complex formation while the hydroxy and hydroperoxide groups were observed to be less reactive. Some proteins were preferential ly denatured by oxidized l i p i d than others as reported by Nishida and Kummerow (1960) and Narayan and Kummerow (1963). Complex formation between egg albumin or lactalbumin and oxidized corn o i l or autoxidized l ino le ic acid was observed by Narayan and 18 Kummerow (1963). No complex could be obtained, however, under similar conditions when Bacto peptone, gelat in, sodium caseinate, lysine or glycine were substituted for albumins. The mechanism of damage to protein by peroxidation of l ip ids has been studied by Desai and Tappel (1963). They measured the addition reaction of peroxidizing l ino lenic acid with cytochrome C in an aqueous emulsion and found a decrease in the so lubi l i ty of the protein and in the content of a l l of the amino acids. They concluded that oxidizing l ip ids act as cross- l inking agents in the polymerization and formation of insoluble masses. Roubal and Tappel (1966) modified this hypothesis after further studies in v i t ro , which have shown that in solution, pro-tein polymerization leading to insolubi l i ty is most l ike ly a free radical-induced process not actively incorporating l i p i d . Using a model system approach, Andrews e_t a l . (1965) reported that two puri f ied proteins (gelatin :and insulin) are•chemically modified in the presence of an autoxidized l i p i d , methyl l inoleate. They also studied the interaction between insul in and methyl l inoleate. They reported that l i p i d oxidation products reacted with the e-amino group of lysine and with the a-amino group of phenylalanine and glycine, the N-terminal amino groups of insu l in . They concluded, on the basis of trypsin hydrolysis and hydrogen fluoride so lubi l i ty tests, that the autoxidation products of methyl l inoleate interact with protein through a cross- l inking reaction. Roubal and Tappel (1966a) examined an emulsion containing a pro-tein and mixture of ethyl esters prepared from menhaden o i l . They 19 reported that f ree- radical intermediates of l i p i d oxidation reacted with proteins forming a complex of the cross-l inked rather than associate type. This indicates that the bonding in the complex was covalent. They also proposed a protein-protein free radical polymerization ,as:the d'ominarit-uimechanism. Roubal and Tappel (1966b) found transient free radicals in per-oxidizing l ip id -prote in reaction systems. The pattern of radical-induced damage to proteins, was similar to that observed in the case of radia-t ion damage; proteins lose so lubi l i ty and constituent amino acids are modified. Evidence for an interaction between proteins and malonaldehyde was presented by Kwon et a l . (1965). Malonaldehyde was investigated be-cause of i t s well known presence in autoxidized unsaturated fatty acids. Buttkus (1967) studied the reaction of myosin with malonaldehyde. Amino acid analysis before and after the reaction (at 100° for 60 sec.) showed that malonaldehyde reacted preferential ly with h is t id ine , arginine, tyrosine and methionine. Crawford et a l . (1967) reacted gelat in . in buffered solutions at pH 4-8 , with malonaldehyde and formaldehyde. Using viscosity measurements, they noted that malonaldehyde did not participate in an intermolecular cross- l inking reaction with gelatin solut ion, where-as formaldehyde did cause an increase in v iscocity . e-Amino lysine groups were shown to be involved in both reactions. Z i r l i n and Karel (1969) studied the oxidation of a freeze-dried model system consisting of methyl l inoleate and gelat in. They found that incubation of the system for 5-10 days at 50°C caused a drop in the 20 viscosity of gelatin in ethanol-rich solvent mixtures, and an increase in the retention time of gelatin on a Sephadex G—150 column, and a re -duction in the melting point of a standard gelatin gel . They also reported that, in some cases, incubation at high relat ive humidity lead to par t ia l insolubi l izat ion of gelatin in water or in acetate buffer. Several investigators (Roubal and Tappel 1966a, Roubal and Tappel 1966b, Roubal and Tappel 1967 and Roubal 1969) suggested that free radicals derived from l i p i d oxidation are the dominant cause of the observed damage to proteins, i . e . , destruction of their amino acids. No definite electron paramagnetic resonance study, however, has been con-ducted to explain such interactions as a result of the low steady state concentration of free radicals . Working with human serum, Harmon and Piette (1966) reported faint electron paramagnetic resonance signals above background noise. These investigators, however, were not able to assign these signals spec i f ica l ly to l i p i d oxidation. Roubal (1969) reported evidence for the formation of radicals in oxidizing unsaturated l ip id -prote in systems, spec i f ica l ly in low moisture systems. Later in 1970 Roubal monitored the free radical concentration in l ip id -prote in mixtures with a low moisture content both during and after the time that the l i p i d actively absorbed oxygen. He indicated that in a dry system the decay in radical content is followed by a r ise in malonaldehyde protein fluorescence. Using the amino acid content at two different periods in the reaction, he concluded that radicals and not aldehyde are the major cause of protein damage. 21 Some studies seem to indicate that protein can react with unoxidized l i p i d s , such as l i n o l e i c , l ino len ic , and oleic acids. Ander-son et a l . (1963, 1964) found that the addition of fatty acid to cod muscle caused reduction in protein extractabi l i ty . Bul l and Breeze (1967a, 1967b) investigated the reaction between fatty acids such as acet ic , propionic, n-butyric, i -butyr ic and n-hep-tanoic acids and egg albumin in aqueous solutions. They found that the fatty acids were bound to the protein and denatured i t . This denatura-tion was accompanied by a signif icant increase in the viscosity of the protein solution. The above mentioned interactions between oxidized l ip ids and pro-teins have undesirable effects on the nutr i t ional value of food in general, and in part icular on the bio logical value of proteins. C. HERRING MEAL Herring meal is valued as a source of high-quality protein. It also contains a considerable quantity of highly unsaturated fat . The meal has important chemical and nutr i t ional characteristics as a result of the react iv i ty of the unsaturated fatty components in the presscake. During processing and storage the polyunsaturated fatty acids are sub-jected to oxidation, the rate and degree of which are affected by the actual amount of polyunsaturated fatty acids. If the oxidation proceeds suff ic ient ly rapidly, the meal may undergo spontaneous heating. Spon-taneous heating leads to a destruction of amino acids, being more pro-nounced for lys ine, tryptophane, cystine, and hist idine (Boge 1960). For 22 most other amino acids investigated, the values are approximately 10% lower than for fresh meal. These chemical findings are supported by feeding tests. The deterioration of the nutr i t ive value of f i sh meal caused by spontaneous heating has been demonstrated by Lea et_ a l . (1960) and Laksevela (1958). Laksevela (1958) reported that the spon-taneous heating of herring meal induced a change in the pattern of available amino acids to such a degree that the meal was unable to sustain growth in chickens. Chemically available lysine was reduced by 40%, but since supplementation with lysine could not compensate for the ensuing reduction of the growth rate of chicken, i t appeared that other factors must consequently be involved. It i s well known that the protein of foods can suffer loss of nutr i t ive value as a result of over heating during processing. When the f ish products formed the only source of protein in a purif ied d iet , the nutr i t ive value was generally diminished with drying procedures which have been previously considered as relat ively harmless (Bender at a l . 1953 and Mi l le r 1956). Carpenter et a l . (1957) have reported on the effects of heating under various conditions on the nutr i t ive value and available lysine of the protein of f ish meal. They tested a number of dehydrated f ish products as supplementary proteins to cereals by the standard gross protein value technique with chicks. Their result indicated that commercial f i sh meal has a lower nutr it ive value than other prepara-tions made by a low-temperature drying process. The nutr i t ive value of vacuum dried cod f i l l e t s , packed in nitrogen at 11% moisture and held at 105°C for 36 hours showed a decrease of approximately 28%. There was a 23 close agreement between the amount of available lysine and the bio logical values. Controlled heating of herring meal dried at low temperature does not lower the nutr i t ive value of the meals. In many cases, chicks grew better with the heated meals, (March et a l . 1961). Tarr et a l . (1951) and Bissett and Tarr (1954) reported no effect on ava i lab i l i t y of essential amino acids in herring meals produced by normal heating methods. Ex-cessive heating, whether through long duration or high temperature, seriously impairs the quality of the product. Bender e_t a l . (1953) recommended that the drying temperature should not exceed 212°F. Studies by Thurston et a l . (1960) on herring and tuna meals showed an effect on pepsin d igest ib i l i t y at higher drying temperatures. The differences were minor below 230°F but became evident at higher temperatures. Rao et a l . (1965) found that the mode of drying of f i sh meals had l i t t l e or no effect on the nutr i t ive value of meal as measured by the chemical indices of available lysine and pepsin d iges t ib i l i t y . The effect of autoclaving on the d igest ib i l i t y of f ish protein concentrate was studied by Knipfel et a l . (1970). They reported a reduction in the d igest ib i l i t y of the f ish protein concentrate as a result of autoclaving. Generally, there are conf l ict ing findings with respect to the effect of storage on the nutr i t ive quality of f i sh meal. Biely e_t a l . (1951) found that the nutr i t ive value of herring meals stored for one year at - 2 5 ° , 21° and 37°C was unaffected when the meals were tested as supplements to a diet well fo r t i f i ed with vitamins. Mi l le r (1955) showed that the value of a dry cod meal as a protein supplement was unchanged during 3 months storage at room temperature. Stansby (1948) on the other 24 hand observed that the apparent fat content of f ish meal, as determined by ether extraction, decreases even during a short storage period. Almquist (1956) reported that in sardine meal the extractabi l i ty of fat and d igest ib i l i t y of protein decreased during storage and concluded that the changes noted were due to oxidation since they were not observed in samples which were stored in sealed glass ampoules. Lea et a l . (1958) prepared herring meals with moisture contents of 11 and 6.2% by hot -a i r drying of the presscake and stored them in ai r and in nitrogen. Oxida-tion was more extensive at the lower moisture content. He also found that available lysine was decreased by 9% after 12 months storage. No decrease occurred in the available lysine content in the whole meals stored in nitrogen or in defatted meals stored in a i r . The interaction of protein with oxidizing fat was studied by Tappel (1955). Insoluble dark-brown copolymers of high oxygen and n i t ro -gen content were formed. Tappel concluded that a considerable destruction of amino acids occurred, since the total amino acids recoverable from the copolymer, even after acid hydrolysis, was about 16% lower than would be expected from their nitrogen content. Lea et a l . (1960) presented evidence of a reaction between fat oxidation products and protein. They found higher bound l i p i d and thiobarbituric-acid values given by oxidized as compared with fresh herring meal. Waissbluth et a l . (1970) found that increasing moisture content had a marked pro-oxidant effect in a l l the oxidative reactions in fishmeal. The effect of storage temperature on the chemical and nutr i t ive properties of herring meal was studied by March et a l . (1961). They found 25 that a more rapid decrease in ether extractabil i ty and in the iodine value of the ether extract occurred at low temperature. They also noted an increase in the HCl-acetone extract during the f i r s t six weeks of storage. This represented the formation of l i p i d complexes with protein. In samples of herring meals stored for 7 to 12 years at room temperature there was marked loss in nutr i t ive value compared with their or iginal evaluation. Infer ior i ty of the meals when used in diets as the only source of protein was accompanied by decreased content of available lys ine, (Laksevela and Aga 1965). Laksevela and Aga also reported that determination of free fatty acids and nitrogen solubi l ized by pepsin gave no useful information. Keeping quality could not be related to raw material or processing and there was no long term effect from the addition of butylated hydroxytoluene as antioxidant. Laksevela (1961), March et a l . (1961) and Lea (1963) reported that the l ip ids extractable from stored herring meal did not depress the rate of weight gain when fed to chicks. The solvent extraction of commercial herring meals had l i t t l e or no effect on their value in chick d iet , (Biely et a l . 1955, March et a l . 1957). Bender and Haizelden (1957) found no consistent effect of defatting a variety of f ish meals on their b io -logical values for rats . In a preliminary experiment, Lea et a l . (1963) found that chloro-form-methanol extracted herring meal promoted poor growth. They reported that further washing of the extracted meal with diethyl ether largely restored i ts value for chicks and concluded that the effect was probably due to the toxicity of residual chloroform. Geisler and Contreras 26 (1967) working on defatted samples of anchovy meals with an increas-ing level of a typical anchovy o i l added to them, reported that the d igest ib i l i t y of the meal decreased in proportion to the amount of fat added. The greatest losses in d igest ib i l i t y of the meal occurred during the f i r s t 15 days while after the f i r s t 30 days the losses were ins ign i -f icant . The peroxide value was also at i t s maximum during the f i r s t 15 days and the percentage of available lysine determined after 5 months of aging diminished as the percentage of added fat increased. In the mean-time the percentage of total lysine showed no change over the five months aging. The material which was not pepsin-digestible contained a s i g n i f i -cant percentage of the total lys ine. Approximately 50% of this lysine appeared to be available. L ip id hydrolysis is known to occur during the cold storage of f i sh (Olley et a l . 1960 and Lovern et a l . 1962). Several investigators (Dyer and Morton 1956, Dyer and Fraser 1959, Olley and Lovern 1960, Olley et a l . 1962 and Olley and Duncan 1965) recognized that the develop-ment of free fatty acids in stored frozen f ish occurs concurrently and possibly prior to protein denaturation. Anderson et a l . (1965) have reported that protein-free fatty acid binding is the driving force in l i p i d hydrolysis in frozen f ish muscle. Olley et: a l . (1962) indicated that the act iv i ty of phospholipase and lipases appeared to be important in producing free fatty acids in Gadoids and related species, during cold storage at - 7 ° and -14°C. On the other hand, the phospholipase act iv i ty seemed to be negligible in some Elasmdbranchs species. Work by Castel l et a l . (1966) showed that 27 the production of free fatty acids may actually inhibit metal-induced oxidation in frozen cod muscle, as indicated by thiobarbituric acid values and odors. Bligh and Scott (1966) reported that the free fatty acid content in cod muscle increased during frozen storage up to nine months. They related these findings to the hydrolysis of phosphatidyl ethanolamine and phosphatidyl choline. From disc electrophoresis of muscle protein of yellow croaker (Pseudosciaena manchurica) and Paci f ic saury (Cololabis saira) f i s h , Soon at a l . (1970) concluded that the protein changed less during storage under refr igeration than when salted or frozen. Storage of f i l l e t s of herring at -15°C showed that l i p i d hydrolysis, as measured from the forma-tion of free fatty acid, occurs simultaneously in dark and white muscle and that i t pract ical ly ceases after 9 weeks of storage, (Bosund and Ganrot 1969). Storing f ish meal at -20°C produced no benef ic ia l ef fect . Losses in growth response to meals stored at -20°C were almost as great as were those to meals stored at room temperature, (Miller and Ambrose 1970). Because antioxidants are widely used to prevent autoxidation of feeds and feed ingredients, they have been the subject of intensive study as a possible means of retarding the oxidative deterioration of f ish meal. Meade (1956), Aure (1957) and Flanzy et a l . (1962) found that treat-ment of f ish meal with antioxidant retards oxidative changes. Lea et a l . (1958) measured the effect on the rate of oxidation of o i l content in herring meal of butylated hydroxytoluene at a concentration of 0.005% in 28 combination with 0.005% c i t r i c acid and found that oxidation was consider-ably retarded. March et a l . (1961) found that butylated hydroxytoluene protected the l ip ids in herring meal against oxidation. Nevertheless the nutr i t ive values of protein in the treated and untreated meals were similar after storage for 9 months. Further studies with herring meal by March et^  a l . (1962) showed that treatment with antioxidant protects against loss of nutr i t ive value of l ip ids during storage. Ut i l i zat ion of the l i p i d fraction of freshly prepared meal was decreased after storing the meal for 11 months. Butylated hydroxytoluene treatment of herring meal protected the fat content against oxidative changes and improved i t s u t i l i za t ion . The u t i l i za t ion of fat in butylated hydroxytoluene-treated meal was s l ight ly less than in herring o i l , after storage for 11 months. Treatment of whole and presscake herring meals with antioxidants increases their available metabolizable energy and nutr i t ive value as protein supplement in chick rations. It also increases the ease of extractabi l i ty and metabolizable energy of the l i p i d fractions of the meal (March ^t a l . 1965). The effect of butylated hydroxytoluene on the protein value of herring meal for young rats was studied by Njaa et a l . (1966). Nitrogen balances were made on batches of meal made from summer and winter herring in which the antioxidant, butylated hydroxytoluene, was added at the 0.03% level before or after drying of the product. They found that butylated hydroxytoluene had l i t t l e effect on protein quality in the meals produced from winter herring, but when added after drying, i t improved the quality of the meals produced from summer herring. 29 When freshly produced f ish meal was treated with 400, 750 or 1000 ppm butylated hydroxytoluene, no evidence of spontaneous heating was reported. The metabolizable energy values of the treated samples were higher than the untreated meal, (Romoser et a l . 1969). With Norwegian herring meal, Opstvedt et a l . (1970) found that the addition of l ,2-dihydro-6-ethoxy-2,2,4-trimethylquinoline (ethoxyquin) improved the protein quality as measured by protein eff iciency rat io and net pro-tein u t i l i za t ion . Slight increases in the metabolizable energy and in the growth rates of broi ler chickens fed the meal at various levels have also been reported. 3. MATERIALS AND METHODS A. PREPARATION OF THE MEALS The experimental herring meals were prepared from freshly-caught f i s h , in a p i lo t - sca le reduction plant at the Vancouver Laboratory of the Fisheries Research Board of Canada (Claggett, 1968), in November 1970. An o i l - f r e e meal was manufactured from herring presscake. Isopropyl alcohol was used as the solvent. The presscake was extracted three times with the boi l ing solvent and then decanted through a f i l t e r . The f ina l cake was dried and deodorized with steam at 100 psig. It was then re-dr ied. This meal i s subsequently referred to as solvent-extracted meal. After the meals were received from the p i lot plant the following treatments were applied: 1. Presscake, undried, used as control and kept at -20°C. 2. Presscake, freeze-dried, vacuum released with ai r and the meal stored under a i r at 21°C. 3. Presscake, freeze-dried, vacuum released with nitrogen and the meal stored under nitrogen at 21°C. 4. Heat-dried herring meal, stored at room temperature (21°C). 5. Heat-dried herring meal, stored at -20°C. 6. Heat-dried herring meal, treated with 0.025% ethoxyquin, stored at room temperature (21°C). 6-Ethoxy-l,2 dihydro-2,2,4-trimethylquinoline, courtesy Monsanto Canada Ltd. 30 31 7. Heat-dried herring meal treated with .025% ethoxyquin, stored at -20°C. 8. Solvent extracted meal stored at room temperature (21°C) . 9. Solvent extracted meal stored at -20°C. The meals were stored in polyethylene-lined multiwall paper bags and were sampled after 2, 8, 16, 24, and 36 weeks for the chemical analyses. B. PREPARATION OF THE MODEL SYSTEM A model system was prepared in the middle of December 1970 as follows: egg albumin (Nutritional Biochemicals Corporation, Cleveland, Ohio) was mixed with herring o i l (freshly extracted from herring meal) in the rat io of 12% o i l and 88% egg albumin. The resultant mixture was divided into two portions. One portion was stored at room temperature (21°C) , while the other was stored at -20°C. Similar ly , herring o i l and egg albumin were stored separately at these two temperatures. The chemical tests were run on the model system and i t s constituents immediately after preparation and after storage for 12, 20, 32 and 40 weeks at the two temperatures. C. CHEMICAL ANALYSES Standard methods (A.O.A.C., 1960) were used for the determination of crude protein, moisture, and ether-extractable fat . The following tests were carried out on the ether extracts: 1. Iodine number was determined by Hanus' method (A.O.A.C., 1960). 2. Peroxide value was determined by Lea's method. (Lea 1931). 3. The ultraviolet absorption spectra were determined with a 32 Unicam Model SP. 800 ultraviolet spectrophotometer, using methanol as solvent (Recknagel and Ghoshal 1966). To determine the thiobarbituric acid value (TBA) in the whole meal, the procedure of Yu and Sinnhuber (1957) was used. Chemically available lysine was determined by the method described by Carpenter (1960) and the values obtained were calculated as a percentage of the crude protein present in the meals. Protein d iges t ib i l i t y , in the fa t -free samples, was measured using pepsin according to the method of the A .O .A .C . , (1960), with the modification that the strength of pepsin solution was decreased to 0.01%. The pepsin used in the present test had an act iv i ty of l/-20,000. Peptide maps were prepared using two-dimentional chromatography and electrophoresis (Katz et a l . 