04...".th ‘ 3' This is to certify that the thesis entitled Comparisons Of Chemical, In Vitro And In Vivo Methods For Estimating Nutritive Values Of Normal And Chemically Treated Haylages presented by Yu Yu has been accepted towards fulfillment of the requirements for Ph- D- degree inDainLScience 1% “W4... V / a?‘ professor Date 2/1/71sl R 0-7639 ABSTRACT COMPARISONS OF CHEMICAL, IN VITRO AND IN VIVO METHODS FOR ESTIMATING NUTRITIVE VALUES OF NORMAL AND CHEMICALLY TREATED HAYLAGES By Yu Yu The present study was designed to gain knowledge about two problems involved in haylage preservation. The objective of the first part of this study was to evaluate and compare several chemical and in vitro methods in estimating five animal responses (nitrogen and dry matter digestibilities, nitrogen balance and retention as a percent of absorbed nitrogen, and maximum dry matter intake) when fed hay or haylage. Chemical analyses evaluated were crude protein content, dry matter solubility in neutral detergent solution, dry matter and nitrogen (N) solubility in acid-detergent (AD), solubility in hot water, acid detergent lignin, N solubility in mineral buffer solution and degree of browning. Analyses obtained from in vitro methods were dry matter and N solubilities when incubated in: acid pepsin, pepsin + pancreatin, rumen fluid + pepsin, and r . . . umen fluid + pep81n + pancreatin solutions. In several Yu Yli cases, one analytical scheme produced two to four vari— ables. Sources of forage samples were: 2h samples witli the majority being alfalfa haylages obtained from Michin- gan State University (MSU), and 66 samples supplied frcnn other U.S. experiment stations. Only limited analyticzrl data were available for those 66 samples. Each in ziggz parameter was regressed with (l) a single laboratory measurement using simple regression analysis technique, (2) with variables obtained from any two analytical schemes using least squares deletion multiple regressicni analysis technique, and (3) with selected, important variables from all analytical schemes using a least squares deletion multiple regression analysis techniqtma. For the 2k MSU forage samples, the best single predictors for (1).lE.KlKE N digestibility was AD in- soluble N as a percent of total N (r = -.92); for (2) dry matter digestibility was AD insoluble dry matter (r = -.90); for (3) N balance was pepsin soluble N as a percent of dry matter (r = .85); for (h) N reten- tion as a percent of absorbed N was hot water soluble dry matter (r = .62); and for (5) maximum dry matter intake was rumen microbial plus pepsin soluble dry matter (r = .82). Forage crude protein content was a very poor predictor for 12.X3K2 N digestibility (P = .19) or digestible N content (r = .59). Yii Yu Predictability of digestible N content was markedly improved by using N fractions determined by either AD or pepsin methods (R2 = .91 and .89). The effect of source of samples was evident witli respect to the regression equations for in_zigg N di- gestibility using AD insoluble N as predictor. Forage samples of MSU contained smaller amounts of insoluble N and had a greater depressions in ND per unit of AD insoluble N than did samples from other sources. A greater depression in ND was observed for forages con- taining 9% AD insoluble N or less than for forages cone- taining greater than 9% AD insoluble N as a percent (If total N. Multiple regressions analyses using variables from two analytical methods indicated that variables obtained from the pepsin incubation were the best pre-— dictors for in KEXE responses. Extremely high predicH:i_ bility (R2) .98) were obtained for multiple regressirnis of 12;XEXE,N and dry matter digestibility using selectexi variables such as neutral detergent insoluble dry maiflser, AD insoluble N, pepsin insoluble N and rumen microbial + pepsin insoluble N. The aim of the second part of this study was to evaluate and compare the value of propionic acid (.h and .8%), ammonium isobutyrate (AIB, .S and 1%) and a. Yu Yu mixture of AIB (.5%) and formaldehyde (1.25% of a 37% solution) in preserving nutritive value of alfalfa hay— lage (50% DM). Levels of propionic acid and AIB were comparable on a molar basis. During a h2-day ensiling period, haylage treated with .5% AIB had the least quantity of heat development as expressed as degree-day above 35 C (66) as compared to other treated haylages (ranging from 119 to 203) and control haylage (322). Heat development was greatest in the upper portion of silo regardless of treatment. None of the chemical treatments were entirely effective in restricting heat develOpment ()»35 C) in the haylage sur— face during the time haylage was being fed. However, results from a refermentation experiment indicated that all treatments retarded heat development (j>35 C) for at least 10 days while control haylage heated to 59 C on the second day of refermentation. The extent of top spoilage was not reduced by treatments except for .8% propionic acid and .5% AIB plus formaldehyde. These two treatments also restricted heat development in the top portion of the silos to a greater extent than for control haylage. Treatments had no marked effect on haylage pH or acetic acid concentration, but did markedly decrease lactic acid concentration. Propionic acid and AIB re- duced total fungal counts to about the same extent; a hO% reduction for the .h or .5% levels and a 75% Yu Yu reduction for the .8 or 1% levels. Treatment with .5% AIB plus formaldehyde had essentially no effect on reduction of total fungal counts. Chemical composition was not markedly different among haylages except that high lignin and AD insoluble N values were found for control, 1% AIB and .5% AIB plus formaldehyde treated haylages. These haylages were also high in quantity of heat development during storage. Maximum haylage dry matter intake by sheep ranged from 3.1h to 3.99 kg per 100 kg body weight with no sig- nificant differences among haylages. Dry matter digesti- bility was not significantly (p >.05) improved by treat- ments, however, N digestibility was significantly (jp< .05) improved by treatments (from 55 to 60.3%). N balance and N retention as a percent of absorbed N were also significantly (p (.05) improved by treatment with .h% propionic acid and 1% AIB. Significant correlation coefficients were observed between N digestibility and AD insoluble N as a percent of total N (r = -.82, p‘(.01) and between N digestibility and degree—days above 35 C (r = -.81, p‘<.Ol). Milk production, composition and effeciency were not significantly (pj>.05) different among treatments. Thus, propionic acid and AIB were equally effective in reducing heat development, total fungal counts and AD insoluble N of haylages and in Yu improving N utilization._ No markedly superior results were obtained by using the higher levels of these chemicals. Yu COMPARISONS OF CHEMICAL, IN VITRO AND IN VIVO METHODS FOR ESTIMATING NUTRITIVE VALUES OF NORMAL AND CHEMICALLY TREATED HAYLAGES BY Yu Yu A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Dairy Science 197A ACKNOWLEDGMENTS The author extends his appreciation to Dr. J.W. Thomas for his advice, guidance and patient counsel throughout his graduate program. His encouragement and enthusiasm have been greatly appreciated. The author is further indebted to the other members of his graduate committee, Drs. J.R. Brunner, R.M. Cook, R. Lueeke, H.A. Tucker and J.M. Wilkinson, for their advice and willing participation in the writer's graduate program. Appreciation is extended to Dr. C.A. Lassiter, Chairman of the Dairy Department, for financial assistance throughout this study. The author also wishes to thank Mr. T. Middleton, P. Tinnimit, R. Greening, T.A. Ferris, L. McGuffey, J. Ball and Dr. Neitzel for their help during this research. The author extends his sincere graditude to his parents and to his wife, Grace, for their continued interest and encouragement throughout the author's education. ii TABLE OF CONTENTS Page LIST OF TABLES 0.0.00.00.00.00...0.0.0.0.......... Vii LIST OF FIGURES ...... ............OOOOOOOOOOOOOOOO Xiv INTRODUCTION .....OOOOOOOOOOOO......OOOOOOOOOOOOOO 1 LITERATURE REVIEW 000.000.........OOOOOOOOOOOOO... 3 Part 1. Regular (High-Moisture) Silages ....... 3 1. Definition and General Characteristics ... 3 II. Reaction Occuring During Silage Fermen- tation00............OOOOIOOOO0.0.0.000... 8 1. Carbohydrates ........................ 8 2. Protein and Amino Acids .............. 10 3. Organic Acids (Non-Nitrogenous) ...... 11 A. Dry Matter and Energy ................ 12 III. Factors Influencing Fermentation and Nutri- tive Value of Resultant Silage ........... 1h 1. Soluble Carbohydrate Content.......... 1h 2. Mechnical Treatment .................. 17 3. Ensiling Temperature ................. 17 h. Moisture Content of Forage ........... 18 Part 2. Low—Moisture Silage (Haylage) ......... 22 1. Definition and General Characteristics ... 22 II. Advantages of Haylage Making ............. 22 III. Disadvantages of Haylage Making .......... 25 IV. Laboratory Estimates of Nitrogen Nutritive Value in Heat Damaged Feeds .............. 33 1. Protein Solubility in Rumen Fluid and Other Related Solvents ............... 3h iii 2. Nitrogen Solution Nitrogen Solution Nitrogen binations Nitrogen \lO\ U147 K») Page Solubility in Acid Detergent OOOOOOOOOOOOOOCOOOOCO000...... 37 Solubility in Acid Pepsin coo-00000000000000.0000.oooooo 39 Solubility in Various Com- of Proteases Solutions ..... . hO Solubility in Rumen Fluid, Pepsin and Pancreatin Solutions ....... hl Hot-Water Insoluble Nitrogen .......... E2 Chemical Determination of Lysine Availability ......OOCCOOCCCOCCOCOOCOO. LLB 8. Degree of Non-Enzymic Browning ........ uh V. Methods Available to Prevent Heating and Spoilage in Haylages ...................... M6 MATERIALS AND METHODS ............................. 59 Part 1. Evaluation of Forage Protein Quality by Laboratory Methods ..................... 59 I. Forage Samples ............................ 59 II. Laboratory Me thods Used to Evaluate Forage Protein Quality O......OOOOOOOOOOOOOOOOOOO. 62 A. Protein S Olubility......OOIOOOOOOOOOO. 62 1. Solubility in hot water ........... 62 2. Solubility in acid detergent SOlution ......OOOOOOOOOOIOOOOO.... 63 3. Solub ility in diluted phosphate- bicarbonate mineral buffer solution .......................... 6h B. In Vitro Protein Digestion ............ 6h 1. Acid pepsin digestion ............. 6h 2. Acid pepsin and pancreatin digestions ........................ 65 3. Rumen microbial and acid pepsin digestions ........................ 66 h. Rumen microbial, acid pepsin and pancreatin digestions ............. 66 5. Rumen ammonia release ............. 67 6. Degree of browning ................ 67 III. Statistical Analyses ...................... 68 Part 2. Haylage Preservation With Propionic Acid, Ammonium I sobutyrate and Mixture of Ammonium Isobutyrate and Formaldehyde .. 69 I. EnSiling TeChniqueS OOOOOOOOOOOOOOOOOOOOOOO 69 iv Page II. Feeding Trial .............................. 72 III. Milk Analysis .............................. 73 IV. Feed Analysis .............................. 73 V. Sheep Digestion and Nitrogen Metabolism Trials oooooooooooooooooooooooo ooooooooo 000. 77 VI. Statistical Analysis ............... ........ 78 RESULTS AND DISCUSSIONS ............................ 80 Part 1. Forage Nutritive Value Evaluation by Several Laboratory Methods .............. 80 1. Relationships Among l2 Vivo Responses ...... 80 ‘11. True Digestion Coefficients of Forage Total Nitrogen and Other Nitrogen Fractions Esti- mations by Statistical Means ............... 82 III. Relationships of Various Nitrogen Fractions to Nutritive Value ......................... 88 (1) Total Nitrogen Digestion Coefficient .. 88 a. Nitrogen digestion coefficient and acid detergent nitrogen ........... 88 b. Nitrogen digestion coefficient and other nitrogen fractions .......... 92 (2) Regressions for Nitrogen Balance ...... 100 (3) Regressions for Nitrogen as A Percent of Absorbed Nitrogen .................. 103 (h) Regressions for Estimating Dry Matter Digestion Coefficients ................ 103 (5) Regressions of Maximum Dry Matter In- takes ................................. 11h IV. Relations Among Nitrogen Containing Fractions 00.0.0000.........OOOOOOOO. ....... llu V. Relations Among Various Dry Matter Solubi- lity Measurements .......................... 119 VI. Regressions of Five In_Vivo Parameters on Thirteen Selected Laboratory Determinations Grouped From An Operationsl Standpoint ..... 120 VII. Multiple Regression of Five In Vivo Para- meters With Measurements of Two Laboratory Determinations ............................. 127 Page VIII. Multiple Regressions of Five In Vivo Parameters Usinquelected Variables —That Produced High Rt Values ................. 136 IX. Comparisons of Multiple Regressions Deve- loped From Different Sources of Forage Samples for Estimating In Vivo Nitrogen Digestion Coefficients ................... 1hh SUMMARY AND CONCLUSIONS .......................... 1&8 Part 2. Haylage Preservation With Various Chemicals ............................. 151 I. Effect of Chemicals on Alfalfa Haylage Temperatures ............................. 151 (A) Effect of Chemicals on Haylage Tem- peratures During Storage .......... 151 (B) Effect of Chemicals on Haylage Surface Temperature After Silos Were Open for Feeding.............................. 162 (C) Effect of Chemicals on Haylage Temperature During Refermentation ... 16h II. Effect of Chemicals on Haylage Dry Matter Losses During Storage .................... 168 III. Effect of Chemicals on Haylage Characteristics .......................... 170 IV. Effect of Treatments on Chemical Compo- Sition Of Haylage OOOOIOOOOOOIOOOOOOOOOOOO 178 V. Effect of Haylage Treatments on Digestion Coefficients and Nitrogen Utilization .... 186 VI. Effect of Chemicals on Haylage Consumption, Milk Production and Composition of Milk of Lactating cows 0 O O O O O O O O O O O O O O O O O O O O O O O O O O 198 SWEY AND CONCLUSIONS 0 O O O O O O O O O O O O O O O O O O O O O O O O O 211 APPENDIX .0000.........OOOOOOOOOOOOOOOO000......O. 211+ BIBLIOGRAPIN 00.00.00.000.........OOOOOOOOOOOOOOO. 2’47 vi Table 10. 11. 12. LIST OF TABLES Chemical Composition Changes of Fresh and Ensiling Perennial Ryegrass ................. Main Products of Carbohydrate Fermentation by Lactic Acid Bacteria ........................ Some Examples of Clostridial Fermentation ... Main Products of Organic Acid Fermentation by Lactic Acid Bacteria ........................ Comparison of Chemical Composition Among Fresh CrOp, Wilted Silage and Haylage .. ..... Proximate Composition of Hay and Haylage .... Laboratory Analyses Used to Evaluate Nitrogenous Feeds OOIOOOOOOOOOOOOOOOOOOOOOO00 Mold-Free Storage Time (Weeks) For Corn With 2h% Moisture Content Stored at Approximately 23 C ........................................ Analytical Values for Forages Used in Protein Solubility Study. Minimum, Maximum and Gross Mean Values are Given for Each Item ......... Digestion Coefficients, Nitrogen Utilization and Maximum Intake of 2h Forages Used in Protein Solubility Experiments .............. Harvesting Dates, Treatment and Dry Matter Content of Alfalfa Ensiled as Haylage, 1973 . Simple Correlation Coefficient (r) Among Five In Vivo Parameters Obtained From Sheep Fed Haylages ......................... ..... ...... vii Page 10 12 23 2h 33 AB 60 61 7O 81 Table 13. 1h. 15. 16. 17. 18. 19. 20. 21. 22. 23. 2h. Page Simple and Multiple Regressions of Digestible Nitrogen (Y) of Forage Nitrogen Fractions for 21 Michigan State University Forages ......... 8h Linear Regression Analysis for In Vivo Nitrogen Coefficients (Y) Using Acid Deter- gent Insoluble Nitrogen as A Percent of Total Nitrogen (X) ........................... 90 Linear Regressions for Estimating Total Nitrogen Digestion Coefficients (Y) Using Various Laboratory Values .................... 93 Regression Coefficients (b0) From the Regression Equations Using N Solubility Determined by Several Solubility Methods as the Predictor ................................ 99 Linear Regressions for Nitrogen Balance Data (Y) on Various Laboratory Values ............. 101 Linear Regressions for Retained Nitrogen as % of Absorbed (Y) on Various Laboratory values 000......COO......OOOOIOOOOOOOOCOO ..... lOLL Linear Regressions for Dry Matter Digestion Coefficients (Y) on Various Laboratory Values. 105 In Vivo Digestion Coefficients of Various Forage Dry Matter Fractions Estimated by Statistical Means ............................ 110 Linear Regressions and Correlation Coeffi- cients of Maximum Dry Matter Intakes ( % Body Weight Determined on Sheep) (Y) on Various Laboratory Values ............................ 