1959) as follows: the ether extracted meals (1 gm) were digested by pepsin (.01% (1/20,'000) in 0.075 N-HCL) for 16 hours at 42-45°C. The enzyme action was stopped by boi l ing for 2 minutes and the samples f i l te red to remove the undissolved substances. An amount of .25 y l of the peptic digest were applied to 18^ x 22% inch sheets of Whatman No. 3 paper. The origin spot was approximately 2 inches below the glass rod over which the paper was hung. Chromatograms were run in the descending direct ion. The solvent used was n-butanol-acetic acid-water (4:1:5). It was used within 10 hours after preparation. Chromatograms were run for 18 hours at room temperature. The paper was then clipped to i t s supporting rod and dried in an oven at 37°C. High voltage electrophoresis, (H.V. Electrophorator Model D (Gibson Medical Electronics)) , was then performed at right angles to the direction of 33 chromatography, in pyr id ine-g lacia l acetic acid-water (1:10:489) buffer. The buffer was adjusted to pH 3.7 using NaOH or HC1. The chromatogram was prepared for electrophoresis by moistening the entire paper with the buffer (pH 3.7) . The paper was placed on a glass plate and a glass rod was put under the origin l ine to prevent puddling. The buffer was pipetted onto the paper. The origin l ine was moistened by the movement of buffer in from both sides and the chromatogram put over the Lucite rack and placed in the tank so that the top and bottom edges were completely immersed. Electrophoresis was run for 1.5 hours at 2500 volts and a current between 75-100 milliamperes. After completion of the electro-phoresis, a glass rod was passed under the top of the paper, which was then clipped to the rod with stainless steel c l i p s , carefully removed and dried in a i r . The peptide maps were developed by dipping the dried chromato-grams in 0.3% ninhydrin-acetone solution and heating at 80°C for 10 min-utes start ing with the oven at room temperature. D. BIOLOGICAL EVALUATION OF THE HERRING MEALS (i) Metabolizable energy Six of the herring meals (Table 1) were analyzed biological ly for their metabolizable energy values after four and nine months of storage. Day-old White Leghorn cockerel chicks were used as experimental subjects. The chicks were reared in e lec t r i ca l l y heated, wire-f loored battery brooders. They were given water and the basal diet ad l ibitum during a four-week pre-experimental period. Composition of the diet is shown in Table 2. The average nitrogen content and metabolizable energy 34 value for the basal diet were 4.16% and 3195 cal/gm dry matter, respec-t ive ly . Three repl icate groups were assigned to each of the treatments. The number of birds per replicate was two. Experimental diets were formulated by adding, on a dry matter basis , the test ingredients at the expense of a portion of the basal diet . The degree of substitution was 25% of the diet . Feed withdrawal was measured and excreta were collected over the following 3-day period. The excreta samples were freeze-dried and ground. The diets and the dried excreta were assayed for gross energy using a Parr oxygen bomb calorimeter. Nitrogen was determined in the diets and in the excreta by the Kjeldahl method. In addition the feed samples were analyzed for their dry matter content. From these results , the metabolizable energy values of the diets on a dry matter basis and corrected for nitrogen retention, were calcu-lated as follows (Hi l l and Anderson 1958): where and Kcal/gm dry diet = X - Y + (8.22 Kcal) Z , X = Gross Kcal/gm dry diet , Y _ (Kcal/gm excreta) (gm excreta) gm dry diet consumed 8.22 Kcal = Caloric value / gm of uric acid nitrogen Z = gm nitrogen retained / gm dry diet consumed , j , (gm nitrogen/gm excreta) (gm excreta) = gm nitrogen/gm diet — ^ s — — — , . ' -gm dry diet 35 To determine the metabolizable energy of the test ingredients on a dry matter bas is , the following formula was used: Kcal/gm test ingredient = Kcal/gm basal diet + Kcal/gm test diet - Kcal/gm basal diet grams test ingredient/gm test diet ( i i ) Protein quality: Experiments were conducted after f ive and ten months of storage to determine the supplementary protein value of the six herring meals indicated in Table 1. The basal diet employed was that used by March et a l . (1965) for assessing fishmeal quality in a way which would give a meaningful i n d i -cation of how effect ively f ish meal protein supplements the cereal protein. The chicks used in the two experiments were one-day-old White Leghorn cockerels. The chicks were kept in e lectr ical ly -heated, wire-floored battery brooders. They were fed a commercial chick starter for three days before being weighed and assigned to groups. The individual chickens were wing-banded, weighed and randomly distributed into groups of 12 chickens each. The experimental diets were composed of a basal mixture to which variable amounts of herring meal, starch, bonemeal and limestone were added to give approximately 3256' cal metabolizable energy per gm, ,1.09%. calcium, . 0C8%%phbsphbrus and ."different Lamounts of herring meals to supply 4%, 7%, and 11% of the protein to the diet . The various compositions of these diets are presented in deta i l in Tables 3(A)-3(L). 36 A diet with no added herring meal but which supplied similar levels of metabolizable energy, calcium and phosphorus was used as a control (Table 3(A)). It should be mentioned, however, that in the second protein quality experiment two kinds of wheat were used in formulating the con-t ro l diet. The metabolizable energy content of the diets was estimated using the values in Tables 4 and 5. The experimental diets were fed ad libitum for 17 days to four replicate groups. The chicks were weighed after 10 and 17 days and the feed intakes were measured. The results were expressed as gain in body weight and as diet eff iciency (gain in body weight / feed consumed). In order to compare the two experiments, (after 5 months and after 10 months of storage), the supplementary protein values were calculated on the basis of the growth response obtained with the diet containing herring meal relat ive to that obtained with the control d iet , and on the basis of the relative eff iciency of diet u t i l i za t ion . Av. gain in wt. of chicks fed diet with S.P.V. _ herring meal ^ gain in weight Av. gain in wt. of chicks fed control diet Body wt. gain/feed consumed (chicks fed S.P.V. _ test herring meal) ^ diet eff iciency Body wt. gain/feed consumed (chicks fed control diet) ( i i i ) D igest ib i l i ty test (in vivo) : The test materials were 11 months-stored herring meals (meals A,B,C,D,E and F, Table 1). D igest ib i l i ty of the different meals was tested 37 by investigating the difference in the Sephadex gel fractionation patterns of the soluble nitrogen content of the intest inal lumen of chicks fed the test proteins. Three-month old White Leghorn cockerels were used as the experimental animal. Two grams of f inely ground test material were administered in pel let form to individual birds after 18 hour fasting period. One ml. of warm water was pipetted into the crops following the administration of the pe l le ts . One and one-half hours after administra-tion of the test meal, the chickens were k i l l e d . The abdominal cavity was opened and the intest inal tract quickly but gently exposed. The contents of the small intestine were washed out with about 100 ml of 0.9% (w/v) NaCl solution into a small beaker and rapidly cooled in i c e -water. The suspension was centrifuged for 20 min. at 1250 G at 10°C. The supernatant f lu id was decanted and the residue resuspended in 10 ml d i s t i l l e d water. After recentrifugation the supernatant was combined with the or ig inal supernatant l iquor. The washed residue was then trans-ferred to a dish for the determination of dry matter and, subsequently, of "insoluble nitrogen" by the Coleman Model 29 A nitrogen analyzer (Dumas method). The supernatant f luids from the intest inal contents were made up to 150 ml and 50 ml aliquots were freeze-dried. Samples of 0.1 gm from the freeze-dried products were redissolved in 20 ml of 0.02M phos-phate buffer at pH 7.6. Soluble nitrogen content was also determined on the freeze-dried soluble intest inal contents. Fractionation on Sephadex Gel: Sephadex (cross linked dextran) gel f i l t r a t i o n medium was used, of f ine bead form, type G-25, part ic le size 20-80 urn (Pharmacia, Uppsala, Sweden). The column was prepared as 38 described by Andrews (1964). The column of height 60 cm and internal diameter at 2 cm, had a void volume of 200 ml. The column was packed in a ver t ica l glass tube across which was fused glass wool to support the gel . A layer of sand (2 cm) was used to prevent blockage of the glass wool by gel par t ic les . Dry gel (100 g.) was suspended in about 1200 ml. 0.02 M-sodium phosphate solution of pH 7.6 and allowed to stand for 24 hours with occasional s t i r r ing . The column was prepared by pour-ing the thin slurry of gel part ic les in buffer solution into the tube, already partly f i l l e d with buffer and at the same time allowing excess l iqu id to percolate through the growing gel bed. The addition of gel was continued unt i l a bed height of 60 cm was obtained and then a solvent reservoir was connected to the top of the column and the flow of buffer maintained for 2 days. The column was calibrated by using markers of known molecular weight to determine approximately the molecular weight of the digestion products. The following compounds were employed: cytochrome C, molecular weight 12400, glucagon, molecular weight 3500, bacitracin A, molecular weight 1470 and a mixture of 18 amino acids. The compounds were dissolved in the buffer solution at a concentration of 1 mg/ml and applied to the column in amounts of 5 ml. Cytochrome C was determined in the effluent fractions by measuring the extinction at 408 nm. Bacitracin A and glucagon were estimated from extinction measurement at 215 nm. Amino acids were determined by addition of ninhydrin reagent (Moore and Stein 1954) to the effluent fractions. 39 The procedure for column runs was as follows, (Ford 1965): an amount of 5 ml of the test solution (0.1 gm/20 ml buffer) was applied to the top of the column and allowed to enter the gel bed. Buffer (5 ml) was then applied and also allowed to enter the gel bed. A further 10 ml of buffer was applied and the column was connected to the buffer reser-voir . The buffer was then allowed to elute the gel and the effluent was collected in th i r ty , 9 ml fract ions, with a fraction collector (LKB 7000 Ultro Rac). One ml of each fraction was heated in a steam auto-clave with 2 ml of 6N-HC1 for 3 hours at 120°. To each hydrolysate was added 2.4 ml 4 N-NaOH, 1 ml of 4 N-sodium acetate buffer solution at pH 5.5, and water to a total volume of 10 ml. A 1-ml portion was taken for the estimation of a-amino nitrogen, by reaction with ninhydrin re -agent. The ninhydrin reagent contained 1 gm ninhydrin and 0.15 gm hydrindantin dissolved in 34.5 ml methyl cellosolve (ethylene glycol monomethyl ether) and 12.5 ml acetate buffer at pH 5.5. To the 1-ml portion, 1-ml ninhydrin reagent was added, mixed well and then heated in boi l ing water for 15 minutes. The tube contents were diluted with 8 ml 50% ethanol. The optical density was read at 570 nm using a Beckman spectrophotometer. A reference curve was prepared with known graded concentrations of leucine, and the amount of a-amino nitrogen present in each fraction calculated in terms of leucine-equivalent by reference to this curve. 40 TABLE 1 HERRING MEALS EMPLOYED IN THE BIOLOGICAL TESTS Meal Type Antioxidant Storage Temperature c A Unextracted 0.025% ethoxyquin 21 B Unextracted none 21 C Unextracted 0.025% ethoxyquin -20 D Unextracted none -20 E Solvent-extracted none 21 F Solvent-extracted none -20 41 TABLE 2 COMPOSITION OF BASAL DIET Component %_ Ground wheat 29.2 Ground yellow corn 37.8 Soybean meal (48% protein) 15.0 Herring meal (72% protein) 10.0 Dried d i s t i l l e r s ' solubles 3.0 Dehydrated cereal grass 2.0 Bonemeal 1.5 Limestone 1.0 Iodized salt 0.5 /kg Manganese sulphate 240.00 mg Vitamin A 4400.00 I.U. Vitamin D 3 440.00 I.C.U. Riboflavin 3.60 mg Zinc bacitracin 9.68 mg Amprolium 124.9 mg 42 TABLE 3(A)* COMPOSITION OF CONTROL DIET AND OF DIETS CONTAINING FISH MEAL A STORED FOR FIVE MONTHS AT 21°C Ingredients Control Diet Levels of Protein Supplied by Fish Meal A  0% Per 100 lb . Ground wheat 65.0 Vegetable o i l 3.0 Corn starch 20.0 Fish meal A Micronutrients 0.5 Iodized salt 0.5 Bonemeal 4.5 Limestone -Ground cellulose 6.5 4% 7% 11% Per 100 lb . Per 100 lb . Per 100 lb . 65.0 65.0 65.0 3.0 3.0 3.0 15.7 12.3 7.8 5.0 8.7 13.7 0.5 0.5 0.5 0.5 0.5 0.5 3.8 3.5 2.5 0.3 6.5 6.5 6.7 Tables 3 (B-L) in appendix, 'pp -• Manganese sulfate 132 mg, menadione 0.484 mg, r ibof lavin 3.96 mg, calcium pantothenate 9.24 mg, folacin 0.55 mg, pyridoxine HCL 2.86 mg, b iot in 0.088 mg, choline chloride 1.32 gm, vitamin E 18.17 I.U., vitamin A 4400 I.U., vitamin D^  440 I .C.U. , vitamin B12 0.01318 mg, amprolium 124.9 mg, Zn bacitracin 27 mg, and oleandomycin 11 mg per kg. 4. RESULTS A. CHEMICAL TESTS (i) Herring meal Two weeks after receiving the herring meal from the p i lo t plant, the presscake meal made from the whole carcass was analyzed and i t s composition was found to be as follows, on a dry matter basis : 78% crude protein, 10.01% ash, and 12.14% ether-extractable fat . The changes in some chemical characteristics during storage of the different herring meals are shown in Figures 1-16. A l l the meals were analyzed for moisture content at the time of each set of analyses. The results of the various analyses are expressed on the basis of the moisture-free meals. There was a progressive decrease in the amounts of ether-extract-able fat in the presscake; freeze-dried meal stored under a i r , and freeze-dried meal stored under nitrogen (meals 1, 2 and 3, respectively) during the nine-months' storage period (Figure 1). The decline was more pronounced in meals 2 and 3 than in meal 1. In the two meals treated with ethoxyquin (meals 6 and 7) and stored at room temperature and -20°C respectively, there was a sl ight change in the amount of ether-extractable fat during the nine months of storage. In meals 4 and 5, which received no antioxidant and which were stored at room temperature and -20°C respectively, a decrease in the ether-extractable fat from 11.5% to 8.7% for meal 4 and from 11.5% to 10% for meal 5 was observed (Figure 2). 43 44 The changes in iodine value of the ether extracts were generally proportional to the changes in the amounts of ether-extractable l i p i d . The iodine values of the l ip ids extracted from the presscake and from the two freeze-dried meals f e l l sharply during the f i r s t two months of storage. The rate of decrease, as indicated in Figure 3, was more pronounced in the two freeze-dried meals (meals 2 and 3) than in the presscake .(meal (1) :•. '• There was a large difference in the iodine value after eight weeks of storage between the l ip ids of the unstabilized herring meals (meals 4 and 5) and those of the stabi l ized ones, (meals 6 and 7) in favour of the ethoxyquin-stabilized meals, Figure 4. The iodine value of the ethoxyquin-stabilized meals (meals 6 and 7) had dropped by about 14 units from their or ig inal values, after sixteen weeks of storage. No signif icant changes in the iodine value occurred in these two meals dur-ing the subsequent storage period. The unstabilized meals (meals 4 and 5) showed different behaviour during the storage period. Lipids from meal 5 (stored at -20°C) exhibited a more rapid drop in the iodine value at the eighth week of storage and then no signif icant changes occurred thereafter. Lipids from meal 4 (stored at room temperature) showed a slower decline of the iodine value during the f i r s t two months of storage as compared to meal 5. The iodine value for meal 4, however, became less than that of meal 5 after the sixteenth week of storage by about 8 units. During the subsequent storage period a sl ight decrease in the iodine value was noticed in the l i p i d of meal 4. The peroxide values were calculated as mill iequivalents per kilogram of o i l , and are given in Figures 5 and 6. Figure 5 gives the results obtained for the presscake and the two freeze-dried meals. There was a considerable amount of peroxide accumulated during the f i r s t month of storage, followed by a decline. For the presscake, however, another maximum of 158.3 mill iequivalent per kilogram o i l sample at twenty-four weeks of storage was noted which then declined to 24 mill iequivalent per kilogram of o i l at the end of the storage period. There was more peroxide accumulated during the storage period for the unstabilized meals (meals 4 and 5) as compared to the stabi l ized ones (meals 6 and 7), Figure 6. However, the peroxide value of meal 4 (which was unstabilized and stored at room temperature) increased from 0 to 77.2 mill iequivalent per kilogram of o i l , a t the sixteenth weeks of stor -age, while the corresponding values for meal 5 (which was also unstabilized but stored at - 2 0 ° C ) , were 0 and 32.0 mill iequivalent per kilogram of o i l . A decrease in the peroxide value was noted for both meals thereafter. The results of the thiobarbituric acid (TBA) test were expressed as absorbance at 535 nm per gram o i l and are presented in Figures 7 and 8 . , For the presscake and both freeze-dried meals (Figure 7), there was a rapid increase in the absorbance at 535 nm during the f i r s t four weeks followed by a decline that continued unt i l eight weeks. In the case of the freeze-dried meals, whether stored under a i r (meal 2) or under n i t ro -gen (meal 3), the decline in TBA value continued unt i l the end of the storage period at 36 weeks. In the case of the presscake, however, the TBA-reactive products accumulated again following the eighth week of storage and reached a peak of 27.6 at the 24th week of storage. After this time absorbance at 535 nm again decreased. 46 It should be noted from Figure 8 that the unstabilized meals (meals 4 and 5) showed more malonaldehyde formation than did the meals treated with ethoxyquin (meals 6 and 7) during storage, regardless of the storage temperature. A decrease in the ultraviolet absorption at 233 nm of hydro-peroxides containing conjugated double bonds, has been taken as an i n d i -cation of oxidation. The results of this test are shown in Figures 9 and 10. As indicated in Figure 9, a large accumulation of the conjugated hydroperoxides for the presscake was'found "to occur atTthe^eighth week of storage, followed by a continuous f a l l in the hydroperoxide con-tent. Freeze-dried herring meals (meals 2 and 3) showed their peaks of hydroperoxide accumulation at the sixteenth week of storage with greater accumulation in the freeze-dried meal stored under a i r (absorbance = 3.0) than in the meal stored under nitrogen (absorbance = 2.46). Ether-extract from herring meal stabi l ized with ethoxyquin and stored at -20°C (meal 6) showed a sl ight increase in the ultraviolet absorption at 233 nm. An extract from the stabi l ized meal stored at room temperature showed s l ight ly more increase in the absorption at 233 nm than that from meal 6 during the f i r s t sixteenth week of storage, Figure 10. On the other hand, the unstabilized meals (meals 4 and 5) exhibited more hydroperoxide accumulation when compared to the stabi l ized herring meals (meals 6 and 7). In meal 4 extract (stored at room temperature), there was a continuous increase in the ul t rav io let absorption at 233 nm up to 24 weeks of storage (absorbance = 2.67) after which time there was a decrease. The unstabilized herring meal stored at -20°C (meal 5) had 47 i t s peak of hydroperoxide accumulation (absorbance = 1.9) at the eighth week of storage, which was followed by a quick drop in the absorption at the sixteenth week and a continuing but slower decline thereafter to a value of 0.82. The available lysine in the meals was calculated as a percentage of the total protein and presented in Figures 11 and 12. Figure 11 i n -dicates that the available lysine content of presscake and Ifreeze-dried meals (stored under a i r and under nitrogen) decreased during the f i r s t 25 weeks of storage, but there was l i t t l e change thereafter. There was a gradual decrease in the available lysine content in the herring meals during the storage period including those stabi l ized and unstabi l -ized meals, stored at both room temperature and -20°C (meals 4, 5, 6 and 7), Figure 12. The rate of decrease in the content of available lysine was more pronounced for the unstabilized meal stored at room temperature as can be seen from the slope of the curve in Figure 12. The pepsin d igest ib i l i t y of the presscake andj thei-two 5freeze-dried meals showed sl ight decreases in values during storage, Figure 13. The changes in the per cent pepsin-digestible protein during storage were minimal for the stabi l ized herring meals (meals 6 and 7) and for the un-stabi l ized herring meal stored at -20°C (meal 5) , Figure 14. On the other hand the unstabilized herring meal which was stored at room temperature, showed a noticeable decrease in the pepsin d igest ib i l i t y from 95% to 92.8% during the storage period. There was a small decrease in the available lysine content of the solvent-extracted meals (meals 8 and 9), as indicated in Figure 15, during 48 storage. The rate of decrease was s l ight ly faster for the solvent-extracted meal stored at room temperature than that stored at -20°C. The pepsin d igest ib i l i t y of the solvent-extracted meals, whether stored at room temperature or at -20°C, changed only s l ight l y , Figure 16. By the end of the storage time both meals contained the same percentage of digestible protein, however, the i n i t i a l rate of decline was s l ight ly faster for the solvent-extracted meal stored at room temperature than for the one stored at -20°C. There was no noticeable change in the peptide maps of the pepsin digest among the differently - treated herring meals nor during storage. The general outline of the maps is presented in Figure 17. 