115 Simple Correlations Among the 31 Measurements Studied 0000......OOOOOOOOOOOOOOOOOOOOO0...... 116 Regression Coefficients of Five In Vivo Para- meters Determined From Sheep on 13 Laboratory Determinations Which are Grouped From An Operational Standpoint ....................... 121 Three Multiple Regressions Calculated by Using 2 Groups of Laboratory Measurements That Gave High R 2 Values for In Vivo Nitrogen Digestion Coefficients (n = 217—....... ....... .......... 128 viii Table . Page 25. Three Multiple Regressions Calculated by Using 2 Groups of Laboratory Measurements That Gave High R2 Values for In Vivo Dry Matter Digestion Coefficients—(n = 21) ...... 131 26. Three Multiple Regressions Calculated by Using 2 Groups of Laboratory Measurements That Gave High R2 Values for In Vivo Nitrogen Balance (n = 21) ................... 13h 27. Two Multiple Regressions Calculated by Using Variables of 2 Groups of Laboratory Measure- ments That Gave Relatively High R2 for In Vivo Nitrogen Retention as A Percent of Absorbed Nitrogen (n = 21) .................. 135 28. Two Multiple Regressions Calculated by Using 2 Groups of Laboratory Measurements That Gave Relatively High R2 Values for Maximum Dry Matter Intake (n = 21)....................... 137 29. Multiple Regressions Calculated by Using Selected Variables That Produced High R for In Vivo Nitrogen Digestion Coefficients ( n = 21 ) .................................. 139 30. Multiple Regressions Calculated by Using Selected Variables That Produced High R 2 for In Vivo Dry Matter Digestibility (n=21).. lhl 31. Multiple Regressions Calculated by Using2 Selected Variables That Produced High R for In Vivo Nitrogen Balance (n = 21) ....... 1&3 32. Comparisons of Multiple Regressions for lg Vivo Nitrogen Digestion Coefficient Developed by Using Different Sources of Samples Forage. 1&6 33. Selected Hourly Haylage and Ambient Tempera- tures (C) During the Storage Period (June 2h t0 AUSUSL 2’ 1973) oooooooooooooooooooooooone 1.52 3h. Average Weekly Temperatures of Haylages During A h2—Day Ensiling Period. Haylages Were Treated With Propionic Acid, Ammonium Isobuty- rate (AIB) and Mixture of AIB and Formaldehyde ..................... ....... .... 159 ix Table 35. 36. 37. 38. 39. h0. h1. h2. 1+3. ML- 1L5. h6. Mean Weekly Temperatures of Haylages Ensiled in Concrete Silos for A h2-Day Storage Period. Temperatures are Presented Based on the Level of Position of Thermocouple in the Silo ..... Mean and Maximum Temperatures Measured 25 cm Below the Surface of Haylages During A Period of 69 Days When Silos Were Open ....... Amount of Dry Matter ReCOvered and Spoiled After h2-Day Storage Period in Concrete Silos of Control and Treated Haylages ............. Dry Matter Content, pH, Organic Acids, Fungal Counts and Recovery of Additives to Haylages .................................. Dry Matter Content, pH, Organic Acids and Mold Counts of Control and Treated Haylages Chemical Composition of Control and Haylages Treated With Propionic Acid, Ammonium Isobutyrate and Mixture of Ammonium Isobutyrate and Formaldehyde ....... Fibrous Constituents and Acid Detergent Insoluble Nitrogen in Control and Treated Haylages. Values are Means of Three Determinations From Three Composite Samples.. Linear Regressions Between Haylage Tem- peratures and Four Analytical Fractions ..... Linear Regressions Between Haylage Tem- perature Measured at Three Silo Levels and Four Analytical Fractions .................... Table of Analysis of Various Used to Analyze Data of Sheep Feeding Trials ................ Dry Matter Intake, Digestion Coefficients, Nitrogen Utilization and Body Weight Changes of Sheep Fed Control and Treated Haylages ............OIOOOOOOOOOIOOCOI00...... Linear Regressions for 12 Vivo Dry Matter and Nitrogen Digestion Coefficients on Two Acid Detergent Insoluble Nitrogen Fractions and Three Temperature Measurements .......... Page 161 163 169 172 175 179 181 183 185 186 187 Table Page M7. Sheep Performance Data Which Were Signi- ficantly Influenced by the Effect of Time (Level Of the SilO)0000000000000000.0000.0000 193 h8. Several Measurements Related to the Levels in the Silo .......................... 19h A9. Consumption of Haylage and Total Dry Matter and Body Weight (BW) Changes of Lactating Cows Fed Control and Treated Haylages ....... 199 50. Milk Production During Preliminary and Experimental Period of Cows Fed Control and Treated Haylages ........................ 201 51. Composition of Milk Produced During Preliminary and Experimental Period of Cows Fed Control and Treated Haylages ....... 20h 52. Comparisons of Various Effects Related to Haylage Treatments When Values of Control Haylage Were Expressed as 100 ............... 206 Appendix Table 1. Description of Samples Used in Forage Protein Quality Evaluation Experiments ...... 21k 2. Approximate Analyses of Samples Used in Forage Protein Quality Evaluation Experiments 0000000000000000000 00000 000000000 216 3. Fibrous Constituents Analysis of Samples Used in Forage Protein Quality Evaluation Experiments 00000000000000000000000.000000000 217 h. Sheep Performance Data of Samples Used in Forage Protein Quality Evaluation Experiments 000......OOOOCOCOCCO0000000000000 218 5. Dry Matter Solubilities of 2h Forage Samples Used in Protein Quality Evaluation Experiments 000000000000000000000000000000000 220 6. Protein Solubilities of 2h Forage Samples Used in Protein Quality Evaluation Experiments 000000000000000000000000000000000 222 xi Appendix Table 7. 10. 11. l2. 13. la. 15. 16. 17. 18. 19. 20. Chemical Composition of Forages Supplied by Dr. H. K. Goering, USDA.................. In Vivo Digestion Coefficients of Forages Supplied by Dr. H.K. Goering, USDA . Composition and Digestion Coefficients of Forages Supplied by Dr. N.A. Jorgensen (Univ. of Wisconsin) and Dr. D.C. Pierson (Univ. of Minnesota) ....... ................ Original Data of Sheep Digestion Trials (Dry Matter Digestibility) ......... ........ Original Data of Sheep Digestion Trials (Organic Matter Digestibility) ............. Original Data of Sheep Digestion Trials (Digestibility of Cell Walls)............... Original Data of Sheep Digestion Trials (Digestibility of Acid Detergent Fiber) .... Original Data of Sheep Digestion Trials (Digestibility of Nitrogen) ................ Original Data of Sheep Digestion Trials (Nitrogen Balance, g. N/Day) ............... Original Data of Sheep Digestion Trials (Nitrogen Retained As A Percent of Absorbed Nitrogen)..................... ..... Original Data of Sheep Digestion Trials (Nitrogen Retained As A Percent of Nitrogen Intake)00..000000000000000...000000....O0OC0 Original Data of Sheep Digestion Trials (Maximum Dry Matter Intake, % Body Weight).. Original Data of Sheep Digestion Trials (Dry Matter Intake During Digestion Trials, % Body Weight).............................. Original Data of Sheep Digestion Trials (Maximum Digestible Dry Matter Intake, % Body Weight) ............. ..... ........... xii Page 226 229 232 23h 235 236 237 238 239 2&0 2M1 2H2 2&3 2AA Appendix Table Page 21. Original Data of Sheep Digestion Trials (Digestible Dry Matter Intake During Digestion Trials, % Body Weight)............ 2&5 22. Original Data of Sheep Digestion Trials (Body Weight Change — g/Day) ............ 2&6 xiii LIST OF FIGURES Figure Page 1. General Relationships Among Forage, Ensiling Conditions, Silage Fermentation and Nutritive Value of Silage .............. 15 2. Development of Heat in A Silo and the Consequences of Heating ..... ............... 32 3. A Concept of the Relationship of Nitrogen Solubility of A Feed to Nitrogen Balance With A Ruminants ........................... 35 &. Preparations of Composited Haylage Samples for Various Kinds of Analyses .............. 7& 5. Temperature of Haylage During Storage Period. Temperature Recording Started on June 22, 1973 But Filing Dates Were Five, Zero, Three, One and One Day Before June 22 for Silos 3, &,5,6,7 and 8 Respectively ........ . ..... ... 156 6. Temperature Developments During Refermenta- tion of Haylage Treated With Various Chemicals 0000000000000000000000000000000000 165 7. Schematic Presentation of Relationships Among Variables Considered to be Important in Haylage Evaluation and Their Simple Linear Correlation Coefficients ............ 195 xiv 2 heat damaged forages. Thus, there is a need for analy- tical methods which must not be too complicated to per- form but can still precisely estimate the in lizg_protein value of heat damaged forages. Although the necessity of air exclusion during ensiling haylage has been recognized by agricultural re- searchers and farmers, instances of severe heat damage in haylage still occur frequently regardless the type of silos (including so called "oxygen-limiting" silos). On the other hand, reductions in.heat develoPment have been observed in haylage treated with various kinds of preser- vatives. However, more work is needed in order to define the interrelationships among application rate of the che- mical, degree of preservation, and nutritive value of preserved haylage as determined by animal feeding trials. Thus, the objectives of this study were (1) to compare the predictive values of several laboratory methods which have not been.thoroughly evaluated to date for in 2132_DM and nitrogen (N) digestibilities, N- balance, and maximum DM intake of haylages which were suspected of having variable amounts of heat damage and (2) to compare the effectiveness of propionic acid, ammonium isobutyrate (AIB) and mixture of AIB and formaldehyde in preserving nutritive values of alfalfa haylages as determined by sheep performances and milk yield of lactating cows. LITERATURE REVIEW Part 1. Regular (High-Moisture) Silages 1. Definition and General Characteristics Silage is a succulent material produced by a process of microbial fermentation of a green crop ( Watson and Nash, 1960). The anaerobic and acidic (pH &.2 or below) condition makes a long period of storage possi- ble. The primary object of silage making is to preserve the material with minimum loss of nutrients and with good resulting palatability. Traditionally however, more hay crops have been preserved as hay rather than silage. Only during recent years, have greater portions of hay crops been harvested for silages. Reasons for this change are: weather condition is not as large a problem when harvesting haylage as when making hay; ensilage causes less harvesting losses; recent development of automated harvesting and feeding systems minimize labor for a silage feeding program; equal or better perfor- mance of animals fed silage as compared to those fed hay in mixed practical rations (Thomas 23 2l° 1969; Roffler ‘gt‘gl. 1967 and Syrjala, 1972). Three types of fermentation can occur in silage (1) Lactic (normal or desirable) fermentation, (2) Secon- dary (butyric or clostridial) fermentation which is conventionally consiered undersirable and can occur both during and/or after the lactic fermentation resulting in degradation of amino acids and lactic acid and (3) A fer— mentation that occurs under aerobic conditions such.as when a silo is opened for feeding (the so called "after fermentation"). Coli-Aerogenes bacteria, yeasts and fungi are primarily responsible for the third type of fermentation (Papendick and Singh-Verma, 1972; Beck, 1963; Gross and Beck, 1970 and Beck and Gross, 196&). The course of a lactic acid type fermentation can be characterized by continued plant cell respiration for a time using up the oxygen.and giving off CO2 and heat. As conditions become favorable, members of acid-producing bacteria (Streptococci and Lactobacilli) increase rapidly. These organisms produce acid until the sugar is exhausted or until the pH becomes unfavorable for their further growth. The fermentation usually is complete at the end of eight days (Barnett, l95&; Conden gt EL. 1969; Langston et 31. 1958; Langston and Bouma, l960a,b). Investigators are still searching for reasons to explain why some silages show predominantly lactic acid type fermentation and others show predominantly butyric acid type fermentation. Many hypothese (e.g. sugar content, ration of protein to sugar content, buffering capacity, initial distribution of microflora on the crops, total counts of lactic acid producing bacteria during early stage of ensilage) have experimentally been found to be inconclusive (Kempton and San Clemente, 1959; Langston gt 21. 1958; Huhtanen.and Pensack, 1963; Ohyama and Masaki, l968a,b,c; l969a,b; and 1971). Langston and Bouma (1960a) have listed several other possible reasons to account for the variability in acid produc- tion, they are: (l) variability in sequence changes of microorganisms, (2) antagonism among certain groups of bacteria early in the fermentation process, (3) deficient nutrients in the plant material for bacterial growth, and (&) occurrence of weakened strains of bacteria. A poor understanding concering butyric acid types of fer- mentation and their undersirability has caused resear- chers to take a new direction in silage preservation. Present work is centered on finding on agents which will selectively restrict the activity of clostridia such as mineral acids, organic acids and wilting of forages. Numerous chemical reactions can occur during silage fermentation and this results in modification of the chemical composition of the original material. Generally, the major changes are reduction of soluble carbohydrate and true protein nitrogen; an increase in non-protein nitrogen as well as organic acids (lactic, acetic, propionic and butyric acids) (Watson and Nash, 1960). An example of chemical composition changes during ensiling is shown in the Table 1(Henderson.§thal. 1972). Table 1. Chemical Composition Changea of Fresh.and Ensiling Perennial Ryegrass. Fresh Silage ........ DM ------ DM 17.75 18.35 Water soluble carbohydrates 17.70 1.22 , Crude protein l&.20 l&.50 Protein-nitrogen (N) 1.8& 0.&5 Volatile - N/total N % o 8.15 Ether-extract 2.&l 3.&5 Crude fiber 26.50 30.&0 Acetic acid - 3.&0 Propionic acid - 0.17 Butyric acid - 0.16 Lactic acid - 10.60 Ethanol - 1.20 pH 6.08 3.9& Ash 7.00 6.90 aData from Henderson 93 El' 1972. J. Sci. Fd Agric. 23:1079. Readily available carbohydrates are lower in silage than in the original forage (17.7 vs. 1.22% DM). These utilized carbohydrates are converted into various kinds of organic acids. Although ensiling usually does not result in loss of crude protein, relative proportions of the nitrOgen fractions do change. For example, fresh ryegrass contains no volatile nitrogen while the resul- ting silage contains 8.2% volatile nitrogen calculated on the total nitrogen basis (Table l). The term "silage quality" is generally used to denote not the nutritive value of the silage, but the extent to which the silage fermentation has proceeded in a desirable manner. Conventionally, good quality silages should have high levels of lactic acid (7.5-12.5 % DM), low pH values ( $ &u2); low level of butyric (< 0.5% DM) and acetic acid (2.5-&.O% DM); and low level of ammonia nitrogen as percent of total nitrogen (5-8%) (Virtanen, 1933; 1952; Watson and Nash, 1960; Flieg, 1938; Breirem and Ulvesli, l95&; Nordfeldt, 1955; Wieringa, 1966 and Nilsson and Nilsson, 1956). Results from animal performance trials generally indicate that both animal digestibility and production are lower when the silage has a high pH value and/or high content of volatile nitrogen (Gordon 23 El’ 196&; Murdoch, 1966), but any negative effect of butyric acid on silage con- sumption has been completely verified. Low partial correlation coefficients have been reported between content of butyric acid and silage dry matter consumption (Emery 23 21. 1966; Kirchgessner gt_al. 1972). Silage production and.preservation have undergone many changes during the last decade, and many questions have been answered but an abundance of questions remain unanswered. Recently, several workers have frequently mentioned the significance of using aseptically grown ' forage or sterilized forage as ensiling material. Basic knowledge about silage fermentation can be obtained by this technique (Huhtanen and Pensack, 1963; Playne gt 31. 