49 12 > cn o »-o < cn \— x UJ or UJ x »-UJ 10 t _L 0 2 4 8 16 24 WEEKS IN STORAGE 36 Fig-1 - Ether-extractable fat from presscake stored at - Z 0 ° C- ( ),freeze-dried meal stored in nitrogen ( ),and freeze-dried meal stored in air (—=—)• 50 14 -13 -6 12 o I" oc Id X \-UJ * 8 - A t 0 2 8 16 24 WEEKS IN STORAGE 36 Fig- 2- Ether-extractable fat from herring meal treated with antioxidant and stored at 21° C- ( A — ) , herring meal treated with antioxidant and stored a t - 2 0 ° C ( o ),herring meal stored at 21° C ( A ),and herring meal stored a t - 2 0 ° C -( — o - - - - ) • 120 110 100 90 UJ 3 < 80 UJ 2 70 Q O 60 50 40 1 0 2 4 8 16 24 WEEKS IN STORAGE 36 Fig- 3- Iodine values of ether-extractable fat of presscake stored at - 20° C- ( freeze-dried meal stored in nitrogen ( — and f reeze-dr ied meal stored in air ( — 0 2 8 16 24 36 WEEKS IN STORAGE Fig- 4- Iodine values of ether-extractable fat of herring meal treated with antioxidant and stored at 21° C- ( A — herring meal treated with antioxidant and stored at - 2 0 ° C ( o — ) , h e r r i n g meal stored at 21°C-( A ),and herring meal stored a t - 2 0 ° C - ( - —O 700 -53 0 2 4 8 16 24 WEEKS IN STORAGE Fig- 5- Peroxide values of ether-extractable fat of presscake stored a t - 2 0 °C-( ),freeze-dried meal stored in nitrogen ( ),and freeze-dried meal stored in air ( )• 80 70 60 CD CO z 50 Id _i I 3 O UJ _l _J s CO* UJ Z> 40 < 30 UJ o x o or ^ 20 10 A l \ \ \ A , \ \ C> 0 2 8 16 24 WEEKS IN STORAGE Fig-6 Peroxide values of ether-extractable fat of herring meal treated with antioxidant and stored at 21° C ( A ), herring meal treated with antioxidant and stored at —20° ( — o — ) , herring meal stored at 21° C- ( A — ) and herring meal stored a t - 2 0 ° C - ( — o — ) • 55 m to IT) < CD »-UJ o z < CO CC O CO m 60 50 40 30 20 10 J—L _L ± 0 2 4 8 16 24 WEEKS IN STORAGE 36 Fig- 7- Storage changes in T B A absorbance at 535 NM calculated per gram of oil in presscake stored a t - 2 0 ° C - ( ),freeze-dried meal stored in nitrogen ( ),and freeze-dried meal stored in air ( )• 56 40 -in ro „ m 30 < CD UJ ^ 20 < CO or o CO CO < 10 0 A ± 0 2 8 16 WEEKS IN * 24 36 oTORAGE Fig- 8- Storage changes in T B A absorbance at 535 NM calculated per gram of oil in herring meal treated with antioxidant and stored at 21° C-( A—) ,her r ing meal treated with antioxidant and stored at - 2 0 ° C-( o ),herring meal stored at 21° C-( A — ) , and herring meal stored at - 2 0 ° C - ( o )• 3-6 3-2 2-8 2-4 2-0 1-6 1-2 0-8 t I L 0 2 4 8 1 6 - 2 4 WEEKS IN STORAGE 36 Fig- 9- Storage changes in absorbance at 233 NM of lipid extract of presscake stored at - 2 0 ° C-( -),freeze-dried meal stored in nitrogen (— and freeze-dried meal stored in air ( =—)• 2-8 2-4 2 0 1-6 I 12 to 5 In ^ 0-8 UJ 0-4 o-o 58 A o / A ' - V / / \ JL _L 0 2 8 16 24 WEEKS IN STORAGE 36 Fig- 10- Storage changes in absorbance at 233 NM of lipid extract of herring meal treated with antioxidant and stored at 21 ° C- ( A — ) , h e r r i n g meal treated with antioxidant and stored a t - 2 0 ° C ( o ),herring meal stored at 2 I ° C ( A — ) , a n d herring meal stored at - 2 0 ° C- ( o )• l i l t • ' 0 3 5 9 17 25 37 WEEKS IN STORAGE •II- Available lysine changes during storage-Presscake stored a t - 2 0 ° C-( -), freeze-dried meal stored in nitrogen ( and freeze-dried meal stored in air { ) Available lysine values expressed as a percentage of the total protein-< > < 3* ^*A— " A L J i i 1 — i 0 3 9 17 25 37 WEEKS IN STORAGE Fig- 12- Available lysine changes during storage-Herring meal treated with antioxidant and stored at 21° C-( A — ) , herring meal treated with antioxidant and stored a t - 2 0 ° C - ( o ),herring meal stored at 21° C-( A ),and herring meal stored a t - 2 0 ° C - ( o — ) • Available lysine values expressed as a percentage of the total protein-61 100 >• - J CO 95 UJ o Q z 90 CO Q. UJ Q. 5« ' ' ' -L L 3 5 9 17 25 37 WEEKS IN STORAGE Fig-13- Pepsin digestibility changes during storage-Presscake stored at - 20 0 C-( -), freeze-dried meal stored in nitrogen ( ), and freeze-dried meal stored in air ( )• Pepsin digestibility expressed as a percentage of the total protein-62 100 p 95 CO uj o CO Q. UJ a. 90 t J I L 0 3 9 17 25 37 WEEKS IN STORAGE Fig-14- Pepsin digestibility changes during storage-Herring meal treated with antioxidant and stored at 21° C-( — - A ),herring meal treated with antioxidant and stored at - 2 0 ° C- ( o ), herring meal stored at 21° C- ( A -— ) ,and herring meal stored a t - 2 0 ° C- ( — o )• Pepsin digestibility expressed as a percentage of the total protein-L ~ J I l I I 0 3 9 17 25 37 WEEKS IN STORAGE Fig-15- Available lysine changes during storage-Solvent-extracted meal stored 21° C- ( —) and solvent-extracted meal stored a t - 2 0 ° C -( )-Available lysine values expressed as a percentage of the total protein-100 5 95 h-co UJ o a z to 9 0 H UJ D_ I ' J I I 0 3 9 17 25 37 WEEKS IN STORAGE Fig-16- Pepsin digestibility changes during storage-Solvent-extracted meal stored at 21° C-( and solvent-extracted meal stored a t - 2 0 ° C -( )• Pepsin digestibility expressed as a percentage of the total protein-3§&very intense (fify in tense O less intense ; ' 3 fa int a. a w o> o o E o o E lec t rophores is Fig-17- Tracing of a typical peptide map of pepsin digests of herring meal-66 ( i i ) Model system The model system prepared was composed of 88% egg albumin and 12% herring o i l . This model system was divided into two portions; one was stored at room temperature while the other was stored at -20°C. A decrease in the ether-extractable fat from 12% to 10.1% occurred during the storage of the model system at -20°C, Figure 18. The model system which was stored at room temperature also showed a similar drop in the ether-extractable fat (from 12% to 9.8%). The iodine value of the freshly produced herring o i l decreased s l ight ly when the o i l was stored alone for 40 weeks whether the o i l was stored at room temperature or at -20°C, Figure 19. Also, there was a considerable decrease in the iodine value of the ether extracts of the two model systems during storage. The iodine values f e l l during the storage period by about 50 units from the or iginal values. After 40 weeks of storage there was no difference between the iodine value of ether-extract from the model system stored at room temperature and that from the model system stored at -20°C. There was, however, a difference between the two model systems in the rate at which the iodine value de-creased during storage, Figure 19. The rate of decrease in the iodine value was more rapid in the model system stored at room temperature than in the system stored at -20°C. As shown in Figure 20, there was a gradual peroxide accumulation in the herring o i l stored alone at room temperature and at -20°C unt i l maximum accumulation at the 32nd week of storage. The peroxide values for the two model systems increased rapidly during the f i r s t twelve weeks 67 of storage and then declined gradually during the subsequent storage period. From the results of the TBA test , i t was noted (Figure 21) that the absorbance of the TBA-reactive compounds at 535 nm increased during the storage period to maxima of 11.2 for o i l stored alone at room tem-perature, 10.2 for o i l stored at -20°C, 12.7 for the model system stored at room temperature, and 11.9 for the model system stored at -20°C at the 20th week of storage. After 20 weeks the TBA value declined in every instance. The absorbance was consistently s l ight ly higher in the two model systems, than in the herring o i l samples stored alone. Some changes in the ultraviolet absorption of the conjugated hydroperoxides (at 233 nm) were observed among the different samples during storage (Figure 22). In samples of herring o i l stored alone (both at room temperature and -20°C) , there was a sl ight gradual accumulation of the diene conjugated hydroperoxides during the course of the storage period. The fat extracted from the two model systems showed more hydro-peroxide formation when compared to the amount of.hydroperoxide bu i l t up in the o i l samples stored alone. For example, at the twelfth week of storage the o i l sample stored at room temperature gave an absorbance at 233 nm of 0.38 while the o i l stored in close association with albumin (the model system) gave an absorbance of 1.7. The corresponding values in case of storage at -20°C were 0.35 and 2.08. The available lysine content of albumin stored at room temperature and that of albumin stored at -20°C was found to decrease from 5.89% to 5.17% and from 5.89% to 5.38% respectively, during the storage period which 68 lasted for 40 weeks (Figure 23). The level of available lysine decreased more rapidly in the two model systems than in the albumin samples stored alone. The loss of available lysine was part icular ly rapid in the model system stored at room temperature. The data in Figure 24 indicated that there were^no^apparent differences in pepsin d igest ib i l i t y between the albumin and the two model systems during the f i r s t 32 weeks of storage regardless of temper-ature. By the 40th week of storage, however, the albumin samples had values about lh units higher than did the model systems. Typical peptide maps of the peptic digests are given in Figures 25-33. In general, a decrease in the number and intensity of the peptide spots occurred as a result of storage. 0 12 20 32 40 WEEKS IN STORAGE •18- Ether-extractable fat from model system stored at 21° C ( A — ) and from model system stored a t - 2 0 ° C ( - - - o — ) • 70 150 -100 50 -12 20 32 WEEKS IN STORAGE 40 Fig-19- Iodine values of herring oil stored at 21° C- (—Ar herring oil stored a t - 2 0 ° C - ( — o — ) , ether-extractable fat of model system stored at 2 I ° C -( — A — ) , a n d ether-extractable fat of model system stored at - 2 0 ° C- ( - - - o - - - ) -0 0 12 20 32 40 WEEKS IN STORAGE Fig- 20- Peroxide values of herring oil stored at 21° C-( A—),herr ing oil stored a t - 2 0 ° C - ( O — ) , ether-extractable fat of model system stored at 21° C- ( — A — ) , a n d ether-extractable fat of model system stored at - 2 0 ° C- ( o — ) • < 4 Ol I I I L— 0 12 20 32 40 WEEKS IN STORAGE FIG- 21- Storage changes in T B A absorbance at 535 NM calculated per gram of oil in herring oil stored at 21 ° C - ( — A — ) , h e r r i n g oil stored at - 2 0 ° C-( o ), model system stored at 21° C ( — A — ) , and model system stored at - 2 0 ° C ( — o — ) • 2-4 O 0 12 20 32 40 W E E K S IN STORAGE 22- Storage changes in absorbance at 233 NM of herring oil stored at 21° C- ( — A — ) , h e r r i n g oil stored a t - 2 0 ° C -( o — ) , l i p i d extract of model system stored at 21° C-( — A — ) , a n d lipid extract of model system stored at - 2 0 ° C ( - - - o — - ) • 74 12 20 32 40 WEEKS IN STORAGE Fig- 23- Available lysine changes during storage-Albumin stored at 21° C- ( A ),albumin stored at - 2 0 ° C- ( - — o ),model system stored at 21 °C-( A ),and model system stored at - 2 0 ° C-( o )• Available lysine values expressed as a percentage of the total protein-75 105 100 12 20 32 WEEKS IN STORAGE 24- Pepsin digestibility changes during storage-Albumin stored at 21 ° C- ( A—) ,a lbumin stored at - 20°C-( o ), model system stored at 2 I ° C ( A — ) , and model system stored a t - 2 0 ° C ( o )• Pepsin digestibility expressed as a percentage of the total protein-76 very Intense intense O less intense Qfaint 0 a. o I s Electrophoresis Fig - 25- Tracing of the peptide map of a pepsin digest of fresh albumin-in tense ^ O less intense O faint 0 77 0 o a 9 E l e c t r o p h o r e s i s Fig- 26- Tracing of the peptide map of pepsin digest of albumin stored at 21° C- for 6 months-o o intense u O less intense fa in t 0 *1 <3 E l e c t r o p h o r e s i s 27- Tracing of the peptide map of pepsin digest of albumin stored at - 20° C- for 6 months-79 very intense tUD> intense o less intense E l e c t r o p h o r e s i s Fig- 28- Tracing of the peptide map of pepsin digest of the model system stored at 21° C- for 6 months-8 0 (SP intense O less intense Qfoint Q Q Q. O w o» o o E o Electrophoresis Fig- 29 - Tracing of the peptide map of pepsin digest of the model system stored at - 2 0 ° C- for 6 months-81 i i intense O 'ess intense O f a i n t D 0 <DHi 0 >» l-c a. o o o E o E l e c t r o p h o r e s i s Fig- 3 0 - Tracing of the peptide map of pepsin digest of the albumin stored at 21 °C- for II months-82 in tense O ' e s s intense O fa int <0 b 0 • I I I % # \ f o I \ I I Q. a w o» O o E o b. l o Electrophoresis Fig- 31- Tracing of the peptide map of pepsin digest of the albumin stored at - 2 0 ° C- for II months-83 (Jj> intense ^ O 'ess intense ^ 0 dD 0 Electrophoresis >» JZ tx o o E o Fig- 32- Tracing of the peptide map of pepsin digest of the model system stored at 21° C- for II months-d j ^ intense O less intense <— Electrophoresis Fig- 33- Tracing of the peptide map of pepsin digest of the model system stored at - 2 0 ° C-for 11 months-85 B. BIOLOGICAL ASSAYS (i) Metabolizable energy tests Experiment 1 (Four-months storage): - The metabolizable energy values obtained for the different herring meal samples tested after four months of storage are summarized in Table 4. The analysis of variance showed that temperature, antioxidant, and their interaction had signif icant effects on the metabolizable energy of herring meals. The data also showed an improvement in the u t i l i za t ion of the ethoxyquin treated meal (meal A) over the untreated meal (meal B) when both meals were stored at room temperature for four months. However, there was no signif icant difference between the stabi l ized and unstabilized meals (meals C and D) stored at -20°C in the metabolizable energy content according to the analysis of variance and Duncan's multiple range test. In addition the average metabolizable energy content of meals A, C and D (3910, 3890 and 3850 cal/gm respectively) were not s ta t i s t i ca l l y d i f f e r -ent. Meal B had average metabolizable energy value of 3395 cal/gm which was s ignif icant ly lower than the values for meals A, C, and D. The average metabolizable energy content of the solvent-extracted meal stored at -20°C (3850 cal/gm) was higher than that of a similar meal stored at room temperature (3590 cal/gm). This difference was s t a t i s t i -cally s ignif icant at the 1% level (Table 4B). Experiment 2 (Nine-months storage): - The results from the second metabolizable energy experiment conducted after nine months of storage are summarized in Table 5. The results showed again that the addition of ethoxyquin prevented the decline in energy ava i lab i l i ty from herring 86 meal stored at room temperature that occurred in the untreated meal stored for the same period of time at room temperature. There were no s ta t i s t i ca l l y signif icant differences in the metabolizable energy values among meals A, C, and D according to the Duncan's multiple range test. The metabolizable energy content of meal B (3280 cal/gm) was s ignif icant ly lower than the values for meals A, C, and D (3935, 3950 and 3905 cal/gm respectively). The solvent-extracted meal which was stored at -20°C for nine months showed a s ignif icant ly higher metabolizable energy content than the solvent-extracted meal stored at room temperature for the same period of time, (3605 vs. 3500 cal/gm respectively). The metabolizable energy values for the two experiments were combined and subjected to a three-way factor ia l analysis of variance, Table 5(C). There were signif icant effects due to temperature, an t i -oxidant, interaction between storage and temperature, interaction between antioxidant and storage, interaction between temperature and antioxidant and interaction between storage, temperature and antioxidant. The herring meals stored at room temperature without antioxidant had the lowest metabolizable energy values at both four months and nine months storage. In addition,,the unstabilized meal which was stored for nine months at room temperature was s ignif icant ly lower in i t s metabolizable energy content than the same meal after only four months storage. There were no major changes in the metabolizable energy content during the storage period as far as meals A, C, and D were concerned, accoding to Duncan's multiple range test , Table 5(c). When the 87 metabolizable energy values for the solvent-extracted meals from the two metabolizable energy experiments were combined for s t a t i s t i c a l analysis, the results indicated that there were signif icant effects due to storage, temperature and their interaction, Table 5(D). The metabolizable energy values for the solvent-extracted meals stored for nine months (meals E and.F) were s ignif icant ly lower than for the same meals stored for only four months. ( i i ) Protein quality tests Experiment 1 (Five-months storage): - The results of the f i r s t supplementary protein quality test , which was conducted after f ive months of storage are given in Table 4. The data were calculated as gains in weight of chicks fed the differently treated herring meals. The results were subjected to two-way analysis of variance. It was found that both storage temperature and antioxidant treatment affected the protein quality of the meal. The effects were signif icant when the herring meals were fed to supply 11% of supplementary protein to the diet. Chicks fed the herring meal stored without antioxidant for f ive months at room tempera-ture (meal B) had the poorest rate of growth at the respective levels of supplementation. The gains in weight of the chicks fed meals A, C and D were not s ta t i s t i ca l l y different at any of the levels fed. There was no signif icant difference in the growth response to the two solvent-extracted meals (meals E and F) according to the analysis of variance. The results of the f i r s t experiment were also expressed in terms of the eff iciency of feed conversion (gain/feed) and are shown in Table 4. 88 There were no significant differences when the values, expressed as efficiency of feed conversion, were analyzed s t a t i s t i c a l l y . The application of Duncan's multiple range test, however, revealed some s t a t i s t i c a l differences. These differences were not consistent for the three levels of supplementary protein and no conclusions could be drawn from these values. The difference in the efficiency of feed conversion between the two solvent-extracted meals was not significant, Table 4 . The differences in the growth response for the six herring meals tested, whether expressed as gain i n weight or as diet efficiency, among the three levels of supplementary protein used ( 4 , 7 and 1 1 % ) were found to be significant as shown in Tables 4 ( G ) - 4 ( J ) . Experiment 2 (Ten-months storage):- The second supplementary protein quality test was conducted at the end of the storage period which lasted for ten months. The data are summarized in Table 5. When the results were expressed in terms of gain in weight, meal B, which contained no antioxidant and was stored at room temperature, showed the poorest response at a l l three levels of supplementation. Meanwhile, when meals A, C, and D were fed to chicks they gave the same growth response at the three levels used. There was no significant difference between the response obtained with meals E and F in terms of gain in weight, Table 5. The results of this test were also expressed as efficiency of feed u t i l i z a t i o n and given in Table 5. The s t a t i s t i c a l evaluation of this data revealed some significant differences among the different herring meals 89 tested. The application of Duncan's test indicated that the eff iciency of feed ut i l i za t ion of meal C (antioxidant-treated and stored at -20°C) was higher than the corresponding eff ic iencies of feed u t i l i za t ion of meals A, B and D, when they were a l l fed to supply 7% of protein to the diet . The difference between the solvent-extracted meals (meals E and F) was signif icant when they were fed to supply 7% or 11% of protein to the diet . Meal F (solvent-extracted meal stored at -20°C) promoted more ef f ic ient feed conversion than did meal E (solvent-extracted meal stored at room temperature). When these two meals were fed to supply 4% of protein to the diet , however, there was no s ta t i s t i ca l l y s ignif icant difference between them, Table 5. The analysis of variance revealed highly s ignif icant differences among the levels of supplementary protein used whether the results were expressed as gain in weight (Table 5(1) and 5(J)) or as diet eff iciency (Tables 5(K) and 5(L))..for the.six herring meals stored for ten months. In order to be able to compare the two supplementary protein quality tests , whether expressed as gains in weight or eff iciency of feed u t i l i z a t i o n , the results were calculated as percentages of the control diet . Definite nutr i t ive changes (Tables 6 and 7) occurred in the meals upon storage as indicated by the loss in growth response and the decrease in the eff iciency of feed u t i l i za t ion in the different herring meals studied. It should also be noted that there was a considerable difference in the growth response and the feed eff iciency between the control diets in both storage experiments. Such a difference may be accounted for by 90 the difference in protein quality of wheat used in formulating the two control diets. ( i i i ) In vivo d igest ib i l i t y The soluble and insoluble nitrogen and dry matter contents of the lumen of small intestines of chickens after one- and one-half hours from receiving the test meals are shown in Table 8. The amounts of insoluble nitrogen and soluble nitrogen in the lumen of the small intestine of birds receiving the untreated herring meals were higher than those re -ceiving the stabi l ized meals regardless of the storage temperature. It was also noted that the antioxidant-treated meal stored at room tempera-ture (meal A) gave more unabsorbed nitrogen (soluble and insoluble nitrogen) than the meal treated with antioxidant but stored at -20°C (meal C). In the untreated meals, the amounts of residual soluble and insoluble nitrogen were greater from the meal stored at room temperature than from the meal stored at -20°C (meals B and D respectively). The intest inal lumen of the chickens given the solvent-extracted meal stored at room temperature (meal E) , contained 0.0357 or 0.0352 gm of undissolved nitrogen whereas in the chickens fed the solvent-extracted meal stored at -20°C (meal F) the levels were only 0.0175 and 0.0510 gm respectively. The amount of dissolved nitrogen which remained unabsorbed in the intest inal lumen was 0.0619 and 0.0510 gm with meal E and 0.0344 and 0.0389 gm with meal F. Figures 34-39 (Test 1) show the composition of the intest inal soluble nitrogen fraction in terms of dissolved protein, peptides, and free amino acids expressed as leucine equivalent, in the intest inal lumen 91 contents of birds fed the various d iets , and of fasted bird which was taken to represent the endogenous secretion of nitrogen. In Table 9 (Test 1) are given the amounts of dissolved proteins, peptides, and free amino acids expressed as area units under the different curves in Figures 34-39. Greater amounts of dissolved proteins, peptides and amino acids were found when the unstabilized meals were fed as com-pared to the stabi l ized ones, regardless of the storage temperature, (Figures 34 and 35). On the basis of the unabsorbed soluble nitrogen residue in the intestine of birds lh hour after receiving the respective meals, i t was found that the antioxidant-treated meal stored at room temperature gave greater amounts of peptides and free amino acids than the meal which was treated with antioxidant but stored at -20°C. It was also noted that in the untreated meals, the residual amounts of peptides and free amino acids were greater from the meal stored at room temperature than from that meal stored at -20°C (Figures 36 and 37). When the two solvent-extracted meals (meals E and F) were compared on the basis of the composition of the residual soluble nitrogen in the lumen of the intestine of chicks lh hour after receiving the different meals, larger amounts of dissolved protein, peptides, and amino acids were found when the solvent-extracted meal stored at room temperature was given than when the solvent-extracted meal stored at -20°C was the test material , (Figure 38). The experiment was repeated with the same herring meals, and the results are given in Tables 8 (Test 1) , 9 (Test 2) and Figures 40-45. The results from the second test confirmed the findings of the f i r s t . 92 TABLE 4 METABOLIZABLE ENERGY VALUES OF HERRING MEALS STORED FOR FOUR MONTHS. GROWTH RESPONSE AND EFFICIENCY OF FEED CONVERSION OF CHICKS FED HERRING MEALS STORED FOR FIVE MONTHS Meal M.E. call gm Meals supplying 4% of protein to diet Meals supplying 7% of protein to diet Meals supplying 11% of protein td diet Gain in wt. (gm) Gain Feed Gain in wt. (gm) Gain Feed Gain in wt. (gm) Gain Feed A 3910a 71.48 a 0.360 a 90.44 a 0.447 a 106.46 a 0.563 3 B 3395b 63.19 3 0.345 a b 80.99 a 0.435 a 100.23 b 0.521 b C 3890a 67.45 a 0.335 b 88.64 a 0.433 a 107.16 a 0.539 a b D 3850a 67.95 a 0.342 b 87.16 a 0.451 a 106.02 a 0.541 a b E 3590b 71.23 a 0.360 a 84.43 a 0.463 3 99.06 3 0.519 a F 3850a 70.44 a 0.364 a 89.98 a 0.458 a 108.06 a 0.533 a Control diet (no fishmeal) 10.4 0.111 The s t a t i s t i c a l analyses of the data in this Table are shown in Tables 4(A)-4(J) in the appendix. Values within each column having the same superscript are not s ta t i s t i ca l l y different (P<0.05 for gain in weights and feed eff ic iency values, P<0.01 for metabolizable energy values). The values for meals A-D have been analyzed separately from the values for meals E and F. 93 TABLE 5* METABOLIZABLE ENERGY VALUES OF HERRING MEAL STORED FOR NINE MONTHS. GROWTH RESPONSE AND EFFICIENCY OF FEED CONVERSION OF CHICKS FED HERRING MEALS STORED FOR TEN MONTHS Meal M.E. Meals supplying 4% Meals supplying 7% Meals supplying 11% cal/gm of protein to diet of protein to diet of protein to diet Gain in wt. (gm) Gain Feed Gain in wt. (gm) Gain Feed Gain in wt. (gm) Gain Feed A 3935a 72.58 a 0 .367 b c 97.56 a 0.467 3 115.78 a 0.539 a B 3280b 65.02 b 0.358 C 85.00 a 0.465 a 106.90 a 0.534 a C 3950a 72.35 a 0.372 a b 95.90 a 0.483 b 111.04 a 0.557 a D 3905a 73.94 a 0.384 a 94.82 a 0.465 a 115.54 a 0.552 a E 3500b 67.15 3 0.365 a 83.29 a 0.455 b 106.23 a 0.532 b F 3605a 70.27 a 0.367 a 86.28 a 0.466 a 112.83 a 0.551 a Control diet (no fishmeal) 19.4 0.173 The s t a t i s t i c a l analyses of the data in this Table are shown in Tables 5(A), (B), (E)-(L) in the appendix. Values within each column having the same superscript are not s ta t i s t i ca l l y different (P<0.05 for gain in weights and feed eff iciency values, P<0.01 for metabolizable energy values). The values for meals A-D have been analyzed separately from the values for meals E and F. 94 TABLE 6 SUPPLEMENTARY PROTEIN VALUES OF HERRING MEALS STORED FOR FIVE MONTHS Meal Meals supplying 4% of protein to diet Meals supplying 7% of protein to diet Meals supplying 11% of protein to diet S.P.V. S.P.V. S.P.V. S.P.V. S.P.V. S.P.V. gain diet ef f . gain diet e f f . gain diet e f f . A 687 324 870 403 1024 507 B 608 311 779 392 964 469 C 649 302 852 390 1030 486 D 653 308 838 406 1019 487 E 685 324 812 417 953 468 F 677 328 865 413 1039 480 95 TABLE 7 SUPPLEMENTARY PROTEIN VALUES OF HERRING MEALS STORED FOR TEN MONTHS Meal Meals supplying 4% Meals supplying 7% Meals supplying 11% of protein to diet of protein to diet of protein to diet S.P.V. gain S.P.V. diet ef f . S.P.V. gain S.P.V. diet ef f . S.P.V. gain S.P.V. diet ef f . A 374 212 503 270 596 312 B 335 207 438 269 551 309 C 373 215 494 279 572 322 D 381 222 488 269 595 319 E 346 211 429 263 547 308 F 362 212 444 269 581 319 TABLE 8 NITROGEN AND DRY MATTER CONTENT OF THE LUMEN OF THE SMALL INTESTINE OF ADULT CHICKENS 1% HOUR AFTER INGESTING THE DIFFERENT HERRING MEALS STORED FOR 11 MONTHS TEST 1: Diet A B C D E F Control (starved) Content of the Small Intestine Total nitrogen Sol . Insol. Sol . Dry matter Insol. .0400 .0240 .4821 .6343 .0441 .0384 .4897 .8707 .0338 .0224 .4656 .4882 .0375 .0339 .4990 .7480 .0619 .0357 .5548 .7659 .0344 .0175 .4346 .3968 .0223 .0081 .4050 .2061 Content of the Small Intestine Total nitrogen Dry matter Sol . Insol. Sol . Insol. .0483 .0227 .5000 .4443 .0556 .03374 .5485 .5240 .0372 .0222 .4846 .4254 .0427 .02434 .4965 .4965 .0501 .0352 .5784 .5833 .0389 .0227 .4359 .5348 .0245 .0089 .4359 .2180 TEST 2: Diet A B C D E F Control (starved) 97 TABLE 9 THE RELATIVE AMOUNTS OF DISSOLVED PROTEIN, PEPTIDES AND FREE AMINO ACIDS IN THE SOLUBLE NITROGEN FRACTION OF THE CONTENTS OF THE SMALL INTESTINE OF CHICKENS, 1% HOURS AFTER INGESTING DIFFERENT HERRING MEALS. THE AMOUNTS EXPRESSED AS AREA UNITS UNDER THE GRAPHS IN FIGURES 34-45 TEST 1: Diet Proteins Peptides Amino acids A 144 524 421 B 255 964 581 C 152 389 303 D 313 560 332 E 460 1566 660 F 186 268 391 Control (starved) 132 278 231 TEST 2: Diet Proteins Peptides Amino acids A 186 553 452 B 299 997 709 C 137 307 290 D 271 557 463 E 449 1263 764 F 129 335 359 Control (starved) 132 294 218 98 ELUATE VOLUME (ML) Fig-34- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens 11/2 hours after ingesting herring meal treated with antioxidant and stored at 21° C ( ),and herring meal stored at 21 ° C- ( )• 99 Fig*35- Fractionation by Sephadex G - 2 5 o f the soluble nitrogen of the intestinal contents of chickens 11/2 hours after ingesting herring meal treated with antioxidant and stored at - 2 0 ° C ( ), and herring meal stored at - 2 0 ° C- ( )• 100 1 139 175 Eiteins . _J1 274 337 391 Peptides Peptides •Amino ac ELUATE VOLUME (ML) ids —*| Fig- 36- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens 1 1/2 hours after ingesting herring meal treated with antioxidant and stored at 21° C* ( ),and herring meal treated with antioxidant and stored a t - 2 0 ° C ( )• 21 - 101 UJ z t -co UJ 1-8 r -z UJ X r -z z 1-5 UJ o o or z UJ 1-2 _ l CD 3 _ l O CO _J < 0-9 r -o J -CO r-z UJ _l 0-6 < > 3 O UJ UJ CIN 0-3 3 UJ _J E 0-0 139 175 Proteins — a — * Peptides Peptides-274 337 Amino acids ELUATE VOLUME(ML) Fig- 37- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting herring meal stored at 21° C ( ),and herring meal stored a t - 2 0 ° C - ( - - ) • 102 Fig- 38- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting solvent-extracted meal stored at 21° C ( ),and solvent-extracted meal stored at - 2 0 ° C-( )• 103 UJ _J m 2 to 5 U J UJ X 0-4 to »-z UJ - I z S z 5 UJ - z o 3 UJ e o o 1 139 175 Eoteins a - * . „ptides Peptides 274 —**- Amino ac 337 391 ELUATE VOLUME (ML) Ids—*| Fig-39- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chicken starved for 18 hours-1 0 4 Fig-40- Fractionation by Sephadex G -25 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting herring meal treated with antioxidant and stored at 21 ° C- ( •), and herring meal stored at 21° C- ( )• 105 Fig- 41- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting herring meal treated with antioxidant and stored a t - 2 0 ° C-( ),and herring meal stored a t - 2 0 ° C-( )• 106 ELUATE VOLUME (ML) Fig-42- Fractionation by Sephodex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting herring meal treated with antioxidant and stored at 2l°C-( ),and herring meal treated with antioxidant and stored a t - 2 0 ° C ( )• 107 Fig-43- Fractionation by Sephadex G-25of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting herring meal stored at 21° C-( ),and herring meal stored a t - 2 0 ° C ( )• 1-4 z co UJ r-z UJ 1-2 X 1-z UJ CD 10 O or z UJ -J CO LU 0-8 o CO _l < r-o 1-— 0-6 CO 1-z UJ - 1 < > 0 4 3 o UJ UJ z o -1 0-2 2 E 0 0 108 139 175 Proteins a 274 I PA eptides Peptides — ELUATE 'Amino acids VOLUME (ML) Fig-44- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chickens I 1/2 hours after ingesting solvent-extracted meal stored at 21° C- ( ),and solvent-extracted meal stored a t - 2 0 ° C - ( )• 109 Ixl CO 3 Fig-45- Fractionation by Sephadex G - 2 5 of the soluble nitrogen of the intestinal contents of chicken starved for 18 hours-5. DISCUSSION A. CHEMICAL TESTS Lipids present in herring meal are l iab le to oxidation because they contain a relat ively high proportion of unsaturated fatty acids and because the meal presents a large surface area. In this study, l ip ids extracted from herring meals showed obvious physical and chemical changes, the degree of which depended upon the temperature at which the meals were stored and whether or not the meals had been treated with an antioxidant. The changes in d i -ethy l ether extractabi l i ty and iodine value (Figures 2 and 4) during storage of the unstabilized meal are in agree-ment with previous findings on Canadian herring meal (Lea e_t a l . 1960, March et^  a l . 1961, 1966) and Norwegian f ish meal (Opstvedt et a l . 1970). Apparently the so lubi l i ty of the l i p i d or iginal ly present in the freshly prepared f ish meal in d i -ethy l ether, declines as a result of oxidation and/or polymerization during storage, (Oshima and Sugamara 1936, Stansby and Clegg 1955 and Almquist 1956). The chemical composition of the o i l i t s e l f i s such that i t i s readily oxidized. In addition to the suscepti-b i l i t y of the o i l content of the meal to oxidation, the presence of hematin in the meal catalyzes the process, (Anonymous, 1958). The disappearance of double bonds as measured by the reduction in iodine value usually serves as a valuable indication of the degree of oxidation and oxidative polymerization of the o i l . Storage temperature 110 I l l was found to be a factor in the progressive decrease in the percentage of ether-extractable material and in iodine value of this fraction of the herring meal. The effect of storage temperature on the unstabilized meals was more pronounced than on the stabi l ized ones (Figure 4). It was noted that the rate of decline in the iodine value was faster for the untreated meal stored at -20°C (meal 5) than for the one which was stored at room temperature (meal 4) during the f i r s t eight weeks of storage. This finding is in agreement with March et a l . (1961) who reported that low temperature ( -20°C) storage promoted more rapid decrease in iodine value of the ether-extract than did storage at 25°C. In this experiment, however, after four months of storage, the iodine value for the unstabilized meal stored at -20°C remained unchanged through the sub-sequent storage period while the decrease in iodine value of the unstabi l -ized meal stored at room temperature continued. The rate of decrease in the percentage of ether-extractable material from the unstabilized herring meals was found to be faster for the meal stored at room temperature (21°C) than for that stored at -20°C throughout the storage period. March et a l . (1962) s imilar ly reported that the fat in the meal stored at -20°C showed less change in extract-ab i l i t y than did that in the meal stored at the higher temperature (25°C) . Astrup (1958) found that the progress of oxidation in herring meals differed depending on the temperature, when oxygen absorption was used as a measure of oxidation. The lower oxygen absorption of the meals stored at the higher temperature indicated the formation of antioxidants. Lea et a l . (1958) reported that the iodine value of the ether-extractable 112 material from herring meal dropped rapidly at the beginning of the storage period at the three storage temperatures used (37° , 25° and 10°C). They also noted that the decrease in the iodine value was slower at the higher storage temperatures up to four months of storage. They suggested that "whereas oxidation of the l i p i d probably proceeds further at lower storage temperatures, coupling of the oxidized o i l to the protein (or i t s destruc-tion) might be greater at the higher temperatures." A decline in o i l extractabi l i ty was only detected at the two higher storage temperatures. Spectrophotometry in the ul t raviolet range i s a useful technique for measuring l i p i d oxidation. Working with l i no le i c acid Frankel (1962) reported that the absorption by conjugated double bonds in hydroperoxides increases, with a maximum in the region of 231-234 nm as the fatty acid oxidizes. Several investigators (Holman, 1946, Holman and Burr 1946 and Holman and Rahm 1966) have reported that the rate of formation of con-jugated dienes correlates closely with oxygen up-take by l i n o l e i c acid. Lea (1953) likewise found close agreement between the iodometric determina-tion of peroxide value and the rate of formation of conjugated hydro-peroxides in l ino le ic acid. In the present study both techniques were used to measure the extent of oxidation in the ether-extractable fat from herring meals and close agreement was found between the two c r i t e r i a . It was also found that the meals treated with ethoxyquin (meals 6 and 7), regardless of the storage temperature, showed l i t t l e oxidation as measured by peroxide values or the ul t raviolet absorption as compared with the untreated meals (meal 4 and 5). There was not much difference in the behaviour of meal 6 and 7 with respect to either their peroxide values or 113 ult raviolet absorption (Figures 6 and 10). On the other hand, differences were apparent between meals 4 and 5. A rapid increase in the peroxide value was noted for meal 4 (untreated and stored at room temperature) unt i l i t reached a maximum at the sixteenth week of storage, after which time the peroxide value declined. For meal 5 (untreated and stored at -2<TC) , an increase in peroxide value was found, but the increase was slower than was the case with meal 4. Peroxide value was, however, maximum at the same time as with meal 4 (although lower than with meal 4), and subsequently decreased. When the oxidation of these two meals (meal 4 and 5) was measured as ultraviolet absorption in the region of 233 nm, a different picture was found (Figure 10). For meal 4, an increase in the absorbance at 233 nm was noted unt i l i t reached a maximum at 24 weeks of storage, while the peak of peroxide value was found to occur at the sixteen weeks of storage. In case of meal 5 the maximum value of u l t ra -v io let absorption at 233 nm was found to occur after eight weeks of storage while as reported above the highest peroxide value for this meal was after sixteen weeks of storage. Despite the somewhat different patterns of oxidative change indicated by the results of the two techniques i t may be concluded that, in general, the higher storage temperature accelerated oxidation of the unsaturated fatty components in the unstabilized meal. Peroxides are the f i r s t products of l i p i d oxidation. They are lab i le compounds which can undergo decomposition and further reaction with other constituents of the meal. Peroxide values therefore.always depend on the difference between the rates of formation and decomposition which in turn would be affected by factors such as the amount of oxygen available 114 for reaction and the moisture content of the meal J (Lea et: a l . 1958). Lundberg and Chipault (1947) in a study of the early autoxidation of methyl l inoleate showed that, whereas a l l of the conjugated dienes are present as peroxides, not a l l of the peroxides are present as conjugated dienes. It has been shown by Holman et d . (1954), and Lundberg et a l . (1946) that the ultraviolet absorption spectrum of an oxidized fat i n -creased as a result of peroxide decomposition by superheated steam. Privett and Blank (1962) showed that the extent of u l t raviolet absorption of autoxidized methyl l inoleate was not simply due to the formation of hydroperoxides but was the result of the entire autoxidation process. They presented evidence for the formation of nonhydroperoxidic compounds which exhibit ul t raviolet absorption during the induction period of methyl l inoleate oxidation. Poor correlation of peroxide value with the amount of oxygen absorbed in stored freeze-dried salmon was reported by Martinez and Labuza (1968) who also showed that the peroxide value did not ref lect the tota l quantities of oxygen absorbed. They explained these findings in terms of decomposition of the hydroperoxides to degradation products, absorption of oxygen by other components beside the l i p i d (Tappel, 1956), and chemical bonding of the peroxides to other components such as proteins, as found by Tappel (1955) and by El-Gharbawi (1965). The thiobarbituric acid (TBA) test has also attracted considerable attention during recent years (Dahle et: a l . 1962, King, 1963, Dunkley and Franke 1967, and Yu and Sinnhuber 1967). The pink colour formed in this test is due to condensation product between 2 molecules of TBA and 115 malonaldehyde, a product of polyunsaturated fatty acid oxidation (Kurtz et a l . 1951, Sinnhuber et a l . 1958). As shown in Figure 8 the untreated herring meals fcieals 4 and 5) showed more oxidative rancidity than the meals which had received antioxidant treatment (meals 6 and 7) regardless of the storage temperature. When the changes of the peroxide value and TBA absorbance for the unstabilized meals were compared i t was noted that both values i n -creased steadily for up to sixteen weeks of storage^ and then started to f a l l . However, the difference in the TBA values between meals 4 and 5 was small, while there was an appreciable difference in the peroxide value between meals 4 and 5, especially at sixteen weeks of storage (Fig-ures 8 and 6). Kenaston et a l . (1955) reported that during the autoxida-tion of l inoleate, TBA values correlate with peroxide content. Dahle et a l . (1962) demonstrated a l inear relationship between TBA values and oxygen uptake for tr iene, tetraene, pentaene and hexaene esters. There is also a l inear relationship between peroxide and TBA values in f ish o i l (Yu and Sinnhuber, 1967). Haase and Dunkley (1969) compared the u l t r a -v iolet absorption method and TBA test for oxidation of an aqueous sus-pension of l inoleate in the presence of either ascorbic acid or cupric ions. They found that the formation of conjugated hydroperoxides is in i t ia ted immediately after the addition of ascorbic acid as a catalyst while the absorbance in the TBA test increases only s l ight ly unt i l some time has passed.. They explained this delay in response as evidence that the TBA test measures secondary oxidation products, not hydroperoxides. It seemed that storage at -20°C favoured rapid formation of conjugated 116 hydroperoxides which later (8 weeks) degrade to TBA reactive materials. In the unstabilized herring meal stored at room temperature (meal 4), the delay in response was in the formation of conjugated hydroperoxides which did not reach a maximum unt i l after 24 weeks of storage, although the absorbance in the TBA test reached i ts maximum value after 16 weeks of storage. In the presence of cupric ions, during the early stages of methyl l inoleate oxidation, the TBA test provides a more sensitive measure of oxidation than ult raviolet absorption by the hydroperoxides (Haase and Dunkley 1969). The explanation given by Haase and Dunkley was that copper acted as a catalyst in the degradation reaction of hydroperoxides to TBA reacting materials (Emanuel and Lyaskovskaya 1967, Ingold 1962), or that copper catalyzes the TBA reaction i t s e l f (Patton and Kurtz 1955), or both. They concluded that the presence of copper or possibly other catalysts control the relationship between the TBA values and ultraviolet absorption by dienes. It i s evident from the peroxide values, TBA values, iodine values and ul t raviolet absorption by dienes (Figures 4, 6, 8 and 10) that the addition of antioxidant ethoxyquin retarded the oxidative changes in the o i l fraction of herring meal during the storage period which lasted nine months. The role of antioxidant in retarding l i p i d oxidation i s not completely understood, however, i t has been suggested recently by Roubal (1971) that i t acts as an inhibitor of trapped free radicals . Storage temperature was found to have no appreciable effect on the stabi l ized meals, but showed some effect on the meals not treated with antioxidant. 117 Available lysine declined during the storage of the different meals tested in this study. It seemed, however, that there was no difference in the rate of decline for the stabi l ized herring meals stored under different temperatures. For meals stored with no added antioxidant, the storage temperature again affected the rate of decline in the content of available lysine in that i t was faster at room temperature than at -20°C (Figure 12). The present findings seem to indicate that storing unstabilized meal at -20°C has some advantage in protecting the a v a i l -ab i l i t y of lysine in herring meal. However, such protection was not evident in the case of meals treated with antioxidant. These results are in agreement with the findings of Lea et a l . (1958, 1960) and March et a l . (1965) who reported losses of available lysine after one year of storage of Canadian herring meal. Working with Pervuvian f ish meal Burke and Maddy (1956) reported marked losses in chemically available lysine in the untreated meal while in the ethoxyquin-stabilized meals the losses were prevented. On the other hand, no difference in the percen-tage of available lysine was found by Opstvedt et a l . (1970) between stabi l ized and unstabilized herring meals during storage. These discrepancies in results could be explained in terms of differences in the extent of oxidation, storage conditions and type of antioxidant employed. In unstabilized herring meal spontaneous heating could easily develop as a result of the oxidation i t s e l f , and could affect the ava i lab i l i t y of lysine (Lea et a l . 