1967). II. Reactions Occuring During Silage Fermentation l, Carbohydrates Glucose, fructose, sucrose and fructosans are the main water-soluble carbohydrates in grass with only glucose and fructose considered major for microbiological purposes. These sugars are fermentable by a variety of microorganisms, of which lactic acid bacteria are the most important. Two fermentative types of lactic acid bacteria are always encountered. One is the homo-fermen- tative type, which forms approximately two moles of lactic acid per mole of glucose fermented (Wood, 1961). The second is the hetero-fermentative type which produces one mole of lactic acid, one mole of CO2 and one mole of ethanol per mole of glucose fermented. Prediction of the final ratio of products of a lactic acid fermentation is impossible because a mixed population always develops. In good silages, total counts of hetero-fermentative bacteria are higher than that of homo-fermentative bacteria which probably reflects the ability of hetero fermentative organisms to withstand a lower pH than homo-fermentatives (Langston, gt_gl, 1958). Main products of carbohydrate fermentation by lactic acid bacteria are given in Table 2. The data indicate that hetero-fermentative lactic acid bacteria ferment fructose and glucose by slightly different path- ways and produce less lactic acid than do homo-fermenta- tive bacteria. Table 2. Main Products of Carbohydrate Fermentation by Lactic Acid Bacteria. Homo-fermentative: (a) 1 Glucose -———9 2 Lactic acid (b) 1 Fructose-———9 2 Lactic acid (c) 1 Pentose ————e 1 Lactic acid + 1 Acetic acid Hetero-fermentative: (a) 1 Glucose -—-9 1 Lactic acid + Ethanol + 1 002 (b) 3 Fructose.———+ 1 Lactic acid + 2 Mannitol + 1 Acetic acid-+ 1 00 (c) 1 Pentose ————9 1 Lactic acid + 1 Acgtic acid Several investigators have found a marked dis- appearance of hemicellulose (polymers of glucose, xylose, arabinose, mannose and galactose plus mixed sugars and uronic acid) during ensilage, which is believed to be hydrolyzed by both microbial hemicellulases (Goering 22 31, 1970) and by organic acidsat the low pH produced during ensilage (Dewar 33 21° 1963). 10 2, Proteins and Amino Acids In well preserved silages or silages made from aspectically grown forages, about 50-60% of the protein is degraded (Mabbitt, 1951). Many results have confirmed the view that plant enzymes are largely responsible for the degradation of protein (Singh, 1962; Bentley 23 El’ 1955; Kemble, 1956; Henderson gt 31. 1972; 1971b and ' Hughes, 1970a). Clostridia are responsible for the major changes in amino acids during ensiling as stated by McDonald 23 El‘ (1968) (Table 3). Table 3. Some Example of Clostridial Fermentation? Organic acids 2 Lactic acid -—e'1 Butyric acid + 2 C02 + 2 H2 Amino acids (a) Coupled oxidation-reduction reaction 1 Alanine + 2 Glycine ——-) 3 Acetic acid + 3 NH + 1 CO2 (b) Deamination 3 Alanine-—+ 2 PrOpionic acid + l Acetic acid + 3 NH + 1 CO 1 Valine __. l Isobrityric agid + 1 NH3 + l 002 l Leucine-—+ 1 Isovaleric acid + l NH3 + 1 002 (c) Decarboxylation Histidine-—+ Histamine Lysine -———9 Cadaverine Arginine-——+~Ornithine -——6>Putreseine Tryptophan-—yTryptamine Tyrosine -——,Tyramine 3 aMcDonald‘gt‘al. 1968. J. Sci. Fd Agric. 19:125. ll Saccharolytic clostridia varieties multiply and .increase the pH value which leads to growth of putrefac- tive clostridia. Destruction of amino acids is by three main pathways, namely, coupled oxidation-reduction reac- tion(Stickland reaction); deamination and decarboxylation (Table 3). Many decarboxylation products of amino acids have been detected in silage such as: amines, cadaverine, putrescine, histamine, r-amino-butyric acid, p-alanine tryamine and tryptamine (Macpherson, 1962; Macpherson and Violante, 1966; Neumark £5 31. 196&). These subs- tances are of interest because of their possible effect on depression of silage intake and the health of animals (Neumark et 51. 196&; Harris et 21. 1966; McCullough, 1966; Thomas gt El’ 1961; and Okamoto 33 El. 196&). 3, Organic Acids (Non-Nitrogenous) Malate and citrate are the most abundant organic acids in a wide range of plant species. These weak acids and their salts form an important buffer system in the plant (Lessard and McDonald, 1966; Playne and McDonald, 1966; Fauconneau and Jarrige, l95&; Wilson and Tilley, 196&). Both homo and hetro fermentative lactic acid bacteria will readily dissimilate malate and citrate by a number of pathways (Table &). The products formed are either neutral (Acetoin, 2,3-butane diol and ethanol), l2 salts of organic acid or alkaline released cations. Because many of the organic acids are present in the plant material in salt form, their destruction by bac- teria acts against preservation, as decarboxylation results in the release of cations and carbon dioxide. Table &. Main Products of Organic Acid Fermentation by Lactic Acid Bacteria? Homo-and Hetero-Fermentative l) l Citric acid;—e2 Acetic acid + l Formic acid + 1 CO 2 or 2 Citric acids—,2 Acetic acid +.1 Acetoin + & CO2 or 2 Citric acid-493 Acetic acid + l Lactic acid + 3 CO2 2) l Malic acid-—+ l Lactic acid + 1 002 or 2 Malic acid-—9 1 Acetoin + & 002 or 1 Malic acid-—e»1 Acetic acid/Ethanol + l Formic acid + 1 CO2 a’McDonald _e_t 3;. 1968. J. Sci. Fd Agric. 19:125. &. Dry Matter and Energy A number of workers have commented on the apparen- tly high gross energy value of silage (Beever 22 21' 1971; Thomas 33 31' 1969; Barry and Fennessy, 1972; Alderman 23 ‘31. 1971; Waldo 93 31. 1969; Waldo gt 31. 1965). In a complete anaerobic biological system, the increase in gross energy during ensiling can be explained 13 biochemically, based on the known reactions. For example, fermentation resulting in high ethanol production, as in the heterolactic fermentation of glucose, will result in increased energy concentrations where as homolactatic fermentations will have little effect on the gross energy value of silages. On the other hand, in some clo- stridial fermentations lactate will be changed to butyrate resulting in some energy loss as hydrogen from the system but with a comparatively greater loss in DM giving an apparent increase in energy density. McDonald 23 El. (1973) conducted several experiments to demonstrate the high energy recovery under a complete anaerobic system. Their results strongly confirm that energy is recoverable in spite of gaseous CO2 loss. However, if oxygen is introduced during ensiling, recovery of energy_and dry matter become unpredictable based on the biochemical model designed for anaerobic fermentation. Thus, in preserving low—moisture silage or stacked-type hays energy and dry matter recoveries should not be calculated from a chemical reaction scheme but from a balance scheme involving H20 and C02(Pedersed, 1971). In fact, many investigations have reported high correlation ( r) 0.7) between 002 production and DM loss during storage (Honig, 1969 and Zimmer, 1969).’ More discussions about the actions of 02 on silage fermentation are given in part II section III. 1h III. Factors Influencing Fermentation and Nutritive Value offiResultant Silage There are several factors which can affect silage fermentation patterns. Important factors are attributed to plant characteristics e.g. ccntent of fermentable sugar and protein, while others are related to ensiling conditions e.g. type of structure and extent of oxygen exclusion, and still others to mangement systems e.g. fineness of ChOp, use of additives. Many factors are interrelated. Figure 1 illustrate some interrelation- ships among forage, ensiling conditions and quality of resulting silage. Some important factors will be re- viewed briefly here. l. Soluble Carbohydrate Content: A wide range of values have been reported as the minimum sugar require- ments for the satisfactory conservation of grass and legume crops. A general value is about 7% of dry weight (Smith, 1962) although many researchers consider this value insufficient (McDonald 23 El' 196&). The soluble carbohydrate content of the green crop is extremely va- riable depending upon species and environment. For example, the soluble carbohydrate content of orchard grass is considerably lower than that of ryegrass at (any given stage of maturity, while timothy and meadow fescue have an intermediate value (Waite and Boyd, 1953). No similar data are available for grasses and legumes 15 .owdHHm no osH¢> o>HuHhusz cad aoHpapathom wwwHHm .mdoHqucoo waHHHucm .owdhom wqoa< naHanOHuaHom Hahocom .H ohstm .ComOhuHa Hdep no access“ a ma zomonan u HuodHcoEEHomH .cHoa OthpsnomH n ¢m> OmH: uoHuosvoam MHH case pnwflo honoHoHum Una .aowoppH: u z .aouuas hue n :9 N coHuostA EmHHonwuo mouaHooahoun omo oHn smashes huaHfinzHoa-z HuHOHKoe mpqwcHEdh whomcom so wasps“ wcHHHOchoo QpHponQ< humpquo OmH mCOHuoadh noEdm H ouauufiz m>H< acknovashom GHOHsmBHm m N Na :Hopon vHoa oHAhuznomH chaumhz o m z onn< hmno uHoa OHGOHQohm GHmHz :ommm QHSQ poem Leanne hp vHom OHpoo< :HmOHhe :ommm nHaopoo UHom OHELom meopoam hamdm muo>ooop psoprsz moHumpmwmwdm UHoa OHpomA How moudeo: so mOHpOHanca a muHmm cad nonaHEHpm oonHwEon oHuwumHnouoam codeHdooccH mUHoa HanocH: vHoa OHuomA nomOHsHHooHEom UochSoLn quOpo omOHsHHoofiao mmohzocH moshuco pde mocH GOHpanapch .nonOH show Haonhn sz\Zu+ pampcoa ohdumHo Nz-cHouon :Q&.u:opcoo cHouoa - uofihu Hapoa caduauomaop H om3\cHop0h ouaaHH vHow othpdm newshoum nomzv opanchnon ousuuhonao oHoa OHCOHQOL EHuuomaQOpm name oHndHom hops: hHQnfin no»a3 cHoa OHuoo :onsHoNo wchpdo .oz. cOHpauHHtho efioa canoes -aa< npzonw co omapm.‘ hpfiHfiunou Hfiom m thuc¢SG moHoon ugh» HHom moHpmHAouowamno acoHpanoo mOHumHhouomAdno mmmHqucoo owe H ma.“ H H a omauom Ho pGoEGOAHEL 16 in U.S. Temperate grasses accumulate higher concentra- tions of soluble carbohydrate than do the tropical gra- sses(Winmann and Reinhold, 19h6; Ojima and Isawa, 1968; Smith 1968a; Wilson and Ford, 1971 and 1973). Barnett (l95u) found that the fermentable carbohydrate content particularly that of fructosan increases with increasing stage of maturity. Temperature also markedly affects the carbohydrate reserves of green crops. Smith (1970) showed that chan- ging timothy plants at inflorescence emergence from a cool to a warm regime decreased water-soluble carbohy- drate content in the stem bases at early anthesis. The effect of water stress on fermentable carbohydrate reserves has been studied by a number of workers with inconsistant results (Eaton and Ergle, 19MB; Brown.and Blaser, 1965; Blaser et'al. 1966; Buckey and Weaver, 1939; Bailey 23.223 1970). The effects of nitrogen (N) fertilization on carbohydrate reserves are complex and variable (Weinmann, 19h8). Generally, N applied at low to moderate rates increases soluble carbohydrate reserves while nitrogen applied at high rates decreases soluble carbohydrate reserves (Adegbola and McKell, 1966, Izumi gt 2l° 1972; White, 1973). The physiological reasons why changes in N variably effect fermentable carbohydrate reserves are not well understood (White, 1973). Several workers have 17 reported that high N fertilization can produce high NH3 levels, high pH abnormal changes in silage fermentation (Wieringa, 1966; Fox and Brown, 1969), which depress_the intake of the silage (Gordon gt 21. l96h; Castle and Watson, 1969). 2. Mechnical Treatment - Chopping and Laceration: Finely chopping or lacerating as compared with coarse chopping has improved silage quality as judged by pH, percentage of N degraded into ammonia, level of butyric acid, lactic acid and total volatile fatty acids (Murdoch, 1965; McDonald gt a1. 1965 and Dulphy and Demarquilly, 1972,1973). Silo capacity for DM is greater when forage is finely chopped (Dulphy and Demarquilly, 1972,1973). Losses of DM during storage are also lower for finely chopped forage. These effects on fermentation are at least partly due to the liberation of cell contents from the herbage by mechnical bruising, the fermentable carbohydrates in the cell contents providing an immediate substrate for bacteria (Murdoch, 1966). Finely chopped silages are not always more digestible than the coarsely chopped one, but voluntary intake of the latter is much higher (Dulphy and Demarquilly, 1973 and Murdoch, 1965). 3. Ensiling Temperature: For many years English investigators recommonded that temperature during ensiling should be allowed to rise to 32-h2 C on the assumption 18 that this range of temperature is optimum for the growth of lactobacilli (Fry, 1885); but clostridia as well as lactobacilli show optimum growth within this temperature range. In terms of the chemical constituents in the silage no advantage is gained by allowing the temperature in the mass to rise so high, in some cases better pre- servation has resulted when the temperature has been held at 27 C (Murdoch, 1960a). Since the temperature rise in silage is largely the result of heat evolved in aerobic respiration of plant and oxidation processes of aerobic microorganisms any increase in temperature repre- sents nutrients loss in silage (Murdoch, 1960). Further- more, high ensiling temperature will seriously depress protein digestibility (Goering 32 31. 1972). 4. Moisture Content of Forage: Moisture content in the forage significantly influences on fermentation pattern(Wieringa, 1958). When moisture content in a forage is reduced to 65% by wilting, the resulting silage shows desirable fermentation characteristics as compared with unwilted silage: a decreased concentration of bu- tyric, propionic, acetic, lactic acid (not always) and ammoniacal nitrogen (NHB-N), and an increased concentra- tion of water soluble carbohydrates hemicellulose and hot-water-insoluble nitrogen (true protein-N) (McDonald 33 El: 1968; Gordon et 31. 1961; Roffler 32 El: 1967; Thomas et 31. 1969). The decreased extent of protein and sugar breakdown is due to increased osmotic pressure which has selective action upon the silage microflora particu— larly upon butyric acid producing organisms. Wilted silage usually has a pH value of about 5 Which is consi- dered undesirable from standards set for regular high moisture silages. This relatively high pH value is pro- bably due to the following reasons: (1) Decreased produc- tion of lactic and other volatile fatty acids. (2) De- creased production of ammoniacal nitrogen. (3) Increased concentration of protein and cations (buffering agent) in the aqueous phase due to wilting (Wiergna, 1961). Wilting is probably the best process to use in preserving . legume crops. Legume crops are low in fermentable sugars and high in protein content (buffering effect) which make these crops difficult to ensile satisfactorily as direct- cut silage when compared with grasses (Murdoch, 1960; DeVuyst gt a}. 1962; and Watson and Nash, 1960). In.the practical situation, wilting saves transport costs of water for the farmer. In addition, as the crop becomes drier seepage nutrients are reduced. Wilted silage also is more suitable for mechanized handling especially in air-tight silos. The odor of wilted silage is not as objectionable as that of direct-cut silage and this has an aesthetic value to farm families and the general 20 public in contact with persons handling silage. Wilted silage, with its relatively high content of fermentable carbohydrate and true protein nitrogen should produce a more favorable rumen fermentation and nitrogen utiliza- tion by the animal than direct-cut silage (McDonald gt ‘21. 196h; Waldo, 1968). However, results from animal trials do not clearly support this concept (McDonald gt a}, 1968; Thomas 33 El. 1969; Sutten and Vetter, 1971 and Roffler et_al, 1967). Although wilted silages have frequently been evaluated as better quality silage than direct-cut silage, animal performance trials have failed to give parallel results. Animals usually consume more silage DM from wilted silage (20-h0% DM) than direct-cut silage (Thomas .33 El: 1961, 1969; Murdoch, l960b, 196h; Brown, 1962; Halley and Dougall, 1962 and Gordon 33 a1. 1961, 1965), but the production data (digestibility, weight gain, milk production and efficiency) are not always most favorable for wilted silages (Thomas gt 31. 1969; Fisher gt El: 1971; Ruszezyc gt 31. 1972; Forbes and Jackson, 1971; Brown, 1961 and Alder gt El. 1969). Actually, Alder et '31. (1969) commented on wilted silage making "wilting should be adopted if it helps the ensiling process, rather than as a necessary technique to ensure maximum voluntary intake with the doubtful anticipation of higher 21 animal production". The decreased digestibility, produc- tion and efficiency observed in animals fed wilted silages could be due to improper ensiling techniques. Wilting can solve some difficulties commonly encountered during ensiling, but some special precautions are needed when wilted forages are ensiled. For example, farmers should check the leaks in the silos, chop the crop reasonably short, and fill the silo rapidly, etc. (Hillman and Thomas, 1973). The degree of compaction inside a silo is usually determined by the moisture content, length of chopped forage and height of the silo. The wetter and the shorter the forage, the better the compaction. Poor compaction will entrap a significant amount of air which will allow plant cells and aerobic microorganisms to continue their wastful oxidation reactions. The heat generated during oxidation reaction can.reduce the nutri- tive value of a resulting silage (Goering gt 21. 