1960) apparently by speed-ing i t s reaction with fat oxidation products. In the present storage experiment, meal 4 (unstabilized herring meal stored at room temperature) showed the greatest losses of available 118 lysine upon storage as well as the greatest oxidation of the ether-extractable fraction of the meal as measured by TBA test , peroxide value, and ul t raviolet absorption. Together, these results show that inactivation of lysine was in some way associated with the oxidation of the o i l . During the storage of f i sh meal, the nutr i t ive value of protein is decreased as a result of oxidation of the l i p i d . L ipid oxidation products such as aldehydes, peroxides and free radicals are capable of reacting with the functional groups of the protein with con-sequent destruction of amino acids or reduction in the b io logical ava i lab i l i t y of amino acids. The chemical determination of available lysine indicated that during storage some of the e-amino groups of lysine were bound to the oxidation products. A reduction in the percentage of available lysine was observed during the storage period (Figure 15) of the solvent-extracted meals (meals 8 and 9). This is in contrast to Lea et a l . (1958) who reported no decrease in available lysine in defatted herring meals. These d i f f e r -ences could be due to variations in the amount of bound fat le f t in the meal after solvent extraction. The pepsin d igest ib i l i t y decreased during the storage of meals 4, 5, 6, 7, 8, and 9 in this study, but the decrease was most pronounced for the untreated herring meal stored at room temperature (meal 4). Similar results were reported by March et a l . (1961) and Opstvedt et a l . (1970) who found that the rate of decrease in pepsin d igest ib i l i t y during storage was faster in unstabilized meal. Geisler and Contreras (1967) re-ported similar findings in a study with anchovy meals. The results of the 119 pepsin d igest ib i l i t y test also indicated that storing the unstabilized meal at -20°C somewhat inhibited the formation of non-digestible nitrogenous compounds compared with the unstabilized meal stored at room temperature. It was observed that meal 4 (unstabilized herring meal stored at room temperature) which showed the most rapid decrease of in v i tro pepsin d iges t ib i l i t y , also showed the most rapid accumulation of oxidation products throughout the entire storage period. The oxidation of unsatur-ated l ip ids produces several chemically active intermediates. These inter -mediates can react with one or more of the reactive groups of the protein (Putnam 1953). Such a reaction would subsequently lead to the formation of quite a number of complex interactions between the oxidation products of l i p i d and the protein. This type of interaction frequently results in poor d igest ib i l i t y of the protein. Cross-linked proteins, for instance, are known to resist hydrolysis by proteolytic enzymes. Oxidative deterioration of the freeze-dried herring meal prepared from the presscake,rwas,evaluated by measurement of-.the_oxldation of the l i p i d s , and changes in the protein fract ion. A rapid drop in ether-extractable materials was noted during the f i r s t eight weeks of storage in the presscake -.(meal'l), .freeze-dried meal stored under a i r (meal 2) , and the freeze-dried meal stored under nitrogen (meal 3), however, the decrease was more drastic for the freeze-dried meals (meals 2 and 3). The changes in iodine values of the ether extracts (Figure 3) were para l le l to the changes in the amount of ether-extractable material. 120 It could be concluded, therefore, that the freeze-dried herring meals favoured more rapid oxidation of the unsaturated fatty compounds, whether the meal were stored under a i r or nitrogen, than did the press-cake. . The results of the peroxide value and TBA tests are presented in Figures 5 and 7. A rapid accumulation of peroxides and TBA-reactive materials was observed in the presscake andlfreeze-dried meals, during , the f i r s t month of storage. In the case of the freeze-dried meals these reaction products subsequently disappeared. In the presscake,1 how-ever, another accumulation of peroxides and TBA-reactive materials was noted at the twenty-fourth week of storage. These results exhibit general agreement between these two c r i t e r i a for following the oxidative changes. Kopecky (1968) has also reported some correlation between peroxide content and TBA value in freeze-dried pork. Autoxidation of l ip ids is often accompanied by an increase in the spectral absorbance at 233 nm even when the amount of peroxides present are unmeasurable (Wills 1965, Maloney et a l . 1966, Thomas et a l . 1968, and Privett and Blank 1962). The formation of conjugated hydroperoxides is in i t ia ted immediately in the presscake after i t s production, (Figure 9). It seemed from the rapid accumulation of oxidation products that high moisture content (ca. 50%) and low temperature promoted rapid oxidation. This is in agreement with the data of Kamiya and Ingold (1964) who found a pro-oxidant effect of 10% water added to a metal-catalyzed model system. Lips (1949) s imilar ly reported that the addition of 12% water to lard acted as a pro-oxidant. Working with chicken fa t , Ph i l l ips 121 and Williams (1952) observed that the addition of 67% water was pro-oxidant. Heidelbaugh et. a l . (1971) concluded that water at a high level acts as pro-oxidant as i t has the ab i l i t y to solubi l ize and mobilize both the reactants and catalysts. For the freeze-dried samples (meals 2 and 3) which represent a very low-moisture system (ca. 3%), the oxidative deterioration was quite serious. Water i s known to retard l i p i d oxidation in many dehydrated and low-moisture food products, (Stevens and Thompson 1948, Marshall at a l . 1945, Matz et a l . 1955 and Martin 1958). Several explanations have been advanced to elucidate the protective effect of water in retarding l i p i d oxidation. Dean and Skirrow (1958) showed that in systems contain-ing signif icant amounts of metal catalysts, water at low concentration acted as an antioxidant by forming hydration complexes with metal catalysts. In studies by Maloney et^  a l . (1966, 1969) the role of water in retarding oxidation of l i p i d has been well demonstrated. The results of the present study may lead to the conclusion that water, at both very low and very high levels , promotes the oxidation of l i p i d s . Rockland (1957, 1969) arrived at a similar conclusion as he found that walnuts and beans tend to oxidize quickly at very low and very high moisture content. Working with a model system Heidelbaugh et^  a l . (1971) found a c r i t i c a l water content and a c r i t i c a l water act iv i ty , corresponding to the inversion of the effect of water depending upon the composition of the model system. The protective effect against l i p i d oxidation when the freeze-dried meal was stored under nitrogen was detected only when the TBA test 122 and ul t rav io let absorption by dienes were the c r i t e r i a used. It should be noted, however, that the nitrogen used was of the general (unpurified) grade which might contain traces of oxygen. This could explain the fa i lu re , according to the other c r i t e r i a , to obtain a protective effect . The rate of decline in the level of chemically determined available lysine was similar for a l l three meals (meal 1, 2, and 3) , (Figure 11). In the solvent-extracted meals (meals E and F) , the rate of change in the available lysine was slower than in the presscake and freeze-dried meals. Lea et_ a l . (1958) found that the amount of available lysine present in normal herring meal decreased during storage. On the other hand they noted l i t t l e change in the amount of available lysine in defatted meal compared to the normal one. As a consequence, they concluded that the decrease in the available lysine i s related to the oxidative changes in the o i l present in the f ish meal. The pepsin d igest ib i l i t y decreased during the storage of a l l three meals (meals 1, 2, and 3). Such a decrease in d igest ib i l i t y could be a result of oxidative deterioration during the storage period. The oxida-tion products may promote certain combinations between terminal groupings or may react with protein end groups. This results in the formation of complexes resistant to proteolytic enzymes with a subsequent reduction in pepsin d iges t ib i l i t y . The freeze-dried meal represents, indeed, a l ip id -prote in system low in moisture content. The decrease in the content of available lysine and in pepsin d igest ib i l i t y could be attributed to free radicals derived from peroxidizing l ip ids in close association with protein as the results 123 indicate considerable l i p i d oxidation. Roubal (1969) found that free radicals are produced as a result of l i p i d oxidation in the presence of protein in frozen or freeze-dried t issue. He suggested that those radicals are pr incipal ly responsible for reducing the protein quality. The interrelationship between the pattern of amino acids and peptides released by digestive enzymes on one hand and the bio logical value of food proteins on the other, has been considered by many workers. Amino acid patterns result ing from in vivo pepsin digestion reveal differences between proteins according to Sheffner et a l . (1956). Never-theless they reported that the total essential amino acid content and/or the patterns existing when the pepsin digests were further digested with trypsin and erepsin no longer revealed such differences. Any treatment of protein which severely decreases peptic d igest i -b i l i t y may be expected to reduce the nutr i t ional quality of the so-treated protein despite further hydrolysis by the action of pancreatic enzymes. It was therefore considered of interest to compare the patterns of peptide-released by pepsin action on the different herring meals under test . The separation of peptides from the controlled peptic digests was performed by two-dimensional paper chromatography and electrophoresis. No major changes in the peptide map were found among the different treat-ments nor upon storage. These results suggest that perhaps this technique s t i l l needs some modification in control l ing the pepsin digestion. Also, in order to detect various peptides, refinements have to be made to improve the sensit iv i ty of this test , so that the changes which occur in the pro-tein fraction of herring meal upon treatment and storage can be detected. 124 The model system was devised to study the effect of proteins on the rate of oxidation of o i l in the course of a long storage period. Storage was at two different temperatures, namely room temperature (21°C) and -20°C. This model system was considerably less complex than fishmeal, even though i t was not possible to eliminate a l l trace contaminants. Most previous studies using the model system approach employed higher temperatures, e . g . , 50°C and the measurements of l i p i d oxidation were conducted over a relat ively short storage period. Iodine value measurements revealed that the o i l became more satur-ated when stored in close association with albumin. This indicates that the presence of protein promotes a rapid decrease in the iodine value of the o i l (Figure 19). The rate of decrease in iodine value was s l ight ly faster in the model system stored at room temperature than at -20°C. Oxidation was also monitored by measurements of peroxide value, TBA test and ul t raviolet absorption spectra in o i l stored alone and o i l stored as part of the model system (in close association with egg albumin). An increase in the absorbance in the TBA test was noted for both o i l and model systems stored at both room temperature and -20°C unt i l the twentieth week of storage after which time the values started to decline. Both model systems showed higher absorbance at 535 nm in the TBA test than did the o i l stored alone at room temperature or -20°C, again i l l us t ra t ing the role of protein in o i l oxidation. It was also noted, in this study, that peroxides accumulated at a faster rate in the model systems than in the o i l (Figure 20). After 32 weeks, however, i t was found that the o i l stored alone exhibited greater 125 accumulation of peroxides than did the o i l stored in the model system. These results seem to indicate that the peroxides formed in the model system interacted with albumin possibly through a free radical mechanism. When the ultraviolet absorption was the cr i ter ion for degree of oxidation, the differences in the rate of oxidation between the free o i l and o i l in the model system were more apparent than when compared with TBA or peroxide values. Johnston et a l . (1961) stated that the u l t r a -v io let absorption characteristics of an o i l provide a more sensitive i n -sight into the oxidation process than does peroxide value. The temperature at which the model system was stored modified the pattern of oxidative change in the o i l component as measured by u l t r a -v io let absorption. In the model system stored at -20°C, the formation of conjugated hydroperoxides reached a peak of absorption at the twelfth week of storage and then declined, while in the model system stored at room temperature, the formation of conjugated hydroperoxides continued to bui ld up unt i l the end of the storage period. A similar pattern was also noted in the herring meals stored at room temperature and at -20°C, (Figure 10).. It may be concluded that storage at -20°C promotes faster degradation of conjugated hydroperoxides than storage at room temperature. According to the estimates of the iodine, TBA, peroxide values and the ul t raviolet absorption character ist ics, i t may be concluded that the presence of protein in close contact with o i l favoured rapid oxidation of the unsaturated fatty components of the system. Labuza et^  a l . (1969) likewise found that protein had an effect on the rate of oxidation in a model system containing 59% egg albumin and 41% l i p i d . The l i p i d used by 126 these investigators was a mixture of methyl l inoleate, methyl oleate and non-oxidizable hydrocarbon, and the tests were carried out at 50°C. A catalyt ic effect on l i p i d oxidation was shown above 40% relat ive humidity. They explained this catalysis as being a result of products produced from protein oxidation, and concluded that the protein may have a modifying effect on the l i p i d oxidation mechanism. A decrease in the chemically-determined available lysine in albumin and model system was evident during the storage period. The rate of decrease was faster for the model system than for albumin stored alone. The data indicated that l i p i d oxidation products reacted with the e-amino group of lysine in the albumin (Figure 23). Similar results were reported by Andrews et a l . (1965) who used a model system approach to study the reaction between an autoxidized l i p i d and insu l in . They found that l i p i d oxidation products react with the e-amino group of lys ine, and also with phenylalanine and glycine, the N-terminal amino groups of insu l in . Insulin which had been in close association with autoxidized methyl l inoleate before treatment with 1-fluoro-2,4-dinitrobenzene did not react with the reagent. As a result they concluded that the reaction sites in the insul in must have already reacted with the oxidation products. Oxidation of l ip id -prote in mixtures results in the formation of free radicals , Roubal et a l . (1966). These free radicals (Roubal 1969) were shown to be the major reactants leading to polymerization of proteins and the destruction of amino acids. The question whether the dominating mechanism in systems containing protein in close contact with oxidizing l i p i d i s radical or non-radical , was raised by Roubal (1970). He concluded 127 that radical attack, but not aldehyde attack, on protein, i s predominately responsible for damage to proteins and overal l losses of amino acids. Interactions among l i p i d , protein and water were also demonstrated in a model system composed of gelatin-methyl l inoleate oxidized for six days at 50°C, (Labuza et a l . 1969). It was found that under conditions of low humidity, which favoured rapid oxidation, the protein was degraded into fragments of lower molecular weight as indicated by the increase in the so lubi l i ty of the protein. On the other hand the so lubi l i ty of pro-tein decreased at high humidity as the oxidation of protein was enhanced and l i p i d oxidation was reduced. In the present experiments, a decrease in the d igest ib i l i t y of the protein of the model system by pepsin was noted. Such a decrease was more pronounced than the decrease in the d igest ib i l i t y of albumin stored alone under the same storage temperature. The difference was evident especially after 40 weeks, when the oxidation of the o i l fraction was more intense (Figure 24). The results i l l us t ra te the effect of o i l on the pepsin d igest ib i l i t y of albumin. A complex formation between the albumin and o i l oxidation products could account in part for the observed reduc-tion in the pepsin hydrolysis of the systems. El-Gharbawi and Dugan (1965) studied the so lub i l i t y of the n i t ro -genous compounds of freeze-dried raw beef during storage in cans with controlled oxygen-nitrogen atmospheres. They found a reduction in the so lubi l i ty of the or iginal ly soluble protein nitrogen after 24 weeks of storage at 80°F and 2% oxygen. They explained this decrease in the so lub i l -i ty as a result of protein denaturation and/or interaction between the protein and the l i p i d oxidation products. 128 An attempt was made to determine whether there were any changes in the pattern of soluble peptides from pepsin digestion of egg albumin stored at room temperature and at -20°C and from the model system stored at these two temperatures. Figures 25-33 represent typical peptide maps of peptic digest of albumin and the model system during the storage period. The d igest ib i l i t y of albumin with pepsin decreased gradually during storage at both temperatures (Figure 24). When the peptide maps from these digests were compared, a gradual reduction in the number and intensity of the peptide spots recovered was noted. Similar behaviour was observed in the model systems during the storage period. Albumin was rendered pepsin-resistant during storage (Figure 24). This decrease in d igest ib i l i t y in the albumin stored alone was not expected, however, i t may be due to the presence of small amounts of carbohydrates in the product used. After storage, the albumin developed a yellowish colour typical of the Maillard reaction. The protein-carbo-hydrate linkage which results from this reaction is known to render the protein resistant to in v i t ro digestion with proteolytic enzymes (Evans and Butts 1951). Because the degree of attack by pepsin was decreased, the digestion mixture would be expected to contain a high proportion of large peptides and low proportion of small peptides. The detection of the peptide spots was done by the usual method using ninhydrin reagent (Kito and Murachi 1969). However, i t has been reported by Kito and Murachi (1969) that higher peptides, cycl ic peptides, peptides whose terminal alpha amino groups are 129 acylated, and some tripeptides produce no colour with ninhydrin. Accord-ingly the only peptides which were visualized in the present work were small ones. A reduction in the number of small peptides recovered on the maps which was associated with an increase in the per cent non-digestible nitrogen may have been an indication of poor d igest ib i l i t y with pepsin after storage. Despite the fact that most of the chemical tests demonstrated greater reduction in lysine avai lab i l i ty and pepsin d igest ib i l i t y of albumin in the model system than in the albumin stored alone, there was no clear difference in the peptide maps of pepsin hydrolysates of albumin stored with or without o i l . It i s possible that with some modification in the conditions of digestion (enzyme, substrate concentration, and detection technique), the sensit iv i ty of this test might be improved. B. BIOLOGICAL TESTS Biological tests generally provide the best assessment of protein quality. Attempts were therefore made to measure the extent of inter -action between oxidized fat and protein in the herring meals studied using b io logical procedures. The tests used in the present experiments were designed to determine whether the nutr i t ional values of the protein and fat were altered during storage as was suggested from the results of the chemical determinations. Because of the limited quantity available of some of the herring meals, only six were selected for b io logical evaluation. These selected meals are indicated in Table 1. 130 The amount of l ip ids present in meals made from herring is re lat ively high and deserves consideration as i t contributes energy to a ration despite the fact that the meals are fed primarily as protein supplements. Consequently, there i s interest in studying the oxidative changes which occur in herring meal l ip ids in terms of the metabolizable energy values of the l ip ids themselves as well as any accompanying decline of protein quality. L ip id oxidation is known to occur very rapidly during the early stages of storage. The use of antioxidant in herring meals to delay oxidation has been amply studied (Meade 1956, Lea et a l . 1958, Flanzy et a l . 1962 and March et a l . 1962). The results obtained in the present investigation showed differences in the metabolizable energy values of the meals in favour of ethoxyquin-treatment when the meals were stored at room temperature. The metaboliz-able energy values of the herring meals stored for four months are given in Table 4. Both ethoxyquin treatment and storage at -20°C resulted in higher metabolizable energy values compared with unstabilized meal stored at room temperature. The s t a t i s t i c a l evaluation of the data revealed no signif icant differences (P>0.01) among stabi l ized meal stored at room temperature, stabi l ized meal stored at -20°C and unstabilized meal stored at -20°C. However, the metabolizable energy levels of these meals were s igni f icant ly higher than that of the unstabilized meal stored at room temperature. The higher metabolizable energy values observed for the antioxidant-treated herring meal can probably best be ascribed to the pro-tection of the residual o i l in the meal. In the untreated meal (stored at room temperature), the fat undergoes oxidative alteration with a consequent loss of i t s b io logical value as an energy source. 131 When the data from the metabolizable energy determination were compared with the values for the ether-extractable fat present in the meals, a close agreement was found. The untreated herring meal stored at room temperature which had the lowest metabolizable energy value also showed the fastest decrease in ether-extractable fat . It could be deduced that when the meal l ip ids become unextractable by d i -ethy l ether due to oxidation during storage they became, in some degree, physiologically unavailable. Opstvedt et a l . (1970), working on Norwegian herring meal arrived at a similar conclusion. They reported losses in metabolizable value of 153 Kcal/kg of dry matter in the untreated meal compared to the ethoxyquin-treated meal. The difference in ethyl ether extractable fat between the two meals was 20g per kg of dry matter i n d i -cating a loss of 128 Kcal of metabolizable energy per kg (assigning 6.4 Kcal/g ethyl ether extract as shown by March el: a l . 1962). This calculated difference was considered to agree with the measured difference of 153 Kcal/kg. De Groote (1968) found that ethoxyquin treatment of Peruvian f ish meal maintained i t s metabolizable energy value. They also reported more losses in d i -ethy l ether extractable fat in the untreated f ish meals compared to antioxidant-treated ones. The results of the protein quality test of herring meals stored for f ive months are given in Table 4. This test was designed to assess protein quality of herring meal when fed to supplement the protein con-tributed by the cereals present in a ration under pract ica l feeding condi-t ions. The diets were formulated in such a way that in no case was the available energy content of herring meal a factor in the response of the 132 chicks. The sensit iv i ty of the protein quality test was altered by testing the meals at different levels in the basal d iet . In the present studies the herring meals tested were used to supply 4, 7, and 11% of protein to the basal diet . The data from the protein quality test were analyzed s ta t i s t i ca l l y to investigate the effect of storage temperature, antioxidant treatment and level of protein contributed by the different herring meals. The analyses indicated that increasing the levels of supplemental protein promoted more rapid growth with more ef f ic ient feed conversion. Ant ioxi -dant treatment was found to have a s ignif icant effect on the protein quality expressed as average gain in weight. However, when the results of protein quality tests were calculated as feed eff iciency ra t io , the interaction between storage temperature and antioxidant treatment was s igni f icant . Using the gain in weight as a c r i ter ion , the unstabilized meal stored at room temperature gave the poorest growth response at a l l levels of supplementary protein. The combination of antioxidant treatment and storage at -20°C had no advantage over the treatment with antioxidant alone with respect to either energy value or growth response. When we look at the results of the metabolizable energy determina-tions and the growth response of the chicks, i t can be seen that the metabolizable energy values and the eff ic iency of u t i l i za t ion tended to be higher in diets containing herring meals treated with antioxidant whether the meal has been stored at room temperature or at -20°C. Also these values were higher for diets containing the unstabilized herring meal stored at -20°C than for the diets containing unstabilized herring meal stored at room temperature. 133 When the metabolizable energy determination and the protein quality test were run on the meals after they had been stored for nine and ten months respectively, s igni f icant ly faster growth and higher metabolizable energy values were obtained with the antioxidant-treated meals (meals A and C) and the unstabilized meal stored at -20°C (meal D) than with the unstabilized meal stored at room temperature (meal B). These results are in l ine with those of March et a l . (1965) who found that antioxidant treatment of herring meals prior to storage had a s tab i l i z ing effect on the metabolizable energy values and S.P.V. for the chicks. They could not establ ish, however, whether the effect of ant iox i -dant on S.P.V. was direct or indirect through stabi l i zat ion of the fat present in the meal. Working on Norwegian herring meal, Opstvedt et a l . (1970) showed that the protein quality of herring meal was improved by antioxidant treatment as measured by protein eff iciency rat io and net protein u t i l i za t ion . They also reported an increase in the metabolizable energy value as a result of antioxidant treatment. Marked improvements in the protein quality (Burke and Maddy 1966) and energy value (de Groote 1968) of Peruvian f ish meal have been found after ethoxyquin-stabil ization. The protective effect of the antioxidant against oxidation of the l i p i d fraction is probably the major factor in maintaining the metaboli-zable energy value of the meal during storage at room temperature. Fat undergoes oxidation in the unstabilized herring meals as demonstrated by the various chemical tests performed and hence loses i t s b io logical value as an energy source. From the results of the present metabolizable energy experiments i t appears that storing the meals at -20°C did protect their o i l content 134 against oxidative changes whether or not the meals were stabi l ized with ethoxyquin. The sl ight decrease in the metabolizable energy value of the unstabilized herring meal stored at -20°C for four or nine months compared to the value of the stabi l ized meal stored at this temperature was not s t a t i s t i c a l l y s igni f icant . The chemical characteristics of these two meals, however, indicated that the unstabilized meal underwent more oxidation than did the stabi l ized one. The ethoxyquin treatment of the meal resulted in higher protein quality as compared with unstabilized herring meal when both meals were stored at room temperature for f ive or ten months. This maintenance in the protein quality could be due to a retardation of the oxidative changes in the l i p i d fraction of the meal. Several researchers have studied the relationship between oxidative changes in the o i l present in f i sh meal and the deterioration in the protein nutr i t ive value. Tappel (1955) working on model systems found that protein interacts with the oxidizing fat leading to the formation of insoluble dark brown copolymers of high oxygen and nitrogen content. Such changes are probably accompanied by a considerable reduction in the nutr i t ive value of the protein. Damage to protein and amino acids by l i p i d oxidation products has been demon-strated in model system studies (Roubal and Tappel 1966). Biely ejt a l . (1955) studied the nutr i t ive value of herring meals including hexane-extracted ones• They found that the presence of o i l in the meals did not affect the growth response of chicks to either freshly prepared or to stored meals when the basal diet was adequately fo r t i f i ed with vitamins. Later, however, Lea et a l . (1958) reported that the 135 available lysine in herring meal was decreased during storage in a i r but did not decline when the meal was defatted or stored in nitrogen. Rand et a l . (1960) stored samples of commercial f ish meal and defatted meal for 12 months at room temperature. When the meals were fed to chicks as the only source of dietary protein, there was a decrease in protein quality in a number of the commercial meals, but not in the defatted ones. Protein quality of f i sh flours made from freshly manufactured sardine meal and from sardine meal stored for 2 months at 28-33°C were compared by Moorjani et_ a l . (1955). Using the "gross protein value" technique they found that the ab i l i t y of f ish flour made from stored sardine meal to supplement a methionine-deficient diet was lower than that of fresh meal. Available lys ine, cystine and methionine were some-what lower in f lour made from stored meal. They concluded that oxidation of l ip ids in f i sh meal is one of the factors responsible for the lower nutr i t ive value of f lour made from stored meal. Narayan et a l . (1958) showed that l i n o l e i c acid could not interact with egg albumin unless the l i no le i c acid was par t ia l l y oxidized. They also indicated that fatty acids such as o le ic acid and laur ic acid did not possess the functional groups necessary for the reaction with proteins. These functional groups however, seemed to be present in autoxidized l i n o l e i c acid . It has been suggested that aldehydes or other products produced during the oxidation of the unsaturated fatty acids could react with the amino or other reactive groups of the protein by a type of aldehyde-amine browning reaction (Tappel 1955). Results obtained by Andrews et: a l . (1965) indicated that inter -mediate products derived from autoxidized l inoleate were responsible for 136 protein alteration ( insolubi l iz ing protein v ia a cross- l inking reaction) in dry methyl l inoleate-gelat in and methyl l ino leate - insul in systems exposed to ai r at 50°C. The interaction between the degradation products of l i p i d oxidation and protein, through alkylat ion, denaturation or cross- l inkage, usually decreased the rate of proteolytic enzyme action. Lineweaver and Hoover (1941), however, reported that alkylated and denaturated proteins are hydrolyzed by proteolytic enzymes much faster than their native counterparts. Cross-linked proteins, on the other hand, are known to resist proteolytic action. Free radicals derived from the oxidation of l ip id -prote in systems have been considered by Roubal (1969) to be the main reason for protein polymerization and.destruction of amino acids. Roubal suggested that some of the free radicals produced during l i p i d oxidation may be more reactive than others and are responsible for decreasing protein quality further. Roubal et a l . (1966) and Z i r l i n and Karel (1969) suggested that oxidation of l ip id -prote in mixtures results in the formation of protein radicals by abstraction of hydrogen from protein by radicals derived from l i p i d oxida-tion products, e . g . , peroxy radicals . Roubal (1970) presented evidence that free radicals , and not aldehydes, are a major cause of protein damage. He also concluded that commonly used food antioxidants such as butylated hydroxyariisole (BHA) and butylated hydroxytoluene (BHT) seem to act as inhibitors of trapped radicals . In the present study the stabi l ized and unstabilized herring meals exhibited no s ta t i s t i ca l l y s ignif icant differences in their metabolizable energy values or their protein quality after f ive or ten months of storage 137 at -20°C. These b io logical data seemed to indicate that treatment with antioxidant had no advantageous effect when the storage temperature of the meals was -20°C. Consequently, i t appeared that both storage at -20°C and treatment with antioxidant helped in preventing losses of available energy and the protein quality of the herring meals studied. However, from the results of the chemical tests performed on the o i l fraction of these two meals, i t has shown that the addition of antioxidant to the meal retarded oxidation even when the storage temperature was -20°C. In v i t ro pepsin d igest ib i l i t y of the herring meal protein was, to a certain degree, related to the biological estimate of protein quality. The untreated herring meal stored at room temperature which had the lowest protein quality also showed the most rapid decrease in the per cent digestible protein throughout the storage period. Thurston et a l . (1960) compared pepsin d igest ib i l i t y and growth response of chicks fed a diet composed of puri f ied ingredients and supplying 20% protein provided solely by f ish meals. They concluded that there was no signif icant relationship between the two c r i t e r i a except in a case when both were extremely low in a scorched meal. The amounts of available lysine present in the herring meals (Figure 12) indicate that stabi l i zat ion with ethoxyquin and/or storage at -20°C provided protection to lysine content during the storage period. The b io logical estimates of protein quality were para l le l to the levels of available lysine as determined chemically in the herring meals. This was in agreement with the data of Carpenter and El l inger (1955) who found 138 a significant correlation between the level of available lysine and the gross protein value in herring meals. Similar relationships have been reported by other investigators (Carpenter et a l . 1957, Boyne et al 1961, and Grace and Richards 1964). No difference in chemically-available lysine was found between sta-b i l i z e d and unstabilized fish meal (Opstvedt et a l . 1970 and Carpenter e i t a l . 1963). However, Lea et a l . (1958, 1960) and March et a l . (1965) reported losses of approximately 8% of available lysine after one year of storage of Canadian herring meal. March et a l . (1965) reported that differences i n S.P.V. between the untreated and the antioxidant-treated meals were not re-lated to the amount of chemically available lysine present after storage. Their explanation was that the amount of lysine was not a limiting factor in the meals under the conditions of their biological test procedure. Lea et a l . (1958) concluded that the decrease in the amount of available lysine, present in herring meal upon storage was associated with the oxidative changes in the o i l and that herring meal stored in nitrogen or defatted meal stored i n air showed slight change in available lysine compared to meal stored in a i r . The reason for the different results obtained in various experi-ments is not clear, but such disagreements may be due to different degrees of oxidation. In addition, the level of available lysine would be expected to decline markedly i f the untreated meals are subjected to spontaneous heating (Lea et a l . 1960). Miller and Ambrose (1970) observed definite nutritive changes in fishmeal after two years of storage. They found that the growth response to the ambient-stored meals was reduced to a greater extent than that to the freezer-stored ones. When herring meal was used as the sole source 139 of protein in a purif ied d iet , March at a l . (1961) found that the meal kept at -20°C from time of processing promoted slower growth than other meals stored at 25°C. Some of the meal which was stored at 25°C were treated with antioxidant. However, they reported that a l l meals were similar in their nutr i t ive value as protein or vitamin B supplement in chick diets. The results of the present experiments showed that herring meal stored at -20°C promoted faster growth than the one which was stored at room temperature. The lack of consistancy in the results obtained in various experiments on the effects of storage temperature and antioxidant treatment on f ish meal quality could be due to differences in the nature of meal l i p i d and consequently in the course of their oxidation. D i f fe r -ences in the antioxidants used may be another possible reason. To further i l lus t ra te the effect of antioxidant treatment and storage temperature on the quality of stored herring meals, the broader patterns of protein digestion in the small intestine of chickens were studied (Figures 34-35). The amounts of insoluble nitrogen and soluble nitrogen present in the lumen of the small intestine were higher from the untreated herring meals than from those treated with antioxidant, at both storage temperatures. Zebrowska (1968) reported a larger accumu-lat ion of insoluble nitrogen in the small intestine of rats given raw soybean meal than in the intestine of those given heated soybean meal. Two possible explanations were offered. F i r s t , the difference might be due to greater amounts of mucosal debris sloughed off during the passage of food. Alternatively , the difference might be the result of slower 140 rate at which raw soybean meal dissolves in the digestive f l u i d . On the basis of the findings of de Muelenaere (1964a) that raw soybean meal was more soluble than heated meal, Zebrowska concluded that in the case of raw soybean meal most of the insoluble nitrogen appeared to be mainly sloughed off mucosal debris. De Muelenaere (1964b) similar ly reported that the amounts of insoluble protein present in the small intestine of rats fed an amino acid diet supplemented with raw trypsin inhibitor were greater than the amounts found when heated trypsin- inhibitor was fed. The fractionation of the unabsorbed soluble nitrogen in the i n -testine of birds one- and one-half hours after ingesting the unstabilized meals (meals B and D) showed relat ively larger proportions of dissolved protein, peptides and amino acids when compared to the stabi l ized ones (meals A and C), (Figures 34, 35, 40 and 41). It seems clear that with these unstabilized herring meals, the whole process of digestion was retarded as compared to the stabi l ized ones. Retardation of digestion may have been responsible for the accumulation of insoluble protein, soluble undigested protein, peptides and amino acids. Furthermore, i t was observed that antioxidant-treated meal stored at room temperature (meal A) resulted in greater amounts of peptides and free amino acids than did the meal treated with antioxidant but stored at -20°C (meal C), (Figures 36 and 42). When the two unstabilized herring meals (meals B and D) were compared i t was found that the residual amounts of peptides and free amino acids were greater from the meal stored at room tempera-ture (meal B) than from the meal stored at -20°C (meal D). These findings 141 indicate that herring meals stored at room temperature for 11 months were less rapidly digested than meals stored at -20°C. In general, the marked differences in the composition of the intest inal contents following ingestion of the various meals ref lect the variation in the rates of digestion and absorption among the differently treated meals. There was a large accumulation of unabsorbed free amino acids and shorter peptides in this study with some of the meals tested. Similarly Buraczewski et^  a l . (1967) reported higher values for unabsorbed free amino acids and peptide fractions in the intest inal lumen of rats given heated f ish protein than in rats given the untreated protein. They suggested that the accumulation of large quantities of peptides may saturate the absorption sites for amino acids. Consequently portions of the unabsorbed free amino acids pass to the large intestine and enter into the metabolism of microflora. It appears safe to assume that proteins which are digested at widely different rates may be equally e f f ic ient ly u t i l i z e d , provided, of course, that the delay in digestion is not too severe. Extremely slow digestion, however, results in loss of nitrogen in the feces as was demonstrated by Buraczewski at a l . (1967) for heat-damaged f ish proteins. It i s well known that for ef f ic ient u t i l i za t ion of the protein, amino acids released during the digestion have to appear simultaneously at the sites of synthesis (Melnick et a l . 1946). Estimates of protein quality of the stabi l ized and unstabilized herring meals stored at room temperature for 11 months, whether based upon their protein quality (Tables 4 and 5) , rate of digestion in the small 142 i n t e s t i n e (Tables 8 and 9), the amount of protein s o l u b i l i z e d by pepsin (Figure 14), or the amount of available lysine (Figure 12) indicated that antioxidant-treated meals are superior i n protein n u t r i t i v e value to s i m i l a r but untreated meals. For the s t a b i l i z e d and unstabilized herring meals stored at -20°C for 11 months, the effect of antioxidant on quality was not so c l e a r l y evident by either the b i o l o g i c a l or the chemical measurements employed. This indicated that storage at -20°C by i t s e l f had a protective effect on the n u t r i t i v e quality of the herring meals. B i o l o g i c a l studies were also conducted on the solvent-extracted meals (meals E and F). After four or nine months of storage, a marked difference i n the metabolizable energy values was observed between these two meals. Meal F, which was stored at -20°C, had a s i g n i f i c a n t l y higher metabolizable energy value than meal E which was stored at room temper-ature. There was no s i g n i f i c a n t difference between protein quality of the meals after storage for f i v e months. However, when the meals were tested for t h e i r protein quality after ten months, the meal which was stored at -20°C had a s i g n i f i c a n t l y higher e f f i c i e n c y of feed u t i l i z a t i o n than the meal stored at room temperature (Table 5). The results obtained from the i n vivo d i g e s t i b i l i t y test (Table 8 and Figures 38 and 44) are i n accord with the metabolizable energy and protein quality observations for the solvent-extracted meals stored for 11 months. Meal E (stored at room temperature) was poorly digested, as indicated by the higher accumulation of soluble nitrogen i n the lumen of the small intestine as compared to meal F (stored at -20°C). i 143 When the protein quality was assessed by two in v i t ro procedures, namely, the amount of protein solubi l ized by pepsin (Figure 16) and the amount of available lysine (Figure 15), insignif icant differences between meal E and meal F were observed during the storage period. Nevertheless, the results of the different b io logical tests indicate that storage of the solvent-extracted meal at room temperature caused more losses of available energy and reduced d igest ib i l i t y to a greater degree than did storage at -20°C. The decrease in the metabolizable energy values and the protein quality of these solvent-extracted meals upon storage (Tables 4 and 5) , contradict the finding of Lea et a l . (1958) who indicated that fat - f ree meals do not change during storage. They compared the oxygen absorption of normal and solvent-extracted herring meal and reported that fat - f ree meal absorbed oxygen at a very low rate. Fat which re -mained in the herring meal after solvent extraction might have been a factor responsible for the reduction in quality noted in the present study. Tappel (1956), however, reported that in dehydrated foods, oxidation of the protein fraction may occur to a greater extent than expected. He also noted that oxygen l a b i l i t y is a general characterist ic of proteins. Storage experiments by Rand et a l . (1960) revealed a decrease in the protein quality of several commercial meals, but not defatted meals when both were fed to chicks as the only source of dietary protein. Moorjani et a l . (1965) concluded that sulfur-containing amino acids became par t ia l l y unavailable in f ish f lour made from sardine meal which had been stored at 28-33°C for two months. They explained these findings as a 144 result of oxidation of the l ip ids during storage. It appears from the b io logical data obtained for the solvent-extracted meals (meals E and F ) , that storage at -20°C had a favourable effect on the nutr i t ive quality of the meal as compared to storage at room temperature. This behaviour would be expected according to the Arrhenius concep t. 6. CONCLUSIONS From the chemical examinations performed in the present study i t could be concluded that (1) the presence of protein in close association with fat promoted rapid oxidation in the fat fraction compared to the rate of oxidation in fat stored alone at the two storage temperatures used, namely 21°C and -20°C. Biological tests showed that meal treated with antioxidant and meal stored at -20°C were superior in nutr i t ive quality to similar but untreated herring meal stored at 21°C. There was a general agreement between the results of the b io -log ica l tests and those of the chemical tests. The results from both types of test indicated that the unstabilized herring meal stored at 21°C was infer ior in nutr i t ive value to the other meals. The d iges t ib i l i t y , the supplementary protein quality and the metabolizable energy value of the solvent-extracted meal stored at -20°C were superior to similar meal stored at 21°C. Oxidative changes occurred in both undried presscake stored at -20 C and fre eze-dried meals stored at 21 C. 145 REFERENCES Almquist, H. J. 1956. Changes i n fat e x t r a c t a b i l i t y and protein digest-i b i l i t y i n f i s h meal during storage. J . Agr. Food Chem. 4 : 638-639. Anderson, M. L., F. J. King and M. A. Steinberg. 