1972). By following correct procedures, Uchida §t_al, (1970) was able to demonstrate that the digestibility of dry matter nitrogen was higher for wilted silage than direct- cut silage (DM digestibility 61 vs. 57%, N digestibility 69 vs. 58%). 22 Part 2. Low-Moisture Silage (Haylage) I. Definition and General Characteristics Haylage is a slightly fermented wilted hay crop silage with a dry matter content of about 50%. Well preserved haylages usually have a light brown color and a very pleasant odor. One example of a comparison of the chemical composition for fresh forage, wilted silage and haylage is given in Table 5. Haylage contains a greater amount of sugar and lesser amounts of organic acids than does wilted silage indicating that less fermentation occurs in haylage than in wilted silage (Table 5). Although.the crude protein content is simi- lar between haylage and wilted silage, the ammoniacal nitrogen fraction is much lower in haylage than in wilted silage suggesting that less protein degradation occurs in haylage than in wilted silage (Table 5). The reduced extent of fermentation and protein destruction observed in haylage ensiling is probably due to the increased osmotic pressure which exerts not only a general inhibitory effect on all silage microorganisms but also a specific restrictive action on butyric acid producing bacteria (Wieringa, 1958). II. Advantages of Haylage Making Only in recent years, in the U.S. has a greater Table 5. Comparison of Chemical Composition Among Fresh Crop, Wilted Silage and Haylage? Freshb Wilted Haylage DMC% 19.00 38.00 51.00 Sugar .20 1.50 3.60 Crude protein 1 .hO 19.60 19.20 Crude fiber 30.90 33.60 32.50 pH d 6.00 u.80 h.80 NH3-N/TN - 15.10 9.90 Butyric acid - 0.79 0.06 Propionic acid - 0.23 0.0h Acetic acid - 2.h3 0.93 Lactic acid — 3.06 2.69 DM loss - 10.20 15.50 aGordon.etHal. 1965. J. Dairy Sci. h8:1062. bFirst cutting alfalfa. c Dry matter. dAmmoniacal nitrogen as a percent of total nitrogen. proportion of the hay crop been preserved as haylage than as silage. Haylage making is often recommended more favorably than is hay making because haylage does not have to be as dry as hay to be taken from the field. This reduces the dry matter losses in the field due to respiration of plant cells, leaf loss, incomplete of pick up by harvesting apparatus and possible rain—damage. Dry matter recovery of field cured hay has been estimated as 73% Which is much lower than the recovery of 87% for wilted silage (Shepherd,gt_al, 195k; Carter, 1960). Dry matter recovery of haylages has not been sufficiently documented, but the average in addition, weather condi- tions necessary for good hay making are never certain; haylage offers an excellent alternative. Also, due to mechanization of the harvesting, delivery and feeding systems much less human labor is required for haylage than for hay. Chemical composition of hay and haylage is similar (Table 6), so the feeding value of well made haylage should be comparable with that of hay. Yet, several feeding trials ranked the feeding value of hay- lage below that of hay (Sutton and Vetter, 1971; Gordon §t_gi, l96h and Roffler 23 al. 1967). Table 6. Proximate Composition of Hay and Haylage? Crude Ether Crude Nitrogen protein extract fiber free ASh extract --------------------- % DM -—-----------—---- Hay 16.18 1.h8 32.65 hl.l8 8.51 Haylage 17.81 2.7h 29.22 39.51 10.72 aData from Gordon‘s: El: 1963. J. Dairy Sci. ue<5>=u11. Haylage should be superior to direct-cut or wilted silage because haylage undergoes a more desirable 25 fermentation pattern (Table l and 5). Gordon.gt‘g1. (196k) observed that the volatile nitrogen content of direct—cut orchardgrass silages ranged from 8 to 37% of the total nitrogen. High volatile nitrogen in silage causes high ruminal ammonia and low nitrogen retention (Waldo and Derbyshire, 1971). Thus, the utilization of nitrogen by ruminants should be better for haylage due to smaller portion of total nitrogen as NHl-N (Gordon, 196h; and Roffler 23 El‘ 1967). Generally, animals tend to consume more dry matter from haylage than from wilted or direct-cut silage. In fact, a linear relationship between intake and silage dry matter content was observed (Thomas et El: 1961 and Gordon 33 El- 1965). But nutrient digestibility and animal production have not been consistently greater for animals fed haylage (Roffler gt a1. 1967, McDonald gt 31. 1968; Sutten and Vetter, 1971 and Hawkins gt El. 1970). The possible reasons for this reduced efficiency in production have not been completely explained but may be partially due to the reduced digestibility of protein and energy in haylage (Gordon, 1968) and discussed in the next section. III. Disadvantages of Haylage Making Since solar energy is the source of energy for 26 removal of water from the forage in the field, there are a interrelationships between general weather conditions and good haylage making. Relatively high dry matter losses (about 15%) during ensiling have been frequently noted when making high dry matter silages (Gordon.gt El. 1965; Murdoch, 1967; and Pedersen et El. 1971) representing gaseous losses (oxidation) and spoilage (mold development). The presence and penetration of air into the silage mass may be modified by the situation outlined below. (1) Forage with a high dry matter content usually has low silage density and hence more air will be entraped. (2) Vari- ation in management factors such as degree of air tight- ness of the silo, fineness of forage chop, the speed of filling the silo, amount of a relatively wetter material in the top area of the silo, covering with a sheet of plastic, distribution of material When filling the silo etc. (Hillman and Thomas, 1973). (3) Variations in aeration when silo is Open for feeding. (h) Air movement caused by diural temperature changes. The heat generated through oxidation reactions in silos can cause two types of problems depending upon temperature. A slight amounts of temperature increase can damage nutrient availability to animals while a large temperature increase can initiate spontaneous ignition. A number of investigators (Cohn, 1890; Miehe, 1930; Festenstein et 21. 1965) have enumerated the se- quential factors responsible for temperature increases sufficient for spontaneous combustion as: (l) respiration of the plant material that occurs to a certain degree until all the material has "died"; (2) microbial metabo- lism that increases temperature as high as 71 C before the thermophilic organisms themselves are killed; (3) chemical oxidation occurs at an ever increasing rate as temperature increases. Temperatures of 70 C readily occur in a silo while spontaneous ignition rarely occurs at that temperature (Koegel, 1971). In fact, the mech- nisms involved in spontaneous ignition are rather compli- cated and still not completely understood (Currie and Festenstein, 1971). In general, biological and chemical factors largely determine whether a sample of haylage is likely to heat, but physical factors will then decide whether the haylage develops a high temperature, Whether this process continues, and what final temperature will be attained. These physical factors are : (l) the quan- tity of crop ensiled is sufficiently large to retain the temperature developed; (2) the crop in a certain physical state where the moisture content is low enough to marked— ly reduce the coefficient of thermal conductivity but still suficiently high to support microbial growth; (3) there is sufficient diffusion of gases to supply the 28 necessary oxygen but insufficient to transfer generated heat away from the mass (Koegel, 1971; Gordon, 1968; Currie and Festenstein, 1971; Gregory gt El. 1963 and Festenstein.gt_al. 1965). Food scientists and organic chemists have known for a long time that sufficient heat can damage food protein during processing (Millard, 1912 and Hodge, 1953). Sufficient heat to damage proteins can result in des- truction of amino acids by oxidation; modification of some of the linkages between the amino acids so that their release is delayed during digestion; formation of linkages that are not hydrolyzed during digestion, i.e. loss of biological availability (Maillard reaction); formation of brown pigments or melanoidins; and loss of palatability (Donoso et_al, 1962; Bender, 1972). Because brown pigments are formed during heat-treatment of a food protein a more general term "non-enzymic browning" has been used to describe the overall reaction (Reynolds, 1963; 1965). The Maillard reaction appears to be the major course for browning development during the heating or prolonged storage of foods and the mechanism follows a common pathway for many foods (Hodge, 1953; Eillis, 1959; Reynolds, 1963, 1965). The primary step involves a condensation reaction between theO(-amino groups of the amino acids or proteins and carbonyl groups of 29 reducing sugars known as the "carbonylamino" reaction and shown in the following reaction: my RNH I HC=0 RNH H 0 CH HC ' _.. ' - 2 '0H)-—-:(Hc':0H) (HCOH)n + RNH2.—-— HCOH.————(HC r— . n_l CH20H (HCOH)n CHZOH HC -————0 CH20H CHZOH Amino Addition Schiffs N-substituted comp. product base glycosylamine Sugar This reaction is then followed by many rather complicated reactions and finally brown pigments are formed. Oven drying of forage or wet feces samples in- creases the yield of lignin, acid detergent fiber (cellu- lose plus lignin), and insoluble nitrogen in acid deter- gnet fiber (ADF-N) (Macdougall and Delong, l9h2; Armitage gt El. l9h8; Van Soest, 1965). The increased apparent lignin and insoluble-N in the ADF was then postulated to be due to the products of non—enzymic browning reac- tion. Results from laboratory work revealed that the reactions involved in forages or feces probably are similar, if not identical, to the Maillard reaction (Van Soest, 1965; Gordon, 1968). Some conditions influencing the extent of formation of ADF-N have been studied in forages (Goering et al. 1973). In general, the reaction 30 was most rapid in the moisture range from 20 to 70% and with temperatures above 60 C but the degree of damage was species dependent. The level of insoluble protein or nitrogen content in the acid detergent fiber in the forages which were suspected to have been heat damaged was found to be highly, negatively correlated with in 1122 nitrogen digestibility ( n = an, r = -.93 )(Goering 23 31, 1972). This observation clearly suggests that increases in the acid-detergent fiber nitrogen value represent a good criterion for estimating the reduction in nitrogen digestibility. There are numerous examples of decreased nitrogen digestibility and/or animal per- formance from animals fed heat-damaged crops (Roffler ‘gt.al. 1967; Zimmerman, 1952; Sutton and Vetter, 1971; Gordon gt 31. 1965; Hill and Noller, 1963; Miller 2t.al. 1967; Hodgson gt 3;. 1935; Wieringa 33 21. 1961 and Bechtel 3133;. 19Ll3, 1915). Wieringa flag. (1961), in a field study, found that the amount of depression of in 3112 nitrogen digestibility was highly correlated with the amount of time during which temperatures were high (number of days above 35 0) although nitrogen di- gestibility was not well correlated with the maximum temperature observed in a silo. Although under farm conditions aeration and heat- ing occur simultaneously, this does not necessarily 31 imply that aeration is required for peotein damage (Millard reaction). According to the reactions known to date, oxygen is not required for production of brown- pigments (Gordon, 1968). Wieringa EE.El° (1961) re- ported that, under laboratory conditions, simply heating silages in sealed jars produced little effect on protein digestibility unless heating was combined with aeration. Since then several workers have conducted many similar types of experiments (Zimmer and Gordon, 196M; Gordon, 1967 and Jorgensen 33 31. 1971) and noted that aeration was not required for heat damage of protein feeds as measured by increases in ADF-N (Gordon, 1967 and Jor- gensen gt 21.-1971). Gordon (1968) suggested that although oxygen is probably not directly involved in the Maillard reaction under practical silage making operations oxygen will be important to the total effect in two ways. Firstly, presence of oxygen retards fermentation of sugars to organic acids; secondly, presence of oxygen allows greater initial heat development and the rate of browning reaction will proceed much more rapidly at higher temperatures. Thus, Browning may be greatly in- creased by aeration, even though it is not required for the primary reaction. A diagrammatic sketch of interrelationships among management factors is illustrated in the Figure 2. 32 dob wow wocmwbmi Homwmmo . bdpmfidmm moos Hommcdo mwwwowosoo cwwbm 09% who bwmfid flabmwowamdllllllk m2+= Smgm obUmUOHdm ifzeEBE bewdwdwbm 6 been mododmosd Humowfidww Swdwomob \ . / ZOHmGCHo oobwoud wHHHHbm mwmom mmdmo ow mwozdfi wwbobomm ow owowvwsm medswdcdwod ow wowmm How oo<0dmmo m owodwo oNHamnHob H modouwo wodaobddeOL T035 33 7 domedwob - HDUHUHdowm Amom wHOdoHSm mum + woacowsm c605 m5o¢m -Hmseesos bddwwmod HHmbe n. mSHUo mowam msmmdm N! U wHOSSHsm bbwuz somedwosm Ll: >Bdwwsa woSfiondcwo @Cmddwww ow mdowwmw HUmHQo ow mHHo Hbmwmomdwdw Homm ow I mdeHmH dewodd -onCooa Swamp Housedwo meCdo m. Uo mwwo mba.d30 005mmnzmbomm ow momwwbm. 33 IV. Laboratory Estimates of Nitrogen Nutritive Value in Heat Damaged Feeds Heat damage may be of little importance with high moisture silage and properly conserved hay. The problem becomes serious with the increased use of low-moisture silage (Donosa gt 21. 1962; Hill and Noller, 1963; and Roffler 23 31. 1967) and artificially dried forages. Laboratory methods used for quantitating nutritive value of nitrogen for ruminants are the solubility of general feed proteins, classification silage nitrogen fractions, and estimation of heat damaged protein (Goering, 1973). Table 7 lists some of these. The more important methods, particularly those recommended for heat-damaged feeds will be briefly reviewed. Table 7. Laboratory Analyses Used to Evaluate Nitro- genous Feeds. General Feeds Silage Water solubility Ammoniacal N or volatile lee rumen fluid base Autoclaved rumen fluid Ionic strength Dilute base Salt solution Pepsin solubility Pepsin + Trypsin solubility Pepsin + Pancreatin solubility Hot water insoluble N Rumen NH3 production Heat Damage Acid-detergent solubility Pepsin solubility Pepsin + Trypsin solubility Available lysine Rumen fluid + Pepsin+ Pancreatin solubility 3A 1. Protein Solubility in Rumen Fluid and Other Related Solvents In order to obtain maximum protein utilization by ruminants, feed protein should be hydrolyzed to NH3 at a slower rate than rumen microbial protein synthesis (Annison et 31. 195k; Chalmers, 1961). The rate of pro— tein hydrolysis in the rumen primarily depends on the solubility of that particular protein. The solubility is, in turn related to the type of protein. For example, albumins are soluble in water while glutelins are in- soluble in water, saline solutions, or alcohol, but soluble in very dilute acids and alkalies. Several re- searchers have shown that properly treating a readily hydrolizable protein with heat or formaldehyde will de- crease the protein solubility in.the rumen and improve nitrogen balance (Goering, 1973; Chalmers gt 31. 195A; Danke 33 31. 1966 and Waldo, 1973). However, When a protein feed is over heated or over treated by formal- dehyde, the protein becomes practically insoluble in the rumen and also resistant to enzymatic hydrolysis in the small intestine of animals (Dinius gt 31. 