1963. Effect of l i n o -l e n i c , l i n o l e i c and o l e i c acids on measuring protein e x t r a c t a b i l i t y from cod sk e l e t a l muscle with the s o l u b i l i t y test. J . Food S c i . 28 : 286-288. , and M. A. Steinberg. 1964. Effect of l i p i d content on protein-sodium linolenate interaction i n f i s h muscle homogenate. J. Food Sc i . 29 : 327-330. , , and F. J . King. 1965. The Technology of Fish U t i l -i z a t i o n , ed. R. Kreuza. pp. 105-110. Fishing News (Books) Ltd. London. Andrews, F., J. Bjorksten and F. B. Trenk. 1965. The reaction of an autoxidized l i p i d with proteins. J . Amer. O i l Chem. Soc. 42 : 779-781. Andrews, P. 1964. Estimation of the molecular weights of proteins by Sephadex g e l - f i l t r a t i o n . Biochem. J. 91 : 222-233. Anonymous. 1958. Oxidation research shows reasons for quality deterior-ation of certain foods. Comm. Fisheries Rev. 20 : (51). 18. A.O.A.C. 1960. O f f i c i a l methods of analyses of the Association of O f f i c i a l A g r i c u l t u r a l Chemists. 9th ed. Astrup, H. 1958. Acta Polytechnica (Stockholm) 242 : 58. Cited after March et a l . 1961. Aure, L. 1957. Oxidation s t a b i l i z a t i o n of herring meal by BHT on a technical scale. Aarsberetn. Norges Fiskeraer 3 : 17-24. Fish. Res. Bd. (Canada) Transl. Ser. No. 214 : 1-9. Backstrom, H. L. J. 1927. The chain reaction theory of negative c a t a l y s i s . J . Amer. Chem. Soc. 49 : 1460-1472. Baliga, B. P. and C. M. Lyman. 1957. Preliminary report on the n u t r i -t i o n a l significance of bound gossypol i n cottonseed meal. J . Amer. O i l Chem. Soc. 34 : 21-24. 146 147 Bender, A. E . , and S. Haizelden. 1957. Biological value of the proteins of a variety of f ish meals. B r i t . J . Nutr. 11 : 42-43. , D. S. M i l l e r , and E. J . Tunnah. 1953. The biological value of f i sh meals. Proc. Nutr. Soc. 12 : i i . B ickel , A. F . , and E. C. Kooyman. 1956. Alkylperoxy-radicals. Part II-Kinetics of autoxidation retarded by 2 : 4 : 6 trialkylphenols. J . Chem. Soc. 1956 : 2215-2221. Bie ly , J . , B. E. March, and H. L. A. Tarr. 1951. The nutr i t ive value of f ish meal and condensed f ish solubles. I l l Effect of heating fat-containing and hexane-extracted meal. ' Progr. Rept. Paci f ic Coast Sta. F ish . Res. Bd. (Canada) No. 89 : 79-81. , , . 1955. The nutr i t ive value of herring meals, 3-The effect of heat treatment and storage temperature as related to o i l content. Poultry Sc i . 34 : 1274-1279. Bissett , M. H. , and H. L. A. Tarr. 1954. The nutr i t ive value of herring meals. I I -Effect of heat on avai lab i l i ty of essential amino acids. Poultry Sc i . 33 : 250-254. Bl igh, E. G . , and M. A. Scott. 1966. Lipids of cod muscle and the effect of frozen storage. J . F ish. Res. Bd. (Canada) 23 : 1025-1036. Boge, G. 1960. Amino-acid composition of herring (Clupea Harengus) and herring meal. Destruction of amino acids during processing. J . Sc i . Food Agr. 11 : 362-365. Bolland, J . L . , and G. Gee. 1946. Kinetic studies in the chemistry of rubber and related materials. II-The kinetics of oxidation of unconjugated o lef ins . Trans. Faraday Soc. 42 : 236-243. , and P. tenHave. 1947. Kinetic studies in the chemistry of rubber and related materials. IV-The inhibitory effect of hydro-quinone on the thermal oxidation of ethyl l inoleate. Trans. Faraday Soc. 43 : 201-210. Boozer, C. E . , G. S. Hammond, C. E. Hamilton, and J . N. Sen. 1955. Air oxidation of hydrocarbons. II-The stoichiometry and fate of inhibitors in benzene and chlorobenzene. J . Amer. Chem. Soc. 77 : 3233-3237. Bosum, I., and B. Ganrot. 1969. L ip id hydrolysis in frozen Balt ic herring. J . Food Sc i . 34 : 13-18. Boyne, A. W., K. J . Carpenter, and A. A. Woodham. 1961. Progress report on an assessment of laboratory procedures suggested as indicators of protein quality in feeding stuf fs . J . Sc i . Food Agr. 12 : 832-848. B u l l , H. B., and K. Breese. 1967a. Binding of fatty acids by proteins. Arch. Biochem. Biophys. 120 : 303-308. 148 B u l l , H. B., and K. Breese. 1967b. Denaturation of proteins by fatty acids. Arch. Biochem. Biophys. 120 : 309-315. Buraczewski, S . , L. Buraczwska, and J . E. Ford. 1967. The influence of heating of f i sh proteins on the course,of their digestion. Acta Biochimica Polonica 14 : 121-131. Biirke, R. P . , and K. H. Maddy. 1966. Verbesserte Fishmehlqualitat durch stabi l is ierung mit Santaquin. 2-Einfluss der stabi l is iarung auf den Futterungswert von Fishmehl. Kraftfutter 49 : 66-74. Cited after Opstvedt; et a l . 1970. Buttkus, H. 1967. The reaction of myosin with malonaldehyde. J . Food Sc i . 32 : 432-434. Carpenter, K. J . 1960. The estimation of the available lysine in animal-protein foods. Biochem. J . 77 : 604-610. and G. M. E l l inger . 1955. Protein quality and available lysine in animal products. Poultry Sc i . 34 : 1451-1452. , , I Munro and E. J . Rolfe. 1957. Fish products as protein supplements to cereals. Br i t . J . Nutr. 11 : 162-172. , C. H. Lea, and L. J . Parr. 1963. Chemical and nutr i t ional changes in stored herring meal. 4-Nutrit ional significance of ox i -dation of the o i l . B r i t . J . Nutr. 17 : 151-169. Caste l l , C. H. , B. A. Moore, P. M. Jangard, and W. E. Neal. 1966. Oxidative rancidity in frozen stored cod f i l l e t s . J . F ish. Res. Bd. (Canada) 23 : 1385-1401. Chargaff, E. 1944. Lipoproteins. Advan. Protein Chem. 1 : 1-24. Chio, K. S . , .and A. L. Tappel. 1969a. Inactivation of ribonuclease and other enzymes by peroxidizing l ip ids and by malonaldehyde. Biochemistry 8 : 2827-2832. . 1969b. Synthesis and characterization of the fluorescent products derived from malonaldehyde and amino acids. Biochemistry 8 : 2821-2827. Claggett, F. G. 1968. Fish meal p i lo t plant for research. F ish. Res. Bd. (Canada), Vancouver Lab. Circular No. 39 : 14-16. Cooper, H. R., and H. W. Melv i l le . 1951. The kinetics of the autoxi-dation of n-decanal. Part I. The mechanism of reaction. J . Chem. Soc. 1951 : 1984-1993. Crank, J . , .and G. S. Park (ed.). 1968. Diffusion in Polymers. Academic Press, London. 149 Crawford, D., T. C. Yu, and R. 0. Sinnhuber. 1967. Reaction of malon-aldehtde with protein. J . Food Sc i . 32 : 332-335. Dahle, L. K., E. G. H i l l , and R. T. Holman. 1962. The thiobarbituric acid reaction and the autoxidations of polyunsaturated f a t t y acids methyl esters. Arch. Biochem. Biophys. 98 : 253-261. Davies, D. S., H. L. Goldsmith, A. K. Cupta, and G. R. Lester. 1956. Radical-capture agents i n t e t r a l i n e . Measurement of r e l a t i v e e f f i c i e n c i e s and correlation with structure. J . Chem. Soc. 1956 : 4926-4933. Dean, M. H., and G. Skirrow. 1958. The cobalt-catalyzed decomposition of t e t r a butyl hydroperoxide. Trans. Faraday Soc. 54 : 849-862. De Groote, G. 1968. Energetic evaluation of unstabilized and s t a b i l i z e d f i s h meals i n terms of metabolizable energy and net energy for main-tenance and growth. Feedstuffs 40 : 26-27, 54-61. de Muelenaeve, H. J . H. 1964a. Effect of heat treatment on the haemag-glutinating a c t i v i t y of legumes. Nature, London. 201 : 1029-1030. . 1964b. Studies on the digestion of soybeans. J . Nutr. 82 : 197-205. Dervichian, D. G. 1949. Structural aspect of lipo-protein association. Disc. Faraday Soc. 6 : 7-21. Desai, I. D., and A. L. Tappel. 1963. Damage to proteins by peroxi-dized l i p i d s . J . L i p i d Res. 4 : 204-207. Detorin, G. A., L. B. Gorbachova, and A. N. Bakh. 1953. Biokhimiya (Moscow) 18 : 618-625. Cited after Narayan et a l . 1958. Dunkley, W. L., and A. A. Franke. 1967. Evaluating s u s c e p t i b i l i t y of milk to oxidized f l a v o r . J . Dairy S c i . 50 : 1-9. Dyer, W. J . , and D. I. Fraser. 1959. Proteins i n f i s h muscle. 13-L i p i d hydrolysis. J . Fish. Res. Bd. (Canada) 16 : 43-52. , and M. L. Morton. 1956. Storage of frozen plaice f i l l e t s . J. Fish. Res. Bd. (Canada) 13 : 129-134. El-Gharbawi, M., and L. R. Dugan. 1965. S t a b i l i t y of nitrogenous com-pounds during storage of freeze-dried raw beef. J . Food S c i . 30 : 817-822. Emanuel, N. M., and Y. N. Lyaskovskaya. 1967. The i n h i b i t i o n of fat oxidation processes. Pergamon Press, Oxford. 150 Evans, R. J . , and H. A. Butts. 1951. Heat in a c t i v a t i o n of the basic amino acids and tryptophan. Food Res. 16 : 415-421. Farmar, E. H. 1946. Peroxidation i n r e l a t i o n to o l e f i n i c structure. Trans. Faraday Soc. 42 : 228-236. , B. A. Sundralingam, and D. A. Sutton. 1942. The course and mechanism of autoxidation reactions i n o l e f i n i c and p o l y o l e f i n i c substances, including rubber. Trans. Faraday Soc. 38 : 348-356. , and B. A. Sundralingam. 1942. The course of autoxidation reactions i n polyisoprenes and a l l i e d compounds. Part I , The struc-ture and reactive tendencies of the peroxides of simple o l e f i n s . J . Chem. Soc. 1942 : 121-139. Flanzy, J . , G. Rocquilin and A. Pinet. 1962. Mesure de l'absorption d'oxygene par les jarines de poisson. Application a leur s t a b i l i -zation par les antioxyge*nes. Ann. Zootech. 11 : 263-272. Ford, J . E. 1965. A microbiological method for assessing the n u t r i -t i o n a l value of proteins. 4-Analysis of enzymatically digested food proteins by Sephadex-gel f i l t r a t i o n . B r i t . J . Nutr. 19 : 277-293. Frankel, E. N. 1962. Mechanisms and products of l i p i d oxidation; Hydroperoxides. In "Lipids and Their Oxidation, Symposium on Foods," ed. H. W. Shultz. Avi Publ. Westport, Conn. Geisler, H. W., and E. Contreras. 1967. The nature of heat damage i n anchovy meals by fat oxidation. Fish. News Intern. 6 : 38-40. Grace, N. D., and E. L. Richards. 1964. The n u t r i t i v e value of meat by-products. J . S c i . Food Agr. 15 : 711-/16. Haase, G., and W. L. Dunkley. 1969. Ascorbic acid and copper i n l i n o -leate oxidation. 1-Measurement of oxidation by U.V. spectrophoto-metry and the TBA test. J . L i p i d Res. 10 : 555-560. Halton, P., and E. A. Fisher. 1937. Studies on the storage of wheat f l o u r . II-The adsorption of oxygen by fl o u r stored under various conditions. Cereal Chem. 14 : 267-292. Harle, 0. L., and J . R. Thomas. 1957. Detection of free r a d i c a l i n t e r -mediates i n the action of oxidation i n h i b i t o r s . J . Amer. Chem. Soc. 79 : 2973-2974. Harman, D., and L. H. P i e t t e . 1966. Free r a d i c a l theory of aging : Free r a d i c a l reactions i n serum. J . Gernotology 21 : 560-565. Heaton, F. W., and M. U r i . 1961. The aerobic saturation of unsaturated f a t t y acids and their esters : Cobalt stearate-catalyzed oxidation of l i n o l e i c acid. J . L i p i d Res. 2 : 152-160. 151 Heidelbaugh, N. D., C. P. Yeh, and M. Karel. 1971. Effect of model system composition on autoxidation of methyl l i n o l e a t e . J . Agr. Food Chem. 19 : 140-142. H i l l , F. W., and D. L. Anderson. 1958. Comparison of metabolizable energy determinations with growing chicks. J . Nutr. 64 : 587-603. Holman, R. T. 1946. Spectrophotometric studies of the oxidation of fat s . Vll-Oxygen absorption and chromophore production i n lipoxidase-oxidized fa t t y esters. Arch. Biochem. 10 : 519-529. , W. 0. Lundberg, W. M. Lauer, and G. 0. Burr. 1945. Spectro-photometric studies of the oxidation of f a t s . I-01eic acid, ethyl oleate and e l a i d i c acid. J . Amer. Chem. Soc. 67 : 1285-1293. , and J . J . Burr. 1946. Spectrophotometric studies of the oxidation of fa t s . Vl-Oxygen absorption'and chromophore production i n f a t t y esters. J . Amer. Chem. Soc. 68 : 562-566. , and J . J . Rahm. 1966. Analysis and characterization of poly-unsaturated f a t t y acids. In "Progress i n the Chemistry of Fats and other L i p i d s , " ed. R. T. Holman. Pergamon Press, New York 9 : 15-90. Johnston, A. E., K. T. Z i l c h , E. Selke, and D. H. Dutton. 1961. Analysis of f a t t y acid oxidation products by countercurrent d i s t r i b u t i o n methods. V-Low-temperature decomposition of methyl l i n o l e a t e hydro-peroxide. J . Amer. O i l Chem. Soc. 38 : 367-371. Kamiya, Y., and K. U. Ingold. 1964. The metal-catalyzed autoxidation of t e t r a l i n . I l l - C a t a l y s i s by manganese, copper, n i c k e l and iron. Can. J . Chem. 42 : 1027-1043. Karel, M. T., T. P. Labuza, and J . F. Maloney. 1967. Chemical changes i n freeze-dried foods and model systems. Cryobiology 3 : 288-296. Katz, A. M., W. J. Dreyer, and C. B. Anfinsen. 1959. Peptide separation by two-dimensional chromatography and electrophoresis. J . B i o l . Chem. 234 : 2897-2899. Keeney, M. 1962. Mechanisms and products of l i p i d oxidation. Secondary degradation products. In "Lipids and Their Oxidation, Symposium on Foods," ed. H. W. Shultz. Avi Publ. Westport, Conn. Kenaston, C. B., K. M. Wilbur, A. Ottolenghi, and F. Bernheim. 1955. Comparison of methods for determining fat t y acid oxidation products by u l t r a - v i o l e t i r r a d i a t i o n . J . Amer. O i l Chem. Soc. 32 : 33-35. K i t o , K., and T. Murachi. 1969. Valylprolylarginine : a ninhydrin nega-tive tripeptide. J . Chromatog. 44 : 205-206. 152 King, R. L. 1963. Oxidation of milk fat globule membrane materials. II-Relation of ascorbic acid and membrane cocentrations. J . Dairy S c i . 46 : 267-274. Kn i p f e l , J . E., H. G. Botting, and J . M. McLaughan. 1970. Fish and other marine products. Assessment of the n u t r i t i v e value of heated f i s h protein concentrate (FPC) by simple i n v i t r o pepsin digestion procedure. J . Assoc. O f f i c . Anal. Chem. 53 : 964-967. Koch, R. B. 1962. Dehydrated foods and model systems. In "Lipids and Their Oxidation, Symposium on Foods," ed. H. W. Shultz. Avi Publ. Westport, Conn. Kopecky, A. 1968. Nahrung 12 : 87. Cited after Chipault, J . R., and J. M. Hawkins. 1971. L i p i d autoxidation i n freeze-dried meats. J. Agr. Food Chem. 19 : 495-499. Krutz, G. W., E. F. P r i c e , and S. Patton. 1951. Studies on 2-thiobar-b i t u r i c acid as a reagent for detecting milk fat oxidation. J . Dairy S c i . 34 : 484. Kwon, T. W., D. B. Menzel, and H. S. Olcott. 1965. Reactivity of malonaldehyde with food constituents. J . Food S c i . 30 : 808-813. Labuza, T. P. 1968. Sorption phenomena i n foods. Food Technol. 22 : 263-272. , N.'.D. Heidelbaugh, M. S i l v e r , and M. Karel. 1971. Oxidation at intermediate moisture contents. J . Amer. O i l Chem. Soc. 48 : 86-90. , J . F. Maloney, and M. Karel. 1966. Autoxidation of methyl lino l e a t e i n freeze-dried model systems. I l - E f f e c t of water on cobalt-catalyzed oxidation. J . Food S c i . 31 : 885-891. , H. Tsuyuki, and M. Karel. 1969. Kinetics of l i n o l e a t e o x i -dation i n model systems. J . Amer. Chem. Soc. 46 : 409-416. Laksesvela, B. 1958. Protein value and amino acid balance of condensed herring solubles and spontaneously heated herring meal, chick experiments. J . Agr. S c i . 51 : 164-176. . 1961. T r i a l s with rancid herring meal f a t . Meld. SSF : 7-9. Cited after Nutr. Abstr. and Rev. 31 : 1434. , and A. T. Aga. 1965. Nut r i t i v e value and a n a l y t i c a l charac-t e r i s t i c s of new and up to 12 year old herring meals. J . S c i . Food Agr. 16 : 743-749. Lea, C. H. 1931. The effect of l i g h t on the oxidation of f a t s . Proc. Roy. Soc. B 108 : 175-189. 153 Lea, C. H. 1953. Recent developments i n the study of"oxidative deter-i o r a t i o n of l i p i d s . Chem. & Ind. (London) 49 : 1303-1309. . 1958. Chemical changes i n the preparation and storage of dehydrated foods. In "Fundamental Aspects of the Dehydration of Foodstuffs." Soc. Chem. Ind. (London). , L. J. Parr, and K. J . Carpenter. 1958. Chemical and n u t r i t i o n a l changes i n stored herring meal. B r i t . J . Nutr. 12 : 297-312. , , . 1960. Chemical and n u t r i t i o n a l changes i n stored herring meal. 2. B r i t . J . Nutr. 14 : 91-113. Lineweaver, H., and S. R. Hoover. 1941. A comparison of the action of c r y s t a l l i n e papain on native and urea-denaturated proteins. J . B i o l . Chem. 137 : 325-335. Li p s , H. J . 1949. The effect of added water and antioxidants on the keeping quality of la r d . Can. J . Res. 27 (Sec. F, 10) : 373-381. Lovern, J . A., and J . Olley. 1962. Inh i b i t i o n and promotion of post-mortem l i p i d hydrolysis i n the fle s h of f i s h . J . Food S c i . 27 : 551-559. Lundberg, W. 0. 1962. Mechanisms of l i p i d oxidation. In "Lipids and Their Oxidation, Symposium on Foods," ed. H. W. Shultz. Avi Publ. Westport, Conn. . 1967. General deterioration reactions. In "Fish O i l s ; t h e i r Chemistry, Technology, S t a b i l i t y , N u t r i t i o n a l Properties and Uses," ed. M. E. Stanby. pp. 141-147. Avi Publ. Co. Westport, Conn. , and J . R. Chipault. 1947. The oxidation of methyl lin o l e a t e at various temperatures. J . Amer. Chem. Soc. 69 : 833-836. , R. T. Holman, and J . J . Burr. 1946. Spectroscopic changes i n fats during r a n c i d i f i c a t i o n . O i l and Soap. 23 : 10-14. Lundgren, H. P., D. W. Elam, and R. A. O'Connell. 1943. Electrophoretic study of the action of alkyl-benzenesulfonate detergents of egg albumin. J . B i o l . Chem. 149 : 183-193. Maloney, J. F., T. P. Labuza, D. H. Wallace, and M. Karel. 1966. Autoxidation of methyl l i n o l e a t e i n freeze-dried model systems. I-Effect of water on the autocatalyzed oxidation. J . Food S c i . 31 : 878-884. March, B. E., J. B i e l y , F. Claggett, and H. L. A. Tarr. 1962. N u t r i -t i o n a l and chemical changes i n the l i p i d f r a c t i o n of herring meals with and without antioxidant treatment. Poultry S c i . 41 : 873-880. 154 March, B. E., J. Bi e l y , C. Goudie, F. E. Claggett, and H. L. A. Tarr. 1961. The effect of storage temperature and antioxidant treatment on the chemical and n u t r i t i v e characteristics of herring meal. J. Amer. Chem. Soc. 38 : 80-84. , , and H. L. A. Tarr. 1957. The n u t r i t i v e value of f i s h meal and condensed f i s h solubles. XI-The protein n u t r i t i v e value of f i s h meals of low and normal fat content. Progr. Rept. P a c i f i c Coast Sta. Nanaimo and Vancouver No. 108 : 24026. , , , and F. E. Claggett. 1965. The effect of antioxidant treatment on the metabolizable energy and protein value of herring meal. Poultry S c i . 44 : 679-685. Marshall, J . B., G. A. Grant, and W. H. White. 1945. Ration b i s c u i t s . Effect of moisture content on keeping quality. Can. J. Res. (Sec. F) 23 : 286-294. Martin, H. F. 1958. Factors i n the development of oxidative rancidity i n ready to eat oat-flakes. J . S c i . Food Agr. 9 : 817-824. Martinez, F., and T. P. Labuza. 1968. Rate of deterioration of freeze-dried salmon as a function of r e l a t i v e humidity. J . Food S c i . 33 : 241-247. Matz, S., C. S. McWilliams, R. A. Larsen, J . H. M i t c h e l l , M. McMullen, and B. Layman. 1955. The effect of variations i n moisture content on the storage deterioration rate i n cake mixes. Food Technol. 9 : 276-285. Meade, T. L. 1956. A new development i n f i s h meal processing. Feed-stuffs 28 : 15-22. Melnick, D., B. L. Oser, and S. Weiss. 1946. Rate of enzyme digestion of proteins as a factor i n n u t r i t i o n . Science, 103 : 326-329. M i l l e r , D. L., and M. E. Ambrose. 1970. Effect of storage on f i s h meal quality as evaluated by chemical indices and chick bioassay. Poultry S c i . 49 : 1005-1010. M i l l e r , D. S. 1955. The n u t r i t i v e value of f i s h proteins. Chem. & Ind. (London) 18 : 501-502. . 1956. The n u t r i t i v e value of f i s h proteins. J . S c i . Food Agr. 7 : 337-343. Moore, N. and W. H. Stein. 1954. A modified ninhydrin reagent for the photometric determination of amino acids and related compounds. J. B i o l . Chem. 211 : 907-913. 155 Moorjani, N., L. Lahiry, R. B. Nair, A. N. Upadhye, and S. V. Rao. 1965. Nu t r i t i o n a l value of f i s h f l o u r . I-Effect of storage of sardine meal prior to i t s extraction with ethanol. Food Technol. 19 : 110-113. Narayan, K. A., and F. A. Kummerow. 1958. Oxidized f a t t y acid-protein complexes. J. Amer. O i l Chem. Soc. 35 : 52-56. , . 1963. Factors influencing the formation of complexes between oxidized l i p i d s and proteins. J . Amer. Chem. Soc. 40 : 339-342. , M. Sugai, and F. A. Kummerow. 1964. Complex formation between oxidized l i p i d s and egg albumin. J . Amer. O i l Chem. Soc. 41 : 254-259. Nishida, T., and F. A. Kummerow. 1960. Interaction of serum l i p o -proteins with hydroperoxide of methyl l i n o l e a t e . J . L i p i d Res. 1 : 450-458. Njaa, L. R., F. Utne, and 0. R. Braekkan. 1966. Effect of BHT ( d i - t e r t -butyl p-hydroxy toluene) on the value of herring meal for the young rat. F i s k e r i d i r . Skr. Tekn. Undersok 5 (2) : 1-11. Cited after Nutr. Abstr. and Rev. 37 : 6345. (1967). Olcott, H. S., and D. K. Mecham. 1947. Characterization of wheat gluten. I - P r o t e i n - l i p i d complex formation during doughing of fl o u r s . Lipoprotein nature of the glutenin f r a c t i o n . Cereal Chem. 24 : 407-414. Olley, J . , and W. R. H. Duncan. 1965. Lipids andprotein denaturation i n f i s h muscle. J. S c i . Food Agr. 16 : 99-104. , and J. A. Lovern. 1960. Phospholipid hydrolysis i n cod flesh stored at various temperatures. J . S c i . Food Agr. 11 : 644-652. , R. P i r i e , and H. Watson. 1962. Lipase and phospholipase a c t i v i t y i n f i s h s k e l e t a l muscle and i t s relationship to protein denaturation. J . S c i . Food Agr. 13 : 501-515. Oncley, J . L. 1954. Lipoproteins of human plasma. The Harvey Lectures, pp. 71-91. Opstvedt, J.,.0. Olsen, and N. Urdahl. 1970. Influence of residual l i p i d s on the n u t r i t i v e value of f i s h meal. I-The effects of adding 1,2-dihydro-6-ethoxy-2,2,4-trimethylquinoline to f i s h meal on protein q u a l i t y , energy value and vitamin E status i n chickens. Acta Agriculturae Scandnavica 20 : 174-184. Patton, S., and G. W. Kurtz. 1955. A note on the thiobarbituric acid test for milk l i p i d oxidation. J . Dairy S c i . 38 : 901. 156 P h i l l i p s , H. J . , and I. L. Williams. 1952. Some factors affecting sta-b i l i t y of chicken f a t . Food Technol. 6 : 74-76. P r i v e t t , 0. S., and M. L. Black. 1962. The i n i t i a l stages of autoxida-t i o n . J . Amer. Chem. O i l Soc. 39 : 465-469. Putnam, F. W. 1953. The chemical modification of proteins. In "The Proteins, Chemistry, B i o l o g i c a l A c t i v i t y and Methods," eds. H. Neurath and K. Bailey. Vol. IB, pp. 893-972. Academic Press, New York. Rand, N. T., V. K. C o l l i n s , D. S. Varner, and T. D. Mosser. 1960. Bi o l o g i c a l evaluation of the factors affecting the protein quality of f i s h meals. Poultry S c i . 39 : 45-53. Rao, D. R., P. V. Kamasastri, and A. N. Bose. 1965. Effects of different types of drying on the n u t r i t i v e value of proteins i n the f i s h meal. Fishery Technol. 2 : 200-203. Recknagel, R. 0., and A. K. Goshal. 