1973; Brown and Valentine, 1972; and Waldo, 1973). Thus, the relationship between protein solubility and net protein utilization by ruminants is not a linear response, in- stead, it is a curvelinear response (Figure 3, Goering, 1973)- 3S I I I I I I I I I I (D I I I o I I I g I I I as I I I H I I I q; OVGI‘ I over I proper I m heated: treated: treatment: 2; I I I | I I I I I | I I I I I I g. ' heat damage or optimal too solubilized over treated solubility low efficiency Protein solubility Source: Goering, 1973. Proc. Wiscon. Conf. on Use of Lab. Analy. In Feeding Programs. p. 63. Figure 3. A Concept of the Relationship of Nitrogen Solubility of a Feed to Nitrogen Balance in A Rumiants. Autoclaved and fresh rumen fluid, water and different salt solutions have been used as solvents of various feed protein. Generally, solubilities of pro- teins in these solvents are good predictors for ruminants N balance of feeds processed properly. When over treated (by heat or formaldehyde) feeds are incubated with these solvents, solubilities of proteins are good predictors of in 1119 N digestibility but not consistantly related to N retention since in_vitro solubility of protein is 36 related to solubility in rumen but does not resemble the enzymatic hydrolysis occuring in.the samll intestine. Little §t_§g, (1963) evaluated the nutritional signifi— cance of soluble nitrogen in protein feeds (soybean oil meal, heated soybean oil meal, linseed oil meal and corn gluten meal etc.) for ruminants. Relationships among protein solubilities in rumen fluid, water and 0.02 N NaOH, as well as in 11239 rumen.ammonia produc- tion, in 31332 cellulose degradation, in viva digestibi- lity, nitrogen balance, and lamb performance were examined. No definite relation between nitrogen solubi- lity and rate of ammonia production was evident; how- ever, nitrogen solubility in rumen fluid was generally more indicative of ammonia formed that solubility in dilute sodium hydroxide or distilled water. Corn gluten meal and heat treated soybean oil meal (110 C for 2h hours) were particularly low in soluble nitrogen and the N was slowly converted to ammonia by rumen micro— biota and was a poor source of nitrogen for 12:21322 cellulose digestion. However, no apparent differences in protein digestibility and nitrogen retention, feed consumption and growth were detected when lambs were fed regular or heated soybean oil meals. Protein solubility in 0.02 N NaOH was used to evaluate the Opti— mum time for treatment of cottonseed meal by 37 autoclaving (Danke gt El. 1966). With increasing the time of autoclaving, nitrogen solubility decreased only slightly, but in a linear fashion. But, the nitrogen utilization data obtained from sheep responed to the treatment time as positive quadratic relationships. Recently, Wohlt gt_a1, (173) suggested that a uniform method of determining protein solubilities is needed. This method should simulate solubility in the rumen and be readily and easily duplicated in any labora— tory. The method pr0posed by them involves a one hour incubation of the protein feed with the Wise Burroughs (1950) mineral mixture diluted to 10% with distilled water. However, the relationship between protein solu- bilities and in XlX2.N utilizations was not stated. The relationship of feed nitrogen solubility to nitrogen utilization is not always predictable with the present methods; however, the solubility of a given pro- tein does give some indication of its utilization. A more exhustive study is needed (Goering, 1973). 2. Nitrogen Solubility in Acid Detergent Solution An acid detergent solution, such as cetyl tri- methylammonium bromide dissoved in l N Hasou is a rela- tively strong detergent and has been used to solubilize plant protein, other nitrogenous compounds and hemi- cellulose (hemicellulose becomes soluble at a low pH) 38 (Goering and Van Soest, 1970). The residue is called acid detergent fiber (ADF or lignocellulose) and can be subjected to further hydrolysis with 72% H2SOH to yield an insoluble acid detergent lignin (ADL). Not all N compounds, however, were solubilized by the solution. About 7% of the total N was found in the residue from unheated forages (Van Soest, 1965). The amount of the acid detergent insoluble nitrogen (ADF—N) became pro— portionally greater when forages were heated to tempera- tures above 50 C. The relationship between ADF—N and ‘in 2119 N digestion coefficient has been evaluated by Goering gt El: (1972). The correlation coefficient between ADF-N as percent of total N (ADF-N x lOO/N) and in 1113 N digestion coefficient (ND) was r = -.93 and the prediction equation was V(ND) = 72.96 —1.02 (ADF—N x lOO/N) for A8 forage samples. The slope in this equation indicates that every percentage unit increment in ADF-N/N will result in reduction of EE;X3X9 ND by one percentage unit. By using this equation, Thomas and Hillman (1972) estimated that about one—third of the haylage made in the State of Michigan was severely heat damaged regardless of silo type while Goering and Adams (1973) found that h0% of the hay crop silage has severe heat damage in.the State of Pennsylvania. An apparent advantage of measuring_ADF-N for the 39 purpose of in 1112 ND estimation is that $2.X£X9 dry matter digestibility (DMD) also can be estimated since one must obtain ADF before measuring of ADF—N and ADF is known to be a good predictor of in XlX2.dPY matter diges- tibility (Van Soest, 1969). One disadvantage of this laboratory method, is the time consumed in analysis which is 2 to 3 days. 3. Nitrogen Solubility in Acid Pepsin Solution Under optimum conditions (e.g. pHéIu, 37 C) pepsin will hydrolyze protein. The amount of nitrogen solubi- lized by this procedure has been adopted by many labora- tories for estimating protein quality in protein rich foods. Goering et a1. (1972) evaluated the suitability of this method for estimating protein digestibility of heat damaged forages. A correlation coefficient of -.91 between pepsin insoluble N x lOO/N and in 1113 ND was found and the prediction equation was ?(ND) = 85.87 - 0.91 (pepsin insoluble N x lOO/N). Marked similarities in prediction regression equations between pepsin inso- luble N/N and ADF-N/N were apparent although ADF-N/N method was slightly more precise. When pepsin insoluble N/N was regressed on ADF-N/N, a slope (bl) of 1.00 was found. This indicates that both N fractions have the same increase with each increment of heat damage. How- ever, the intercept (bo) was + 17. This means that to about 17% of the nitrogen is pepsin insoluble when acid detergent insoluble is 0 and no explanations were offered (Goering gt El° 1972). The extent of dry matter solubility by incubation with pepsin can also be obtained as well as N and a high correlation between voluntary intake of digestable DM and DM solubility by pepsin has been reported (Donefer gt EL. 1966). When DM solubility as well as N solubility is determined using pepsin than an additional advantage to the use of this procedure in forage analysis would be apparent. h. Nitrogen Solubility in Various Combinations of Proteases Solutions Other protease hydrolysis incubation procedures have been proposed to study and measure protein dena— turation resulting from food and forage processing methods. Saunders et_al, (1973) implied that a two enzyme sequence would approximately simulate the gastro- intestinal tract digestion process. The use of pepsin and pancreatin in sequence has been developed for protein quality evaluation based on the composition of amino acids released after digestion (Akeson and Stahmann, l96h). Papain hydrolysis has also been used for protein quality evaluation, but the ranking of quality by this method was not as well correlated with in vivo ND as the pepsin : pancreatin method (Buchanan MI and Byers, 1969)- Recently, Saunders gt El. (1973) compared the correlation coefficients for leaf protein between pro- tein digestibility measured in rats and two sequential incubations using pepsin : trypsin or pepsin : pancrea- tin. A slightly higher correlation coefficient was noted for the pepsin : trypsin sequential digestion value than for the pepsin : pancreatin value ( r = .91 vs. .87). Papain hydrolysis was also performed but no value reported. Generally, the digestion coefficients calculated from papain hydrolysis were low and not well correlated with the in 111g values. 5. Nitrogen Solubility in Rumen Fluid, Pepsin and Pancreatin Solutions ngyitrg rumen fermentation (A8 hours) followed by pepsin digestion (h8 hours) is a widely used labora- tory technique to estimate the in Kiyg_dry matter digestibility (DMD) of feeds for ruminants (Tilley and Terry, 1963). This method precisely predicts DMD and has become the method of choice used to predict animal digestibility by animal nutritionists and plant breeders (Van Soest, 1973). Recently, Antongiovanni et 31. (1971) evaluated this procedure and a similar procedure to predict in 22KB nitrogen digestibility. They expected a closer simulation to in Kilo condition by using pancreatin incubation sequential to the regular h2 two-stages of the lg XiEEE incubation. Additional amounts of protein were hydrolyzed when pancreatin was added after the pepsin digestion, but the degree of this additional hydrolysis varied among samples, and most unfortunately, their samples did not have in XEXE nitro- gen digestibility (ND) values. Thus, evaluation of the precision of this method in predicting in viva ND is impossible. The most apparent drawback for this pro- cedure is the long time involved for three sequential incubations. A complete analysis requires about six days. 6. Hot-Water Insoluble Nitrogen The fraction of N insoluble in hot water has been used to estimate true protein nitrogen in forages or silages. The forage sample is boiled in water, which coagulates the protein causing it to become water in- soluble and remain in the cell of the sample (Goering and Van Soest, 1970). This procedure has been routinely used for silages to fractionate nitrogen into a protein and non-protein-nitrogen (NDN) fraction (Waldo and co- workers, l973a,b,c). For example, they found.that hot water insoluble nitrogen decreased during the fermenta- tion period of direct-cut silage while in formic or formaldehyde treated silage hot water insoluble nitrogen tended to be higher. Although no correlation coefficient MB was given, the N balance data were generally parellel with the amount of hot water insoluble N in direct-cut or formic acid treated silages, but in formaldehyde treated silages a less definite relationship was found (Waldo, 1973). This procedure has a practical advantage over others in that it is rather simple, fast, inexpensive and requires little equipment. 7. Chemical Determination of Lysine Availability Severe overheating of protein supplemental feeds results in seriously depressed availability of all amino acids with lysine apparently more heat sensitive than some of the other essential amino acids (Meade, 1972). There are several chemical method available to measure the lysine availability in proteins of animal, cereal or oil-seed origin based on quantitative reactions of the fi-NH3 in the lysine molecule. Exampillary methods are: l-fluore-2,h-dinitrobenzene (FDNB) method (Carpen- ter, 1960); 2,h,6-trinitrobenzenesulfonic acid (TNBS) method (Tsai gt El: 1972); alkylation method (Finley and Friedman, 1973), remazol brilliant blue R method (Pruss and Ney, 1972), and modified FDNB method (Blom gt_al, 1967). Comparable results have been reported when different methods were used and they are all correlated with the protein utilization by young AA monogastric animals (Finley and Friedman, 1973). To date these measurements and comparisons were limited only to certain typical protein sources (e.g. bovine serum albumin, casein). All these methods depend on development of a color at a specific wavelength, accurate spectrophotometric measurements on extracts of plants or plant proteins and especially heat damaged plant pro— tein should not be expected because of interfering agents such as plant pigments, carbohydrates hexosamines and brown pigments (Maillard reactions) (Allison gt 31. 1973; Blom gt 21. 1967). At the present time, a method developed by Finley and Friedman (1973) is probably the most reliable for estimating available lysine in samples high in carbohydrate and pigments. However, this method has a serious drawback since it requires an expensive amino acid analyzer. Perhaps because of analytical difficulties, there are no data in the literature con- cerning the relationship between lysine availability and protein digestibility of forages by ruminants. 8. Degree of Non-Enzymic Browning Using laboratory model systems, the brown pigments produced by the Maillard reactions can be quantitated by measuring the absorbance of an extract at about A20 nm (Reynolds, 1965; Keeney and Bassette, 1959; Eichner and Karel, 1972; Pokorny and Janicek, 1971; Janicek and 1+5 Pokorny, 1970; Pokorny 32 21. 1973a,b,c,d; Baloch gt 21. 1973; Yanagita gt El. 1973). For example, Yanagita 22 21, (1973) reported that when mixture of casein and ethyl loinoleate was subjected to heat treatment, the degree of browning (Eh3O/16 mg N/lO ml) increased with increasing reaction temperatures while the quantity of available lysine and digestibility in 31332 by pepsin and trypSin decreased with increasing reaction temperatures. Browning pigments occur widely in processed foods, but special separation and isolation procedures are needed before the degree of browning can be accura- tely measured. This is because of the strong inter- ference by various pigments. For example, carotenoid pigments in dehydrated carrot have high absorption near wavelengths of the browning products (331°Ch.22.2l° 1973; Hendel et_§1, 1950). Unfortunately, there is no special procedure developed for measuring the degree of browning in heat—damaged hay crops. Farmers and animal nutritionists usually estimate extent of heating by odor and sight of the sample. This is obviously not as accurate as the other laboratory methods (ADF-N or pepsin solubility). In fact, Goering and Adams (1973) found that several haylage samples with two ADF-N values were much darker in color than expected based on their ADF-N value. A6 V. Methods Available to Prevent Heating and Spoilage in Haylages Under farm conditions, heating and spoilage in haylage are due to the presence of oxygen and the most effective means to solve such problems would be to exclude air. This can be partially acomplished by good management practices e.g. chop forage to reasonably short, fill silo rapidly, cover the top with a plastic sheet, (Hillman and Thomas, 1973) and by using oxygen- limiting silos. Alternatively, one can use preservatives to prevent the heating and spoilage by restricting respi- ration of plant cells and microbes. There are several chemicals having fungistatic and fungicidal properties such as volatile fatty acids, sorbic acid, benzoic acid, sulfur dioxide etc. that might be added to ensiled forages (Nickerson and Sinskey, 1972). Any useful chemi- cal should have the following characteristics: (1) prevent spoilage and deterioration over a wide range of moisture and temperature conditions at application rates which are economically competitive; (2) be palatable and nontoxic to livestock and leave no residue in animals fed the material; (3) be reasonably safe to handle and require a minimum of special equipment for application, distribution and storage. According to Sauer (1973), the best mold inhibi- tors, considering efficacy, cost and safety are M7 propionic acid and closely related compounds. These include propionic, acetic, formic and isobutyric acids, and a few salts such as sodium propionate and ammonium isobutyrate. The salts are less effective pound for pound, but are relatively non-corrosive and safer to handle compared to the acids. The mode of action of these acids is not known for certain. The lowered pH, dissociated and undissoci- ated molecules of the acid all play important roles in preservation. They may destroy cell membrances or inhibit some microbial enzyme systems concerned with metabolism or reproduction (Nickerson, 1972). The effectiveness of various preservatives in preventing spoilage of silage and high moisture grain has been studied and results reported recently (Sleiman, 1972; Britt, 1973; Waldo gt El: 1973a,b; Fox 32 21. 1972; Henderson 33 EL: 1972; Huber, 1970; Bade gt 21. 1973; Jones, 1970; Papendick and Singh-Verma, 1972; Sauer, 1973; Thomas gt 21. 1973). However, pr0per comparison of effectiveness of different preservatives is difficult because lack of consistency in dosage rate, testing material, moisture level and storage conditions. The preservation value of different chemicals has inconsis- tant rankings under different conditions (Goering and Gordon, 1973). Thus, for illustrative and comparative A8 purposes an experiment which tested a large number of preservatives under the same storage condition should be selected. One experiment conducted by Sauer (1973) was choosen for this purpose although the material used was wet shelled corn grain (Table 8). Table 8. Mold-Free Storage Time (Weeks) for Corn With 2A% Maisture Content Stored at Approximately 23 C. Chemical Applied Application Rate (1. 135—— . Methylene bis propionic acid . Pro ionic acid 800 Propionic acid:20% Acetic acid 80% Propionic acid:20% Formic acid 80% Propionic acid:20% Isobutyric acid 50% Propionic acid:50% Isobutyric acid 20% Propionic acid:80% Formic acid 50% Propionic acid:50% Formic acid 50% Propionic acid:50% Acetic acid 20% Propionic acid:80% Isobutyric acid Ammonium Isobutyric acid (AIB) Isobutyric acid Acetic acid (glacial) Formic acid . 20% Propionic acid:80% Acetic acid . Ammonium acetic acid 17. None MN |'\)(\) +-+ I I II I \ptr -QCD¢%O HIJFJHFJFJH OWntruMbtscro(DHJOWHITOHuIA OCDCDHFJRDerhJFWpCbOVQCD aSauer, D.B., A report from U.S. grain marketing research center Agricultural research service. USDA. October, 1973. Based on several such trials the authours conclude that: (1) salts are inferior to the acids form, (2) A9 propionic acid and methylene bis propionic acid are superior to other related acids, (3) there are no synergistic effects using acid mixtures. Results from this experiment are somewhat contradictory to a general believe that antimicrobial activity of the fatty acids increases with increasing chain length. Butyric acid is, for example, more effective than propionic acid (Galbraith gt 31. 1972). Although butyric acid was not tested in Sauer's study, the large difference in potency of mold growth inhibition between propionic acid and isobutyric acid obviousely can not be interpreted from the chain length point view. Acetate and propionic are naturally occurring compounds in silages and in the rumen, and are a readily digestible energy source. They are intermediate products in carbohydrate and protein digestion and are produced in particularly large quantities in ruminants. The quanti- ties consumed in acid treated forage or grain would not be sufficient to disturb normal acid levels in the digestive system (Sauer, 1973). In a preliminary study Thomas gt 31. (1973) evaluated several preservatives for mold inhibition using haylage ensiled in barrels (A2-6l% DM). Results were comparable to Sauer's study. Propionic acid and a mixture of acids (propionic acid 80% : Acetic acid 20%) 50 were very effective in reducing temperature during a 55 day period. Ammonium propionate was less effective than propionic acid at the same molor application rate while formic acid was almost completely ineffective. Voluntary intake of propionic acid treated haylage was much greater than the control when fed to sheep. No marked effect of acid treatment was noted on dry matter digestibility. Similar results about the value of acids were found by Sleiman (1972) using partially wilted rye grass (28% DM) with minimum conpaction in piles and barrels. In a recent report Goering and.Gordon (1973) evaluated a number of preservatives under laboratory conditions. Their results ranked the effectiveness of chemicals in retarding mold growth of alfalfa haylage in following order: propionic aCid.>ammonium isobutyrate) sodium propionate > acetic : propionic (50:A0) acidS). formalin >sodium chloride. There were some indications that formic acid reduced NH3-N content in unwilted silage (18% DM) more effectively than did propionic acid. The lactobacilli counts were also higher in formic acid treated silage than in propionic acid treated silage. However, in wilted silage (38% DM) the reverse situation was found with propionic acid producing the better results (Papendick and Singh-Verma, 1972). Nevertheless, marked improvements 51 in intake, dry matter digestibility and nitrogen utiliza- tion have been reported when propionic acid was applied to unwilted Italian ryegrass (20% DM) (Cottyn gt 31. 1972). Formic acid has been successfully used as an addi- tive to direct—cut grass silage in Norway and Great Britain (Saue and Breirem, l969a,b; Castle and Watson, l970a,b; Henderson and McDonald, 1971; Carpintero gt 31. 1969). In the United State, thorough and systematic evaluations of formic acid as hay-crop silage additive have been conducted only by Waldo and coworkers (Waldo §t_al, 1966; 1971; l973a,b,c,d; Derbyshire 33 El: 1971). The possible actions of formic acid on silage fermenta- tion and animal production are: 1) inhibition of plant protease activity after cutting, thus stoping protein degradation; 2) immediate decrease in silage pH to A.2 which is favorable only to lactic acid bacteria; 3) exertion of selective pressure on clostridia which reduces the production of butyric acid and ammoniacal nitrogen and preventing lactic acid degradation; A) reduction in dry matter loss during ensiling; 5) slight improvement in voluntary intake; 6) slight improvement in digestibi— lity; 7) significant improvement in efficiency of animal gain and 8) significant improvement in N balance (Waldo, 1973). However, equal milk production for cows fed formic acid treated direct-cut silage (.5%, w/w) and 52 control silage was reported by Thomas gt 31. (1969). In general, formic acid treated unwilted silage is superior to unwilted control, wilted control, un- wilted molasses treated, unwilted sodium metabisulfite (Na28205) treated unwilted silages (Carpintero gt a1. 1969; Waldo, 1973; Henderson.§t.§l. 1972). However, Reid (1971) reported the metabolizable energy values of formic acid treated and untreated control were not significantly different. In respiration calorimetric experiments using sheep, the proportions of the metabo- lizable energy stored as new body tissue were 53.A% for the untreated silage and 5A.6 for the formic acid silage. Thus Reid (1971) concluded that the main advantage of formic acid treated silage was increased consumption of the treated silage. When formic acid was added to wilted silages improved recovery, gain and efficiency were found (Derbyshire_et.al. 1971). Since formic acid did not reduce spoilage of haylages in barrels (Thomas .et‘al. 1973) its use as haylage additive is not promising. Recently, formaldehyde has been reinvestigated as a direct-cut or wilted silage additive. Formaldehyde may possess two advantages over formic acid such as (1) reduced cost ($5 vs. lO/ton forage DM) and (2) improved protein utilization. Formaldehyde binds with forage protein to protect the protein from microbial degradation S3 in the rumen and allow subsequent digestion of the protein in the abomasum and intestine resulting in more efficient use of ingested N (Ferguson.g£ El: 1967; Ferguson, 1970; Waldo, 1973). However, when too much formaldehyde is used, the forage protein will be over protected and even resistant to enzymatic hydrolysis in the small intestine. For example, when alfalfa (27% DM) was treated with formaldehyde at the rate of 16.6 g per 100 g crude protein when ensiled voluntary intake as well as dry matter and nitrogen digestibilities were significantly reduced as compared with values for un- treated silage (Brown and Valentine, 1972). When formaldehyde was applied to a grass-legume mixture (20% DM) at a lower rate of 2 g per 100 g crude protein plus about the same amount of formic acid, the silage had less protein degradation and more net lactic acid pro- duction in the silos than did the control (Waldo EE.§1- 1973c). Nitrogen digestibility was slightly depressed (63 vs. 65%) by treatment, but daily nitrogen retention was significantly increased (31.3 vs. 25.7 grams). Daily gain by heifers was increased from 6A3 to 750 grams by this treatment (Waldo 32 El: 1973). The authors concluded that formaldehyde at this concentration practi- cally stopped nitrogen degradation during silage fermen- tation and aided protein by-pass through the rumen Sh producing increased nitrogen retention. 0n the other hand, Thomas (l96A) found no significant differences in DM intake and daily gain between dairy heifers fed formaldehyde treated direct-cut alfalfa silage (2 g formaldehyde/100 g forage protein) or untreated silage. The potential use of formaldehyde as silage additive, particularly for legume crop silage, appears optimistic for direct—cut silage. However, there is no data available concerning formaldehyde for haylage preserva- tion. However, application rate will be critical in its effect on ND. Although ammonium isobutyrate (AIB) was classed as ineffective in preserving high moisture corn grain in a study by Sauer (1973). Goering and Gordon (1973) on the other hand, found that AIB was superior to propionic acid in inhibiting of heat and mold development in hay- lage. In addition to the effect on inhibiting mold development, AIB has nutritional significances to ruminants. It contains 8A% crude protein equivalent which will contribute to the total nitrogen content in the ration. The carbon skeleton of AIB is a branched four carbon moiety which has been considered as one of the important growth factors for cellylulytic bacteria in the rumen (Bryant, 1973). Ammonium isobutyrate also has some convenience in field application being rela- tively non—corrosive and practically odorless. 55 Another chemical, sulfur dioxide (SO theore- 2). tically should be an effective haylage preservation. Sulfur dioxide or the sulfurous acid salt (e.g. Na28205) is widely used as an anti—microbial agent to preserve various types of food (e.g. fruits and wine) (Nickerson and Sinskey, 1972). Sulfur dioxide is a multi—functional preservative acting to (l) inhibit the non-enzymic brow- ning reaction; (2) inhibit various enzyme-catalyzed reactions, notably enzymic browning; (3) act as a anti- oxidant and a reducing agent; (A) inhibit and control growth of microorganisms (Dunn, 1956; Nickerson and Sinskey, 1972). In practice, when Na2S205 was applied at low levels (S A% wet basis), no significant improvements were found in silage quality, storage recovery or animal performance (Carpintero et 51. 1969; De Vuyst gt_§l, l967a,b; Zelter, 1961; Murdoch egg. 1956; Levitt 3331. 1962). However, when Na28205 was added at higher levels, silage quality was improved to some extent, but silage voluntary intake was significantly reduced (Levitt 33 El. 1962; Cowangt_a_l_. 1952; Allred 9351;. 1955). Alderman gt 31. (1955) and Cowan gt_gl. (1953) both commented on the use of Na28205 for high dry matter silage. They suspected that the instability of Na28205 in treated silage was due to the progressive decomposi- tion of N82S2OS through oxidation reactions. 56 There are other direct-cut silage additives which can be divided into two groups. The first group includes materials that will stimulate lactic acid fermentation such as molasses, sugar, glucose, apples or cereal products. However, these additives generally have no direct effect on clostridia and they can be beneficial to silage fermentation only under strict anaerobic con- dition since sugars are not only good substrates for fermentation but also an excellent energy source for oxidation by molds or fungi (Wieragne gt 21. 1961). The second group includes various kinds of strong mineral acids which directly increase the acidity of the silage and hence depress all microbial activity, particularly the clostridia which are sensitive to low pH values. Strong mineral acids (e.g. 2 N H01 or 1A N H280“) were used successfully during the early 1900s (Virtanen, 1929). However, the uses of strong mineral acids and phosphoric- sulfuric mixtures has been gradually replaced by weaker organic acid for several reasons. The improvement of harvesting techniques such as fine chopping of silages made it possible to replace strong mineral acids by weaker acids, because the acid can react more easily with chopped or crushed crop. According to Virtanen (1969), mineral acids exert their silage preservation effects only by reducing pH and have no specific anti-microbial 57 properties. The mineral acids gradually react with the basic components of the crop and thus the final acidity of the silage depends on the original content of organic acids. On the other hand, the preservative effect of formic acid is partly due to a reduction of pH, and also to a selective anti—bacterial effect. There are several instances suggesting that animals developed metabolic acidosis when they were fed silages treated with mineral acids (McCarrick 22 El- 1965; L'estrange and Murphy, 1972). Also, in field situations, strong mineral acids are more difficult to handle than weaker organic acids. In addition, several organic acids possess special fungistatic or fungicidial properties which add to their use fulness thus suitable as haylage additive, while mineral acids have no such selective action. This review has indicated that several organic acids (e.g. propionic acid) can be applied to direct- cut silage at rather low levels yet still obtain com- parable or even better results than those obtained by using sugars or mineral acids or both. In fact, Saue (1968) has demonstrated that formic acid treated silage was better than molasses or mineral acid treated silages for lambs. In conclusion, this literature review indicates that the ND of haylage is commonly lower than that of 58 hay or direct-cut silage. This decreased ND can not be detected by the traditional total N analysis. Although a few laboratory methods have been used successfully to estimate ND of haylages, other laboratory methods should be evaluated. Many antifungal agents have been demonstrated to be benefial in preserving nutritive value of direct-cut silage. However, only limited experiments have been carried out using haylage. Thus, the present experiment was designed: l) to evaluate several chemical and in [11:59 laboratory methods to estimate the nutritive value of haylages; and 2) to evaluate propionic acid, ammonium isobutyrate and formaldehyde in preserving nutritive value of alfalfa haylage. MATERIALS AND METHODS Part 1. Evaluation of Forage Protein Quality by Laboratory Methods I. Forage Samples Twenty—four forage samples that have been studied for various experimental objectives by research personnel in the Michigan State University, Dairy Science Depart- ment were selected for the present study. Among these 2A samples, eight were preserved as staCked alfalfa hays, eight were preserved as alfalfa.haylages in cement silos, three were preserved as alfalfa haylages in pilot type silos (about 50 kg fresh haylage capacity) and five were preserved as sun cured alfalfa hay (Table 9). At least A samples were suspected to have been heat damaged based on data of storage temperatures, un- usually low voluntary intake, nitrogen(N) and dry matter (DM) digestion coefficients, while the other 10 samples were considered normal since they had normal nutritive values when fed to sheep. Each sample had values for proximate analysis (crude protein, ether extract, crude fiber, N free extract and ash), fibrous constituents analysis (cell wall constituents, acid detergent fiber, hemicellulose, cellulose and lignin) and sheep 59 60 .mcwee mam mHmmflpcoAdm CH meSHw> n .aoppws hams Am.mv AH.mmv AHN.®V Ao.H:v Am.o:v Am.wv Ao.omv m.©I®.MH b.0HIH.®m N.HIH.:H m.HmIO.mN O.mMIN.Oo o.©I®.OH H.mmIo.®m IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII an R IIIIIIIIIIIIIIIIIIIIIIIIIII------II--- n... Amv omehmn IOHHm poaflm thpNHE Amy emeth mmwwm IOHHm pdoEoo ImHHwMHm A®.®mv Awm.mv AH.ONV DANNV Amy hmfl Umxowpm ho m.mm-m.m: mo.o-mm.m c.mm-m.oH w:-sw Amvses ooaSoIssm -seHseHs :m .................. so a ------IIIIIIIIIII pomauwo newspxo cHopoam vow mm mommfiwM>WOmwmn mmao.m eHmem oeamIz aenpm 06590 dig a o.pmo Mahem . o m .oc proe .EopH Seam pom Q0>Ho oas .sssaaaz .sosem seHHansHom seasons as some moSHw> use: mmoaw paw Edewaz momwaoa pom mode> HwOthch¢ .o eHan 61 performance (maximum intake, digestion coefficients of major nutrients and N utilization) as shown in Appendix Table 1,2,3, and A. The maximum, minmum and mean values of forage constituents and sheep performance data are summarized in Table 9 and 10 respectively. Table 10. Digestion Coefficients, Nitrogen Utilization and Maximum Intake of 2A Forages Used in Protein Solubility Experiments. Items Range Mean Digestion Coefficients (%) Dry Matter A3—70 57 Organic Matter AAe7l 59 Cell Wall Constituents A0-62 52 Acid Detergent Fiber A2-56 A8 Nitrogen (N) Utilization Digestibility (%) A0-77 6A N-balance (g/day) -2.8-9.1 3.5 N-balance x lOO/N absorbed -8.9-28.3 15.5 Maximum DM Intake (%BWa) 1.91-A.6A 3.52 8Body weight. These samples were deemed sufficient in number, range in analytical values and digestibility to be representative of these existing on farms and adequate to characterize protein quality. Forage samples were ground through a 1 mm screen 62 of a Wiley mill and stored in sealed glass jars. In addition to the 2A samples obtained from this Department, Dr. H.K. Goering of USDA Maryland kindly supplied data for AA forage samples which in turn, were furnished by several experiment stations (South Dakota, Nebraska, Iowa and Utah). These samples had complete proximate analyses, nitrogen solubilities in acid de- tergent and acid pepsin solutions, and.l£.KlX2 nitrogen digestion coefficients. Also, data for 18 forages were supplied by Dr. N.A. Jorgenson of University of Wiscon- sin and Dr. D.C. Pierson of University of Minnesota. Data obtained included acid detergent insoluble nitrogen fractions and lE.XlXE nitrogen digestion coefficients. Data supplied from the other sources were combined with the data of this study and used for predicting in 1313 N digestion coefficients by acid detergent insoluble N as a percent of total N. II. Laboratory Methods Used to Evaluate Forage Protein Quality A. Protein Solubility 1. Solubility in hot water: Two gram sample of air dry, finely ground (passed 1 mm screen, Wiley mill) forage was boiled and refluxed with distilled water for one hour and insoluble material was recovered by filtra- tion (Whatman filter paper No.5A, 15 cm). Content of 63 dry matter (oven drying at 90-100 0 for 2A hours) and nitrogen (macro-kjeldahl procedure) in the insoluble solids were determined. Percent of hot water insoluble nitrogen (N) was calculated either as a proportion of total dry matter (DM)(Equation 1) or as a preportion of total N (Equation 2). Hot water insoluble N, %DM = Gram N x 100/oven—dry sample weight (g)----(1) Hot water insoluble N, %N = Gram N x 100/gram N in oven-dry sample weight-(2) Hot water soluble N was calculated as the differ- ence between total nitrogen and insoluble nitrogen (Equation 3). Hot water soluble N, %DM = Total N, %DM - Hot water insoluble N, %DM ------ (3) In addition, hot water soluble dry matter was also calculated (Equation A). Hot water soluble DM,% = 100 - (W - W /oven-dry sample wgight x (100)---(A) Where: WP= Weight of filter paper plus insoluble residue; W = Weight of filter paper. t 2. Solubility in acid detergent solution: Two gram sample of air dry, finely ground (passed 1 mm screen, Wiley mill) forage was boiled and refluxed with acid detergent solution (1 N H2SOA containing A9 g cetyl trimethylammonium bromide/liter) for one hour according to procedure outlined by Goering and Van Soest (1970). Acid detergent insoluble residue was recovered by 6A filtration (Whatman filter paper, No. 5A, 15 cm) and analyzed for dry matter (oven drying) and nitrogen (macro-Kjeldahl procedure). Calculations of insoluble N, soluble N and soluble DM were the same as used for the hot water solubility. 