1966. Quantitave estimation of peroxidative degeneration of rat l i v e r microsomal and mitochondrial l i p i d s after carbon tetrachloride poisoning. Exptl. Mol. Pathol. 5 : 413-426. Rockland, L. B. 1957. A new treatment of hygroscopic e q u i l i b r i a : Application to walnut and other foods. Food Res. 22 : 604-628. • 1969. Water a c t i v i t y and storage s t a b i l i t y . Food Technol. 23 : 1241-1251. Romoser, G. L., W. A. Dudley, and R. R. Burke. "1968. Autoxidation of f i s h meal. Fish. News Intern. 7, No. 1 : 27-29. Roubal, W. T. 1967. Oxidation deterioration of fle s h l i p i d s of P a c i f i c cod (Gadus macrocephalus). J . Amer. O i l Chem. Soc. 44 : 325-327. . 1969. Trapped radicals i n dry l i p i d - p r o t e i n systems under-going oxidation. J . Amer. O i l Chem. Soc. 47 : 141-144. . 1970. Free r a d i c a l s , malonaldehyde and protein damage i n l i p i d - p r o t e i n systems. Lipids 6 : 62-64. , and A. L. Tappel. 1966a. Damage to proteins, enzymes and amino acids by peroxidizing l i p i d s . Arch. Biochem. Biophys. 113 : 5-8. , . 1966b. Polymerization of proteins induced by free r a d i c a l l i p i d peroxidation. Arch. Biochem. Biophys. 113 : 150-155. 157 Roubal, W. T., and A. L. Tappel. 1967. Damage"to ATP by peroxidizing l i p i d s . Acta Biochem. Biophys. 136 : 402-403. Salwin, J . 1959. Defining minimum moisture contents for dehydrated foods. Food Technol. 13 : 594-595. Sheffner, A. L. , G. A. Eckfeldt, and H. Spector. 1956. The pepsin-digest-residue (PDR) amino acid index of net protein u t i l i z a t i o n . J . Nutr. 60 : 105-120. Sinnhuber, R. 0., T. C. Yu, and Te Chang Yu. 1958. Characterization of the red pigment formed i n the 2-thiobarbituric acid determination of oxidative r a n c i d i t y . Food Res. 23 : 626-634. Soon, I. I., C. H. Min, and H. Y. I I . 1970. Comparative studies on disk electrophoretic analysis of f i s h muscle proteins i n stock patterns. Korean J . Nutr. 3 : 107-112. Cited after Nutr. Abstr. and Rev. 41 : 2389, (1971). Stansby, M. E. 1948. Report on fat i n f i s h meal. J . Assoc. O f f i c . Agr. Chem. 31 : 606-610. , and W. Clegg. 1955. Determination of o i l i n f i s h meal. Comm. Fisheries Rev. 17 : 16-19. Tappel, A. L. 1955a. Studies of the mechanism of vitamin E action. I l l - I n v i t r o copolymerization of oxidized l i p i d s with proteins. Arch. Biochem. Biophys. 54 : 266-280. . 1956b. Unsaturated l i p i d oxidation catalyzed by hematin com-pounds. J . B i o l . Chem. 217 : 721-733. . 1956. Freeze-dried meat. III-The mechanisms of oxidative deterioration of freeze-dried meat. Food Res. 21 : 195-206. Tarr, H. L. A., H. M. B i s s e t t , B. A. Southcott, J . B i e l y , and B. E. March. 1951. The n u t r i t i v e value of f i s h meal and condensed f i s h solubles. II-Effect of heat on herring meal. Progr. Rept. P a c i f i c Coast Sta., Fish. Res. Bd. (Canada) No. 87 : 42-46. Thomas, H. V., P. K. Mueller, and R. L. Lyman. 1968. Lipoperoxidation of lung l i p i d s i n rats exposed to nitrogen dioxide. Science 159 : 532-533. Thurston, C. E., L. E. Ousterhout, and P. P. MacMaster. 1960. The n u t r i t i v e value of f i s h meal protein. A comparison of chemical measurements with a chick-feeding test. J . Assoc. O f f i c . Agr. Chem. 43 : 760-762. U r i , N. 1956. Metal ion catalysis and p o l a r i t y of environment i n the aerobic oxidation of unsaturated f a t t y acids. Nature 177 : 1177-1178. 158 U r i , N. 1958. The mechanism of the oxidation of l ino le ic acid with particular reference to metal catalysis . In "Fourth International Conference on Biochemical Problems of L ip ids , " ed. H. M. S inc la i r , pp. 30-35. Academic Press, New York. . 1961. (Chapter 2): Physico-chemical aspects of autoxidation. (Chapter 4): Mechanism of antioxidation. In "Autoxidation and Ant i -oxidants," Vol . I, ed. W. 0. Lundberg. Interscience Publ . , New York. Venolia, A. W., and A. L. Tappel. 1958. Brown colored oxypolymers of unsaturated fats . J . Amer. O i l Chem. Soc. 35 : 135, 138. Waissbluth, M. D., L. Guzman, and F. P. Plachco. 1971. Oxidation of l ip ids in f i sh meal. J . Amer. O i l Chem. Soc. 48 : 420-424. Wi l l s , E. D. 1965. Mechanisms of l i p i d peroxide formation in tissues. Role of materials and haematin proteins in the catalysis of the ox i -dation of unsaturated fatty acids. Biochem. Biophys. Acta 98 : 238-251. Yu, T. C , and R. 0. Sinnhuber. 1957. 2-thiobarbituric acid for the measurement of rancidity in fishery products. Food Technol. 11 : 104-108. , . 1967. An improved 2-thiobarbituric acid (TBA) procedure for the measurement of autoxidation of f i sh o i l s . J . Amer. O i l Chem. Soc. 44 : 256-258. Zebrowska, T. 1968. The course of digestion of different food proteins in the rat . Fractionation of the nitrogen in intest inal contents. Br i t . J . Nutr. 22 : 483-491. Z i r l i n , A . , and M. Karel. 1969. Oxidation effects in a freeze-dried gelatin-methyl l ineleate system. J . Food Sc i . 34 : 160-164. APPENDIX 159 160 TABLE 3(B) COMPOSITION OF DIETS CONTAINING FISH MEAL B STORED FOR FIVE MONTHS AT 21°C Ingredients Levels of protein supplied by f ish meal B 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 16.3 13.3 9.4 Fish meal B 5 8.8 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.8 3.5 2.5 Limestone - - 0.3 Ground cel lulose 5.9 5.4 5.1 TABLE 3(C) COMPOSITION OF DIETS CONTAINING FISH MEAL C STORED FOR FIVE MONTHS AT -20°C Ingredients Levels of protein supplied by f ish meal C 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.8 12.4 7.9 Fish meal C 5 8.7 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.8 3.5 2.5 Limestone - - 0.3 Ground cel lulose 6.4 6.4 6.6 *Micronutrients as in Table 3(A). TABLE 3(D) COMPOSITION OF DIETS CONTAINING FISH MEAL D STORED FOR FIVE MONTHS AT -20°C Ingredients Levels of protein supplied by f ish meal D 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.8 12.4 8 Fish meal D 5 8.8 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.8 3.5 2.5 Limestone - - 0.3 Ground cellulose 6.4 6.3 6.5 TABLE 3(E) COMPOSITION OF DIETS CONTAINING FISH MEAL E STORED FOR FIVE MONTHS AT 21°C Ingredients Levels of protein supplied by f ish meal E 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 16.2 13.1 9.1 Fish meal E 4.9 8.6 13.4 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.8 3.4 2.4 Limestone - - 0.3 Ground cel lulose 6.1 5.9 5.8 *Micronutrients as in Table 3(A). 162 TABLE 3(F) COMPOSITION OF DIETS CONTAINING FISH MEAL F STORED FOR FIVE MONTHS AT -20°C Ingredients Levels of protein supplied by f ish meal F 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.9 12.6 8.3 Fish meal F 4.9 8.6 13.4 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.8 3.4 2.4 Limestone - - 0.3 Ground cellulose 6.4 6.4 6.6 TABLE 3(G) COMPOSITION OF DIETS CONTAINING FISH MEAL A STORED FOR TEN MONTHS AT 21°C Ingredients Levels of protein supplied by f ish meal A 4% per 100 lb . 7% per 100 lb . 11% per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.7 12.3 7.8 Fish meal A 5 8.7 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.6 3 2.1 Limestone - - 0.5 Ground cel lulose 6.7 7 6.9 *Micronutrients as in Table 3(A). TABLE 3(H) COMPOSITION OF DIETS CONTAINING FISH MEAL B STORED FOR TEN MONTHS AT 21°C Ingredients Levels of protein supplied by f ish meal B 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 16.4 13.5 9.8 Fish meal B 5 8.8 13.7 Micronutrients* 0.5 0.5 0.5 Iodized sal t 0.5 0.5 0.5 Bonemeal 3.6 3 2.1 Limestone - - 0.5 Ground cellulose 6 5.7 4.9 TABLE 3(1) COMPOSITION OF DIETS CONTAINING FISH MEAL C STORED FOR TEN MONTHS AT -20°C Ingredients Levels of protein supplied by f ish meal C 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.6 12.2 7.6 Fish meal C 5 8.7 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.6 3 2.1 Limestone - - 0.5 Ground cellulose 6.8 7.1 7.1 *Micronutrients as in Table 3(A). 164 TABLE 3(J) COMPOSITION OF DIETS CONTAINING FISH MEAL D STORED FOR TEN MONTHS AT -20°C Ingredients Levels of protein supplied by f i sh meal D 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 15.7 12.3 7.8 Fish meal D 5 8.8 13.7 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.6 3 2.1 Limestone - - 0.5 Ground cellulose 6.7 6.9 6.9 TABLE 3(K) COMPOSITION OF DIETS CONTAINING FISH MEAL E STORED FOR TEN MONTHS AT 21°C Ingredients Levels of protein supplied by f ish meal E 4% per 100 lb . 7% per 100 lb . 11% per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 16.3 13.3 9.4 Fish meal E 4.9 8.6 13.4 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.5 2.9 1.9 Limestone - - 0.5 Ground cellulose 6.3 6.2 5.8 *Micronutrients as in Table 3(A). TABLE 3(L) COMPOSITION OF DIETS CONTAINING FISH MEAL F STORED FOR TEN MONTHS AT -20°C Ingredients Levels of protein supplied by f ish meal F 4% 7% 11% per 100 lb . per 100 lb . per 100 lb . Ground wheat 65 65 65 Vegetable o i l 3 3 3 Corn starch 16.1 13 9 Fish meal F 4.9 8.6 13.4 Micronutrients* 0.5 0.5 0.5 Iodized salt 0.5 0.5 0.5 Bonemeal 3.5 2.9 1.9 Limestone - - 0.5 Ground cellulose 6.5 6.5 6.2 *Micronutrients as in Table 3(A). 166 TABLE 4(A) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE FIRST EXPERIMENT, TREATMENTS A, B, C AND D Source of variation SS d. f . MS F Temperature 140019.84 1 140019.84 135.40** Antioxidant 229677.60 1 229677.60 222.14** T x A 167673.79 1 167673.79 162.17** Error 8271.36 _8 1033.92 Total 545642.59 11 * **Significant at 1% leve l . Duncan's multiple range test: Treatment A C D B Mean value 3910 3890 3850 3395 Mean values sharing the same l ine are not s igni f icant ly different at 1% leve l . TABLE 4(B) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE FIRST EXPERIMENT, TREATMENTS E AND F Source of variation SS d. f . MS Temperature 99135.19 1 99135.19 164.81** Error 2406.07 4 601.51 Total 101541.26 5 **Significant at 1% leve l . 167 TABLE 4(C-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (EXPRESSED AS GAIN IN WEIGHT) FOR TREATMENTS A, B, C, AND D TABLE 4(C-a): Meals fed to supply 4% protein to the diet . Source of variation SS d. f. MS F Temperature 0.55 1 0.55 0.01 N.S. Antioxidant 60.72 1 60.72 1.16 N.S, T x A 77.31 1 77.31 1.47 N.S. Error 630.44 12 52.54 Total 769.02 15 N.S. Not s igni f icant . TABLE 4(C-b): Meals fed to supply 7% protein to the diet . Source of variation SS d. f . MS Temperature 19.10 1 19.10 0.49 N.S. Antioxidant 119.14 1 119.14 3.03 N.S. T x A 63.32 1 63.32 1.61 N.S. Error 472.66 12 39.39 Total 674.22 15 N.S. Not s igni f icant . TABLE 4(C-c): Meals fed to supply 11% protein to the d iet . Source of variation SS d. f . MS F  Temperature 42.19 1 42.19 4.93* Antioxidant 54.32 1 54.32 6.35* T x A 25.91 1 25.91 3.03 N.S. Error 102.64 12 8.55 Total 225.06 15 N.S. Not s ignif icant * Significant at 5% leve l . Duncan's multiple range test : Treatment C A D B Mean value 107.16 106.46 106.02 100.23 Mean values sharing the same l ine are not s igni f icant ly different at 5% leve l . 168 TABLE 4(D-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (EXPRESSED AS GAIN IN WEIGHT) FOR TREATMENTS E AND F TABLE 4(D-a): Meals fed to supply 4% protein to the diet. Source of variat ion SS d.f . MS Temperature 1.25 1 1.25 0.17 N.S. Error 42.94 6_ 7.16 Total 44.19 7 N.S. Not s igni f icant . TABLE 4(D-b): Meals fed to supply 7% protein to the diet . Source of variation SS d. f . MS Temperature 61.38 1 61.38 0.31 N.S. Error 1198.59 6_ 199.77 Total 1259.97 7 N.S. Not s igni f icant . TABLE 4(D-c); Meals fed to supply 11% protein to the diet . Source of variation SS d. f . MS Temperature 161.91 1 161.91 2.39 N.S. Error 406.45 6_ 67.74 Total 568.36 7 N.S. Not s ignif icant 169 TABLE 4(E-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (EXPRESSED AS DIET EFFICIENCY) FOR TREATMENTS A, B, C AND D TABLE 4(E-a): Meals fed to supply 4% protein to the d iet . Source of variat ion SS d.f . MS F Temperature 0.000729 1 0.000729 4.50 N.S. Antioxidant 0.000064 1 0.000064 0.40 N.S. T x A 0.000506 1 0.000506 3.12 N.S. Error 0.001949 12 0.000162 Total 0.003248 15 N.S. Not s ignif icant Duncan's multiple range test : Treatment A B D C Mean value 0.360 0. .345 0.342 0.335 Mean values sharing the same l ine are not s igni f icant ly different at 5% leve l . TABLE 4(E-b): Meals fed to supply 7% protein to the diet . Source of variation SS d. f . MS F Temperature 0.000001 1 0.000001 0.00 N.S. Antioxidant 0.000036 1 0.000036 0.03 N.S. T x A 0.000900 1 0.000900 0.67 N.S. Error 0.016147 12 0.001345 Total 0.017084 15 N.S. Not s igni f icant . 170 TABLE 4(E-c) : Meals fed to supply 11% protein to the diet . Source of variation Temperature Antioxidant T x A Error SS 0.000018 0.001540 0.001958 0.005147 d.f . 1 1 1 12 MS 0.000018 0.001540 0.001958 0.04 N.S. 3.60 N.S. 4.57 N.S. Total 0.008663 15 N.S. Not s igni f icant . Duncan's multiple range test: Treatment A D C B Mean values 0.563 0.541 0.539 0.521 Mean values sharing the same l ine are not s igni f icant ly different at 5% leve l . 171 TABLE 4(F-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (EXPRESSED AS DIET EFFICIENCY) FOR TREATMENTS E AND F TABLE 4(F-a): Meals fed to supply 4% protein to the diet. Source of variation SS d . f . MS Temperature Error Total 0.000036 0.001303 0.001339 1 6 0.000036 0.000217 0.17 N.S. N.S. Not s ignif icant . TABLE 4(F-b): Meals fed to supply 7% protein to the diet. Source of variation SS d. f . MS Temperature Error Total 0.000060 0.001158 0.001218 1 6 0.000060 0.000193 0.31 N.S. N.S. Not s igni f icant . TABLE 4(F-c) : Meals fed to supply 11% protein to the diet, Source of variation SS d.f . MS Temperature Error Total 0.000406 0.001454 0.001860 1 6 0.000406 0.000242 1.68 N.S. N.S. Not s igni f icant . 172 TABLE 4(G) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS GAIN IN WEIGHT FOR TREATMENTS A, B, C AND D Source of variation SS d.f . MS F Levels 11223.46 2 5611.73 133.91** Temperature 44.93 1 44.93 1.07 N.S. Antioxidant 226.72 1 226.72 5.41* L x T 16.91 2 8.45 0.20 N.S. L x A 7.49 2 3.75 0.09 N.S. T x A 159.07 1 159.07 3.79 N.S. L x T x A 7.55 2 3.78 0.09 N.S. Error 1508.66 36 41.91 Total 12891.79 47 N.S. Not s ignif icant * Signif icant at 5% level ** Signif icant at 1% level TABLE 4(H) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS GAIN IN WEIGHT FOR TREATMENTS E AND F Source of variation SS d. f . MS Level 4284.68 2 2142.34 24.50** Temperature 126.00 1 126.00 1.44 N.S. L x T 98.55 2 49.27 0.56 N.S. Error 1573.80 18 87.43 Total 6083.03 23 N.S. Not s ignif icant ** Significant at 1% level 173 TABLE 4(1) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS DIET EFFICIENCY FOR TREATMENTS A, B, C AND D Source of variation SS d. f . MS F Level 0.306380 2 0.153190 237.14** Temperature 0.000305 1 0.000305 0.47 N.S. Antioxidant 0.000567 1 0.000567 0.88 N.S. L x T 0.000443 2 0.000222 0.34 N.S. L x A 0.001074 2 0.000537 0.83 N.S. T x A 0.003120 1 0.003120 4.83* L x A x T 0.000244 2 0.000122 0.19 N.S. Error 0.023244 36 0.000646 Total 0.33537 47 N.S. Not s igni f icant . * Significant at 5% leve l . ** Significant at 1% leve l . TABLE 4(J) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE FIRST EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS DIET EFFICIENCY FOR TREATMENTS E AND F Source of variation SS d . f . MS F Level 0.108403 2 0.054202 249.78** Temperature 0.000113 1 0.000113 0.52 N.S. L x T 0.000390 2 0.000195 0.90 N.S. Error 0.003914 18 0.000217 Total 0.112820 23 N.S. Not s ignif icant ** Signif icant at 1% level 174 TABLE 5(A) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE SECOND EXPERIMENT, TREATMENTS A, B, C AND D Source of variation SS d.f . MS F Temperature 310826.64 1 310826.64 253.94** Antioxidant 371184.19 1 371184.19 303.25** T x A 280816.20 1 280816.20 229.41** Error 9792.25 8 1224.03 Total 972619.28 11 **Significant at 1% level Duncan's multiple range test : Treatment C A D B Mean value 3950 3935 3905 3280 Mean values sharing the same : l ine are not s igni f icant ly different at 1% leve l . TABLE 5(B) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE SECOND EXPERIMENT, TREATMENTS E AND F Source of variation SS d. f . MS F Temperature 16569.01 1 16569.01 9.41* Error 7037.16 4 1759.29 Total 23606.17 5 *Significant at 5% leve l . 175 TABLE 5(C) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE FIRST AND SECOND EXPERIMENTS, TREATMENTS A, B, C AND D Source of variation SS d.f . MS F Time 426.56 1 426.56 0.37 N.S. Temperature 434042.27 1 434042.27 384.50** Antioxidant 592411.54 1 592411.54 524.70** Time x T 16804.22 1 16804.22 14.88** Time x A 8450.25 1 8450.25 7.50* T x A 441236.98 1 441236.98 390.80** Time x T x A 7253.01 1 7253.01 6.40* Error 18063.61 16 1128.98 Total 1518688.44 23 N.S. Not signif icant * Significant at 5% leve l . ** Significant at 1% leve l . Duncan's multiple range test: Treatment A, D„ D 2 2 1 2 1 1 1 2 Mean value 3950 3935 3910 3905 3890 3850 3395 3280 Mean values sharing the same l ine are not s igni f icant ly different at 1% leve l . TABLE 5(D) ANALYSIS OF VARIANCE OF THE METABOLIZABLE ENERGY VALUES OF THE FIRST AND SECOND EXPERIMENTS, TREATMENTS E AND F Source of variation SS d . f . MS F Time 86234.83 1 86234.83 73.06** Temperature 98380.77 1 98380.77 83.35** Time x T 17323.43 1 17323.43 14.68** Error 9443.22 8 1180.40 Total 211382.25 11 **Significant at 1% leve l . Duncan's multiple range test : Treatment F^ F 2 E 1 E 2 Mean value 3850 3605 3590 3500 Mean values sharing the same l ine are not s igni f icant ly different at 1% leve l . TABLE 5(E-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (EXPRESSED AS GAIN IN WEIGHT) FOR TREATMENTS A, B, C AND D TABLE 5(E-a): Meals fed to supply 4% protein to the diet . Source of variation SS d . f . MS F 75.52 4.43 N.S. 35.70 2.09 N.S. 83.63 4.91* 17.05 Total 399.45 15 Temperature 75.52 1 Antioxidant 35.70 1 T x A 83.63 1 Error 204.60 12 N.S. Not s igni f icant . * Signif icant at 1% level Duncan's multiple range test : Treatment D A C B Mean value 73.94 72.58 72.35 65.02 Mean values sharing the same l ine are not s igni f icant ly different at 5% leve l . TABLE 5(E-b): Meals fed to supply 7% protein to the d iet . Source of variation SS d. f . MS Temperature 66.34 1 66.34 0.96 N.S, Antioxidant 186.19 1 186.19 2.70 N.S, T x A 131.79 11 131.79 1.91 N.S. Error 827.99 12 69.00 Total 1212.31 15 N.S. Not s ignif icant TABLE 5(E-c) : Meals fed to supply 11% protein to the diet . Source of variat ion SS d.f . MS Temperature 15.32 1 15.32 0.29 N.S. Antioxidant 19.14 1 19.14 0.36 N.S. T X A 178.89 1 178.89 3.33 N.S. Error 643.90 12 53.66 Total 857.25 15 N.S. Not s ignif icant 177 TABLE 5(F-a, b and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (EXPRESSED AS GAIN IN WEIGHT) FOR TREATMENTS E AND F TABLE 5(F-a): Meals fed to supply 4% protein to the diet. Source of variation SS d. f . MS Temperature 19.53 1 19.53 1.57 N.S. Error 74.62 6_ 12.44 Total 94.15 7 N.S. Not s ignif icant TABLE 5(F-b): Meals fed to supply 7% protein to the diet . Source of variation SS d. f . MS Temperature 17.73 1 17.73 0.23 N.S. Error 464.79 6_ 77.46 Total 482.52 7 N.S. Not s igni f icant . TABLE 5(F-c) : Meals fed to supply 11% protein to the diet. Source of variation SS d. f . MS Temperature 87.19 1 87.19 1.00 N.S. Error 522.48 6_ 87.08 Total 609.67 7 N.S. Not s ignif icant TABLE 5(G-a, b, and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (EXPRESSED AS DIET EFFICIENCY) FOR TREATMENTS A, B, C, AND D TABLE 5(G-a): Meals fed to supply 4% protein to the diet . Source of variation SS d.f . MS F Temperature 0.000976 1 0.000976 13.01** Antioxidant 0.000010 1 0.000010 0.13 N.S T x A 0.000474 1 0.000474 6.32* Error 0.000905 12 0.000075 Total 0.002365 15 N.S. Not s igni f icant . * Significant at 5% level ** Significant at 1% level Duncan's multiple range test: Treatment D C A B Mean value 0.384 0.372 0 . 3 6 7 0 . 3 5 8 Mean values sharing the same l ine are not s ignif icant ly different at 5% leve l . TABLE 5(G-b): Meals fed to supply 7% protein to the diet . Source of variation SS d. f . MS F Temperature 0.000250 1 0.000250 2.29 N.S. Antioxidant 0.000396 1 0.000396 3.63 N.S. T x A 0.000263 1 0.000263 2.41 N.S. Error 0.001311 12 0.000109 Total 0.002220 15 N.S. Not s igni f icant . Duncan's multiple range test: Treatment C A D B Mean value 0.483 0.467 0.465 0.465 Mean values sharing the same l ine are not s igni f icant ly different at 5% leve l . 179 TABLE 5(G-c): Meals fed to supply 11% protein to the diet. Source of variation Temperature Antioxidant T x A Error Total SS 0.001332 0.000090 0.000001 0.002511 0.003934 d.f . 1 1 1 12 15 MS 0.001332 0.000090 0.000001 0.000209 6.37* 0.43 N.S. 0.01 N.S. N.S. Not s igni f icant . * Significant at 5% leve l . TABLE 5(H-a, b, and c) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (EXPRESSED AS DIET EFFICIENCY) FOR TREATMENTS E AND F TABLE 5(H-a); Meals fed to supply 4% protein to the diet. Source of variation SS d.f . MS Temperature Error Total 0.000050 0.000264 0.000314 1 6 0.000050 0.000044 1.14 N.S. N.S. Not s igni f icant . TABLE 5(H-b): Meals fed to supply 7% protein to the diet . Source of variation SS d.f . MS Temperature Error Total 0.000253 0.000173 0.000426 1 6 0.000253 0.000028 9.04* * Significant at 5% leve l . TABLE 5(H-c): Meals fed to supply 11% protein to the diet. Source of variation SS d. f . MS Temperature Error Total 0.000666 0.000402 0.001068 1 6 0.000666 0.000067 9.94* * Significant at 5% leve l . 181 TABLE 5(1) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT, (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS GAIN IN WEIGHT FOR TREATMENTS A, B, C, AND D Source of variation SS d. f . MS F Level 13701.08 2 6850.54 147.11** Temperature 143.52 1 143.52 3.08 N.S. Antioxidant 191.92 1 191.92 4.12* L x T 13.66 2 6.83 0.15 N.S. L x A 49.11 2 24.55 0.53 N.S. T x A 385.33 1 385.33 8.27** L x T x A 8.98 2 4.49 0.10 N.S. Error 1676.49 36 46.57 Total 16170.09 47 N.S. Not s igni f icant . * Significant at 5% leve l . ** Significant at 1% leve l . TABLE 5(J) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS GAIN IN WEIGHT FOR TREATMENTS E AND F Source of variation SS d.f . MS F Level 6766.68 2 3383.34 57.35** Temperature 107.61 1 107.61 1.82 N.S. L x T 16.84 2 8.42 0.14 N.S. Error 1061.88 18 58.99 Total 7953.01 23 N.S. Not s igni f icant . ** Significant at 1% leve l . 182 TABLE 5(K) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS DIET EFFICIENCY FOR TREATMENTS A, B, C, AND D Source of variation SS d. f . MS F Level 0.2478 2 0.1239 953.08** Temperature 0.0023 1 0.0023 17.69** Antioxidant 0.0002 1 0.0002 1.54 N.S. L x T 0.0003 2 0.0002 1.15 N.S. L x A 0.0003 2 0.0002 1.15 N.S. T x A 0.0001 1 0.0001 0.77 N.S. L x T x A 0.0005 2 0.0003 1.92 N.S. Error 0.0047 36 0.0001 Total 0.2562 47 N.S. Not s igni f icant . ** Significant at 1% leve l . TABLE 5(L) ANALYSIS OF VARIANCE OF THE PROTEIN QUALITY IN THE SECOND EXPERIMENT (INCLUDING 3 LEVELS OF SUPPLEMENTARY PROTEIN), EXPRESSED AS DIET EFFICIENCY FOR TREATMENTS E AND F Source of variation SS d. f . MS F Level 0.121848 2 0.060924 1296.25** Temperature 0.000793 1 0.000793 16.87** L x T 0.000176 2 0.000088 1.87 N.S. Error 0.000839 18 0.000047 Total 0.123656 23 N.S. Not s ignif icant ** Significant at 1% level 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0101245/manifest

Comment

Related Items