3. Solubility in diluted phosphate-bicarbonate mineral buffer solution: An aliquot of forage sample containing 25 mg N was weighed and transferred into a 200 ml centrifuge bottle. Phosphate-bicarbonate buffer solution was prepared according to Burroughs (1950) and this original buffer solution was diluted ten fold with distilled water ( 1 part original buffer + 9 parts distilled water). One hundred m1 of this diluted buffer solution was added to the centrifuge bottle containing the forage sample. Bottles were then immersed in a 39 C incubator for one hour. After the incubation period, the solution was centrifuged at 15,000 x g for five minutes. Extracted nitrogen was determined on a 5 ml aliquot of the super- natant by the macro-Kjeldahl methods Percent total solu- ble nitrogen (N) was calculated as mg N extracted in the 100 ml divided by the 25 mg N. B. In Vitro Protein Digestion 1. Acid pepsin digestion: Fifty m1 of acid pep- sin solution (0.01% w/v pepsinA in 0.1 N HCl) were APepsin: 1:10,000. Nutritional Biochemicals, Cleve- land, Ohio. 65 added to a 125 m1 Erlenmeyer flask containing 0.5 g air dry, ground forage sample. The mixture was incubated at 39 C for 20 hours (Goering and Van Soest, 1970). The indigestable material was recovered on Whatman filter paper No. 5A, 15 cm and analyzed for dry matter (oven drying) and nitrogen (macro-kjeldahl procedure). Percent of pepsin insoluble N, soluble N, and soluble DM were calculated according to similar equations as used in the hot water solubility procedure. 2. Acid pepsin and pancreatin digestions: These pro- cedures were outlined by Saunders gt 3;: (1973). In a 50 ml centrifuge tube, 500 mg of air dry, ground forage sample was suspended in 15 ml of 0.1 N HCl containing 1.5 mg of pepsin and incubated at 37 C for three hours. The solution was neutralized with 0.5 N NaOH and treated with A mg of pancreatinA in 7.5 ml of 0.2 M phosphate buffer pH 8.0. The mixture was incubated for an addi- tional 2A hours at 37 C. The solids were separated by centrifuging (20,000 x g for 5 min.) and washing with water (5 x 30 ml). The solids were finally filtered through a 1.2Jp.filter (Millipore), oven dried, weighed, and analyzed for nitrogen (macro-Kjeldahl procedure). Calculations of insoluble N, soluble N and soluble DM were similar to those described for the hot water solubi- lity procedure. APancreatin, Grade III, Nutritional Biochemicals, Cleveland, Ohio. 66 3. Rumen microbial and acid pepsin digestions: Fresh rumen fluid was obtained from a fistulated cow, two hours after a morning feeding of mixed alfalfa-grass hay. Rumen fluid was filtered through four layers of cheese— cloth, and the filtrate was then immediately added to tubes containing forage samples and a mineral buffer solution (modified Terry and Tilley method, 1963; out- lined by Goering and Van Soest, 1970). The tubes were then capped with a rubber stoper fitted with two openings: A bunsen valve and a gassing tube connected to a common manifold. The manifold was connected to a supply of carbon dioxide. The tubes were gently gassed with 002 during the first 20 min of a A8 hour incubation at 39 C. At the end of this incubation period, rumen microbial digestion was stopped by adding 2 m1 acid pepsin solution (.5 g pepsin in 6 N HCl) and the acid pepsin digestion continued for A8 hours at 39 C. The insoluble material was isolated by filtration (Whatman filter paper No. 5A, 15 cm) and analyzed for DM and N. Calculations of inso- luble N, soluble N and soluble DM were made as described for the hot water solubility procedure. A. Rumen microbial, acid pepsin and pancreatin digestions: Procedures involved in the first two stages for rumen microbial and acid pepsin incubations were practi- cally the same as those described in the previous method. 67 after a total of 96 hours of rumen microbial and acid pepsin digestions, the incubation mixture was neutra- lized with 0.5 N NaOH and treated with A mg of pancrea- tin in 7.5 ml of 0.2 M phosphate buffer pH 8.0. The mixture was incubated for an additional 2A hours at 39 C. Indigestible material was filtered on Whatman filter paper No. 5A, 15 cm, and analyzed for dry matter (oven drying at 100 C for 2A hours) and total nitrogen (macro- Kjeldahl procedure). Percent of insoluble N, soluble N and soluble DM were calculated using equations similar to those described for the hot water solubility procedure. 5. Rumen ammonia release: Experimental procedures for quantitating the amount of NH; release were essentially those used in the rumen fluid-acid.pepsin incubation except that incubation time was 3 hours and no urea was added to the mineral buffer solution. Rumen microbial digestion was stopped by addition of 1 ml saturated mercuric chloride (HgClZ). The incubation mixtures were centrifuged at 25,000 x g for 10 min. and duplicate 5 m1 supernatant aliquots were assayed for ammonia nitrogen by steam distillation (AOAC, 1965). Total ammonia nitro- gen production was expressed as a percent of total nitro- gen in the sample. 6. Degree of browning: To estimate the extent of "browning" a 500 mg forage sample was incubated with 68 50 ml phosphate buffer at pH 7.8 containing 3.5 mg pronaseA at 37 C for 5 hours. The solution was then filtered through a 15 cm filter paper (Whatman No.1) and the filtrate was made to 50 ml. One ml of this diluted filtrate and 3 m1 of distilled water were pipetted into a photometer tube. After thorough mixing, the optical density of this diluted solution was read in a spectrophotometerB at a wavelength of AAO millimicrons. III. Statistical Analyses A total of 31 variables were determined for each sample. These variables were also combined into 13 groups based on laboratory operations. Simple linear regressions of five in £133 parameters (DM and N diges- tion coefficients, N balance, N retention as a percent of absorbed N, and maximum DM intake) on each of 31 variables were calculated. Five in 1119 parameters were also regressed on each of 13 groups. Multiple and linear regressions were calculated for the amount of digestible nitrogen (g/100 g dry matter) on selected various nitro- gen fractions. ‘In £112 parameters were then regressed on variables obtained from random combinations of any two groups. APronase. A5,000 PUK/g. B grade, Calbiochem. San Diego, California. BColeman model Junier II., Coleman Instruments Co., Maywood, Ill. 69 The next step was regressions selected variables derived from the combinations of more than three groups with in_vivo responses. A least square deletion program was combined with multiple regression analyses. Part 2. Haylage Preservation With Propionic Acid, Ammonium Isobutyrate and Mixture of Ammonium Isobutyrate and Formaldehyde I. Ensiling Techniques First crop alfalfa (85%) was harvested in the late bud stage of maturity. Forage was cut, crushed, and windrowed with a A m swath mower. When forage moisture content was reduced to about 50%, forage were picked up, chopped and blown into front unloading wagons with a field forage harvester. The chopper was set for about 1 cm theoretical length of cut. Each loaded wagon was weighed on a platform scales and empty wagons were weighed several times during the filling operation. Forage samples were obtained at the blower for DM analysis. The assignment of silos, dates, treatment and amounts ensiled are given in Table 11. The amount of additive (Table 11) for treated forages was calculated to the nearest 0.1 pound after each load was weighed. Chemicals at low application rates (i.e. O.A% propionic acid and 0.5% ammonium isobutyrate) were diluted 1:1 with water. Chemical solutions were sprayed on the 70 .QOHpSHom peprpGoocoo &wm m CH CmeQoHE .mchfiwH .hanEoo madameum aoHaawo Seam pomwnoadm ma: ophnowaanme .eommeCoe .mHnmSez .hameoo.vqm cowaw .m.3 hp GOHHSHOm COHumapaeocoo Rae a CH poflchhdm .opwa%pdnomH ESHGoEE¢ "mH¢U .mwxoe .HpmHano mammoo .hqwmsoo HwOHEoflo ommGwHoo hp UQOHGASm .omuom we OHpma w QH pHow oHpoow was UHom oHGonoam mo oHSpNHS "moumoEonoo .AUeHHmCQ may 3\3 mo mHmmn so was Ho>eH psoEpwoaen .oNHm CH E NH N m .mOHHm opoaocoo uanambw ssH.om 0mm.mm em ones oasoaaoam 3.0 mm.Hm\r m ~:H.mm omm.mm me mH< em.o mm.Hm\c a smm.cm oos.o: :m osoz om-mH\b c mmm.am mnm.mm cm mas as om-ma\c m ooossooHdsaom mom.em omm.mm mm wmm.H+omH< Rm.o mm.mm\c : Hes.om ooo.om mo ooaos oesoaaota sm.o mH\c m antenna same“. he. sham. h. Hapoe Hence a . .MNOH .emmthm mm UeHHmcm wwmeH¢ we unopcoo soups: ham one pcospmeae .mopwa mchmo>amm .HH oHQmB 71 forage as it entered the blower from the wagon by a precalibrated pumpA. Formaldehyde and ammonium isobu- tyrate were mixed together after each had been weighed individually. Forage in the silo was leveled and tramped after every two leads. Thermocouples (copper-constantan) were placed near the center of the silo at approximately 1.5 m vertical intervals after leveling and tramping so that by the end of the ensiling operation, five thermocouples were placed in each silo with two in the bottom potion. Leads from each thermocouple were extended outside the silos and attached to a 2A channel potentiometerB which automatically converted electrical potential in to temperature and recorded the temperature at preset time intervals. During the entire A2-day of haylage storage period temperature was recorded ten times per day for the 2A leads. Once daily recordings were obtained for the fifth or other bottom thermocouple lead for each silo and ambient using a portable potentiometerc. When silos were opened for feeding, temperature measurements by recording potentiometer ceased but portable potentio- meter readings continued for the bottom thermocouple ASupplied from the W. R. Grace and Company, Tennessee. BSupplied from the Department of Agricultural Engineering, Michigan State University. CBrown portable potentiometer Model 126 W2. Minneapolis- Honeywell Regulatory Co., Philadelphia, Pa. 72 leads for three additional weeks. During the entire period when the six silos were being emptied, tempera- ture 25 cm beneath the surface of the haylage was obtained about two times per week with a mercury thermo- meter. All material removed from the silos as weighed and recorded as silage or spoilage. Incomplete weights were obtained on material remaining in the silo at the end of the feeding trial. II. Feeding Trial Forty—eight lactating dairy cows, averaging more than 18 kg of milk per day, were assigned to one of six haylage treatment groups (Table 11) in a randomized block design. Milk yields during a 2 week standarization period were used as a blocking factor. Treatment groups were also balanced for age, genetic groups, and days after parturition. During the standardization period cows were fed haylage ad libitum (allowing 10% orts) from a general herd supply. In addition they received a grain mixture (16% crude protein) at a rate of l kg/3 kg milk. Refusals were weighed 5 times per week. During the experimental period (A9 days) the assigned haylage was fed ad libitum. Grain mixture (16% crude protein) was 73 fed to cows at rate of 1 kg per 3 kg milk. Refusals of haylages and grain were weighed five times per week. Cows were weighed on two consecutive days at four days after initiation of the trial and at the end of the trial. Milk weights were recorded five days per week during standardization and experimental periods. III. Milk Analysis Composite (AM and PM) samples of milk were taken from each cow at biweekly intervals during the feeding trial and twice during the standardization period. To- tal solids were determined by drying 2 ml for 2 hours in a forced air oven at 100 C. Butterfat was determined by a butterfat auto analyzerA. IV. Feed Analysis Haylages were sampled from the feeding cart three times per week during the trial and stored in a refri- gerator at A C. Samples were composited biweekly and used for laboratory analyses. For various kinds of analyses different aliquots were taken from the composite (Figure A). An small aliquot was used for dry matter AMark II. A/SN. Foss Electronic Company, Hillerod, Denmark. 7h .methwG< mo mwGHm mSOHaw> sow monEdm emehwm UepHmomEoo mo mGOHquonam .2 oastm GomoapHG H z GomOHpHQ w oHQSHOmGH pcomaopob ©H6< A, AmQ¢VAonHw pnomaopep mucopdepmcoo powApNe GHopoam pomppxo aoQHy QHGwHH CH0< HHmz HHoo ooAMImz 06390 aoflpm 06590 £m¢ a a a $11 . a _ % a a mHmhwmcw wwwmwwfld mpGoSpHpmcoo COprOHmeGoUH op wonm mdownHm mUHow use mpafioo OHnwwao UHoS new ewes canoes oHaraHo> ma wearasHm s i i n AwGHhAU Conroy A0 0.: vv — a GOprQHELepob wcHhaU someowapxo popes QpHS heppws ham aHm booaom :om m z 0.0 QOHpmNHQeonom H n i a _ monEwm omehwm wouHmomEoo 75 determination by oven drying at 100 C for 2A hours. About one kg of sample was spred evenly on aluminium pans and air dried at room temperature. A fan was usually used to speed the drying process. The dried samples were then ground through a 1 mm screen in a Willey mill and used for proximate analyses (AOAC, 1965), fibrous constituent and acid detergent insoluble nitrogen analyses (Goering and Van Soest, 1970). For organic acid determinations a 10 g haylage sample was mixed with A0 ml of 0.6 N H2SOA and stored for 2 or more days at A C. The mixture was then filtered through two layers of cheesecloth and the filtrate was further cen- trifugated at 12,000 x g for 20 min. The clear superna- tant was used for determination of lactic acid by the procedures of Baker and Summerson (l96A). Volatile fatty acids (acetic, propionic, butyric, isobutyric, valeric and isovaleric acidfi'were quantified in a gas chromatograph as follows. Three,u1 samples were injected into a Hewlett—Packard, F and M gas chromatographA using a glass column (2.A m) packed with chromosorb 101 (80/100 mesh). The injection port temperature was set at 3A0 C, the column temperature at 285 C, and the flame detector at 320 C. Nitrogen was used as the carrier gas and AHewlett-Packard, F and M Scientific 00., Model A02. 76 flow rate was 30—A0 ml per min. Sample volatile fatty acid concentrations were calculated by comparing peak height with those of analytical grade acids made into a dilute "standard" solution. Haylage samples were prepared for pH measurement and number and type of fungi by homogenizing 20 g of haylage and 180 ml of distilled water into a Sorvall Omni—MixerA for 3 minutes at 8,000 r.p.m. with the home- genizing cup immersed in ice. The pH of the homogenized material was measured with a Sargent pH meterB using a combination electrode. The homogenate was then filtered through one layer of cheesecloth and the filtrate was used for the pour-plate technique for estimating the number of fungi present. Three to five serial dilutions of l to l x lO-u up to l to l x 10-6 of filtrate with sterile water were normally required in order to have proper concentrations of fungal spores and mycelia in the plate (approximately 20-50 colonies per plate). The growth medium used was potato dextrose agarC. One hundred mg of Novobiocin calciumD was used per liter AIvan Sorvall, Inc., Newton, Conn. BE.H. Sargent and Company, Chicago, Ill. CBBL, Cockeysville, Maryland. DThe Upjohn Company, Kalamazoo, Michigan. 77 of agar. Poured plates were incubated at room tempera- ture for approximately 9 days or until sufficient fungal growth was noted. The colonies was counted using a colony counterA and identified according to genus. V. Sheep Digestion and Nitrogen Metabolism Trials Twelve sheep averaging about 20 kg body weight were fed general supply (herd) haylage for a week before silos were Open then after silos were open, two sheep were assigned to one of the six experimental haylages. Weighed amounts of these haylages were fed to sheep and refusals were also weighed daily for a 7-day preliminary period. Dry matter content of each haylage was estimated from a composite sample taken during the preliminary period and maximum dry matter intake for each sheep was thus obtained. On day 8 the amount (90% of the maximum) to be fed during the next 10 days as removed from the silo and an.amount weighed out into plastic bags to be fed daily to daily each sheep and stored frozen or refrigerated. Also, haylage samples were taken during this weighing operation and handled in a manner similar to that described in the section of Feed Analyses. Sheep were transferred from calf pens into digestion crates on day 8 but no collection of feces and urine were made AFisher Scientific Co., New York, N.Y. 78 during the next three days. During the last 7 days, total feces were collected daily and temporarily stored at A C. Urine was collected in a plastic bottle containing 20 ml of 50% H2804 (v/v), and the daily volum measured and recorded. Ten percent (by volume) of the daily urine volume was saved in a plastic bottle and stored at A C. This composite urine sample was thoroughly mixed and a small portion (about 20 ml) used for total nitrogen analysis. At the end of trial, feces collected from 7 days were weighed, thoroughly mixed and a small portion was used for dry matter determination (oven drying). Also, a aliquot of about 500 g were air dried ((lAO C). These air dried feces samples were then ground through a 1 mm screen in a Willey mill and used for analyses of chemical and fibrous constituents. Sheep were weighed on day 7 and day 11. The entire l7-day feeding trial was repeated two times with random reallocation of sheep to haylages. A 5-day adjustment period was allowed between trials. VI. Statistical Analysis All data obtained from the milk production trial were analyzed on a Central Data Corporation 6500 computer at the Michigan State University computer center. Analysis of variance and analysis of covariance programs 79 were used and where statistical significances were noted, differences between treatment means were tested by Duncan's multiple range tests (Steel and Torrie, 1960). Two-way analysis of variance was used to analyze the data from sheep digestion trials. When statistical significances were noted differences between means were tested by Duncan's multiple range tests (Steel and Torrie, 1960). RESULTS AND DISCUSSIONS Part 1. Forage Nutritive Value Evaluation by Several Laboratory Methods 1. Relationships Amonngn Vivo Responses Simple correlation coefficients among five in 1319 parameters are in Table 12. Nitrogen digestion (ND) co- efficients were correlated with dry matter digestion (DMD) coefficients (r = .86) and nitrogen (N) balance (r = .6) but were not well correlated with N retention calculated as a percent of absorbed N and maximum dry matter (DM) intake (r = .12 and .33) respectively. N balance was Significantly correlated with DMD (r = .83) and also co- rrelated with maximum DM intake (r = .75) and N retention as a percent of absorbed N (r = .72). A significant but relatively low correlation coefficient (r = .65) was noted between DMD (ranging from A3 to 70%) and maximum DM intake (ranging from 1.91 to A.6A kg/lOO kg body weight). The close relationship between DMD and ND can be expected since in normal forages, proteins are primarily associated with materials of cell contents which are also highly digestible dry matter (Van Soest, 1969). The high correlation between DMD and nitrogen balance supported 80 81 mod Va u c So.ova n o ”mood ya u o “Hooéva u a “moodva .mooflm adom so woman on 00% Comp meg onh00 Some .momwaom :m .GmwOAHHG II III II G venaomnw ARV R mm mpCoHOHMMooo vochpoa GOHumoMHU GomOAsz moamenIz QSQ. Cemoasz ass.o ems.o nmc.o mm.o ems.o om:.o NH.o wmw.0 o00.0 some N pflmHoz hvon o emancH sounds haw ESEHNME A 0 ARV z cenaomnm mo & mm UeQHmpoa z Ahw0\z mv ooCdenImz o QEQ muQeHOHameoo QOHpmepr ampere ham A 0 UoGHmpno whopoewawm o>H> .mommthm 00m meonm Seam mM.e>Hm mcoE< A90 paeHonmeoo GOHpmHomaoo oHQEHm .NH oHQwe 82 the concept of a Significant interaction between rumen nitrogen and carbohydrates (Waldo, 1968). There are several factors which could limit the degree of corre- lation between ND and N balance. These factors are: protein solubility, greater experimental error involved in N balance determinations, physiological state of animal and feeding level (Chalmers, 1961). The insignificant correlation between ND and N retention as a percent of absorbed N might be anticipated since these two determinations represented two different stages of overall N utilization by the animal, one in the gastrointestinal tract and the other within the body proper. The significant positive correlation between DMD and maximum dry matter intake observed in this study statistically confirmed the concept that dry matter di- gestibility is one of the important factors which con- trol the voluntary intake of roughage by ruminants (Van Soest, 1969), but contrary to the concept that diges- tibility decreases as intake of a given roughage in— creases (Campling, 1970). II. True Digestion Coefficients of Forage Total Nitrogen and Other Nitrogen Fractions Estimations by Statistical Means When digestible nitrogen contents (g total 83 nitrogen/100 g forage DM x apparent nitrogen diges- tibility) were regressed on amount of total nitrogen or other nitrogen fractions (g N/100 g forage DM), the resulting regression coefficients b.I (Slope) can be interpreted biologically as an estimate of true diges- tibility of the nitrogen or nitrogen fraction while bO (intercept) can be interpreted as the metabolic fecal nitrogen or grams of nitrogen excreted per 100 grams of feed containing no nitrogen. Generally, studies of the regressions of digestible nitrogen content of feed 2 ( i.e. on total nitrogen content have resulted in R proportion of the sum of squares attributable to re— gression) as high as 0.99, a bO value of about - 0.5 to - 0.6, and b1 values greater than 0.9 (Holter and Reid, 1959). However, Goering gt El: (1972) have re- ported very low values of b1 (0.79) and R2 (0.71) when regression analyses were performed on heat damaged forages. Using the same model, true digestion coeffi- cients of total nitrogen and metabolic fecal nitrogen for the samples in the present study are Table 13. The value of b (—0.83) and b1 (0.91) would suggest O that these forage samples were probably normal but the 2 (0.35) indicated, on the other hand, extremely low R that the forages were somewhat abnormal. Other regressions were calculated (Table 13) by 81+ .opasHumo no aoaao cascaduu "mam u .GOHnnopwoa on oHnuuanaupa nohdadn no Eda on» uo GOthonoan ummo .choonocdn u.:dgn .ucoHoHuuooo cOHpuHoaaOO HuHuaaa "0.0.5”Ho .HoHnoaoHa scans ncoade .AucoHOHhuooo GoHuuomHu oshuv oQOHn ano .oHndHoaGH u.HOmcHn .Aowdoau haw w 00H\z m comoaqu Hdoou OHHondpoEv uaooaoucH uonn .oHndHom u.Homw .owaaom haw m 00H\w no woman noaH¢> HH.I om.HI 00. H0. Hoo.I z mHomcH pcowaououIcHo< 2 .Hon anomaopoquHo¢ m am. on. mH.H mm.H- z macs scamsouoeIoaoa m mm. mm. Ho. mm.n szcomOAuHc Hence H . . . m . H o m m 0 0 m D .0.0 m n n Hmm cm 0 o 9 NR HK GOdezvm mpGoHonmooo :onmomwom .mowdaom thmaobHcD ououm deHnon Hm you mcoHpooam cowoaqu owdaom no AMV cowoaqu oHnHumowHQ mo mGOHmmoamem oHQHquz 0nd OHQBHm .MH 0Hnda 85 separating soluble and insoluble nitrogen fractions in an attempt to identify these fractions responsible for the relatively poor fit of the simple regression. Equa- tions 2,A,6,8 and 10 imply that digestible nitrogen (Y) originates entirely within the acid detergent (AD) so- luble N, pepsin soluble N, rumen microbial + pepsin soluble N, rumen microbial + pepsin + pancreatin soluble N, pepsin + pancreatin soluble N or hot water insoluble N fractions. Equations 3,5,7,9 and 11 imply that diges- tible nitrogen originates in the soluble fraction des- cribed above as well as in the insoluble fraction. All equations that included a soluble nitrogen fraction gave b1 values either above 1.0 or extremely close to 1.0. These high b1 values (true digestion co- efficients) were somewhat unrealistic based on biological basis Since the maximum value of digestibility is 1 or 100% but they certainly indicated that these nitrogen fractions were not only soluble but also highly diges— tible. In this study the estimated true N digestibilities (b1) of AD soluble N and pepsin soluble N fractions were all greater than 100%. This is contrary to the report of Goering gt 21. (1972). They did not find increases in b1 values for AD soluble N and pepsin so- luble N fractions greater than 1 when they changed from total N analyses to the AD and pepsin soluble N frac- tions. 86 With the exceptions of hot water and pepsin + pancreatin incubation values marked improvements of R2 were noted when soluble N fractions were used in re— gressions as compared with the R2 when total nitrogen was used (Equation 2,A,6, or 8 vs. 1). The highest R2 value (0.89) was found for Equation A which is based on the pepsin soluble N fractions. All of the b2 values (true digestion coefficient estimates of insoluble N fractions) were negative ex- cept for the values for the pepsin + pancreatin and hot water solubility methods. The negative true diges- tion coefficient estimates suggested that the insoluble fractions separated by chemical or in 11:33 enzyme di- gestion methods were also indigestible in 1132. The positive coefficients for hot water and pepsin + pan- creatin insoluble N fractions, on the other hand, in- dicated that these fractions were digestible in 1319. For the hot water insoluble N fraction, one should ex— pect the positive estimates of in 1112 digestion co- efficient since the hot water insoluble N fraction consists of true protein N. The positive coefficient for pepsin + pancreatin insoluble N fraction was not anticipated, since based on theoretical consideration the action of these two enzymes should be similar to that for in vivo digestion. However, if the in vitro 87 incubation conditions were not optimal for enzyme di- gestion results could be unrelated to in XEXQ diges- tion. The degree of indigestibility of insoluble N fractions varied extensively (Table 13). For example, the true digestibility of acid detergent insoluble N fraction was estimated to be -126% (Equation 3) while the value for pepsin insoluble N fraction was only —2% (Equation 5). The greater value of indigestibility from the AD insoluble N fractions suggested that this method is more sensitive t°.lB.ZlX2 N digestibility than pepsin insoluble N fraction. A slight increase in insoluble N fraction would result in a large reduction of digestible protein content. The R2 value was markedly improved by utilizing the insoluble fractions as predictors in addition to the soluble portion for acid detergent solubility method (R2 = 0.76 for Equation 2 vs. 0.91 for Equation 3) but improvement was not large for pepsin, pepsin + pancrea- tin, rumen microbial + pepsin + pancreatin, rumen mi- crobial + pepsin and hot water solubility methods (Equation 5,7,9 and 11 vs. A,6,8 and 10 respectively). The extent of additional improvements in R2 due to in- clusion of the insoluble N fraction are related to the partial correlation coefficients of these insoluble N 88 fractions. Among all insoluble N fractions only the acid detergent insoluble N and hot water insoluble N fractions possessed relatively high partial correlation coefficient values (—0.78 and 0.68 respectively). Separation of total N into soluble and insoluble N fractions increased the R2 value and reduced standard error of estimate (SEE) when digestible nitrogen content was predicted using both fractions. However, the value for true digestion coefficients (b0) of these fractions are difficult or impossible to interpret. Similar state- ments were made by Goering gt El: (1972). Values of R2 and SEE in Table 13 suggested that digestible nitrogen content would be estimated most precisely by Equation 3 based on acid detergent soluble and insoluble N fractions. From a practical predictive standpoint, Equation A based on pepsin soluble N alone would be prefered, but from an operational standpoint acid detergent solubility is a less troublesome deter- mination than pepsin and other protein solubility methods. III. Relationships of Various Nitrogen Fractions to Nutritive Value (1) Total Nitrogen Digestion Coefficient a. Nitrogen digestion coefficient and acid detergent nitrogen: Two regression equations predicting nitrogen 89 digestibility from acid detergent nitrogen as percent of total N (AD insoluble N x 100/N) were calculated from two different sources of data. Equation 1 was computed based on the data of the present study (Table 1A) while Equations 2 and 3 were derived from the combined data of the present study plus that of Goering (Maryland), Jorgensen (Wisconsin) and Pierson (Minnesota). An ex- tremely high correlation coefficient (r) or adjusted r (adjusted for sample number) was found in all three cases. Similar regression coefficients (b0 and b1) were found for Equations 2 and 3 but both were different from the coefficients of Equation 1. Equation 1 had a greater value for bO (N digestion coefficient when the sample contains no AD insoluble N) than did Equation 2 and 3 (91 vs. 73 and 76). The slope (b1) was much steeper for Equation 1 than for Equation 2 or 3 (2.2 vs. 1.1). Using equation number 1, developed from the present study, the nitrogen digestion coefficient would be 0 when the value for acid detergent insoluble N as percent of total N reached a value of A2, while the nitrogen digestion coeffieient should not be 0 until this percent reached about 72 with equations 2 and 3. The discrepancies observed in regression analyses were probably due to the differences in the extent of heat damage (amount of AD insoluble N/N) among samples 90 . H A n0 oaonw .Aonv uncommpGH .aomOApHc Heron we pnooaem a me oHnsHomGH pGCwACHCU .AGOonHm .0.Qv anemonGHS .AcomocwnOh .<.zv GHmcoomHz e cacao 0 .NSoH .mmNHnmm .Hom hhme .0 .poanHnsm neuwH 0am .mm.mm.wGHa000 Song wonprpo moSHw>o. .Sedmaoeas: charm damages: 3 .AoomH .oHasoe 0cm Hompmv sonesq onEmm pom popm5m0¢ .pnoHOHmmooo GOprHepaoo woundn0 n w m0.I m0.I ow Umaofluo + H> SH mow mHthwQ< GOHmmoawom aonHH .dH oHan < 91 of different sources. Many research investigators supplied animal data and samples for analysis by Goering whereas the MSU data were from a more uniform experimen- tal situation. In the present study, the range of AD- insoluble N/N was from 0.71 to 22.0% while the range in the data of Goering gt g1. was from 1.78 to 75%. Con- sidering the regression equations and the amount of AD- insoluble N/N in the original samples, one is forced to conclude that Equation 2 or 3 should be used to estimate in 1113 N digestion coefficients of forages which have been severly heat damaged (i.e. AD-insoluble N/N greater than 22%). These equations also indicate that the ne- gative effect of AD- insoluble N on in 1112 N digestion coefficient is variable and dependent on the absolute level of AD-insoluble N/N in the samples. Acid deter- gent insoluble N as a percent of total nitrogen reduced ‘in 111g N digestibility to a greater extent in forages with only mild heat damage (b1 = 2.19) than it did in forages severly heat damaged (b1 = 1.08) where AD- insoluble N/N exceeded 22%. Since Van Soest (1965) has indicated that normal (not damaged by heat) forages contain about 7% of AD- insoluble N/N, the combined samples were separated into "normal forage" (AD-insoluble N/Ns9) and "abnormal forage" (AD-insoluble N/N>9). Regression equations 92 relating in 1113 ND to AD-insoluble N/N were computed for these two sub samples and are in Table 1A as equa- tion A and 5. The precision for predicting of l&.XiXQ ND from AD-insoluble N/N was appreciably lower for normal forages than for abnormal forages (adjusted r = -0.A8 vs. -O.91). The bO also indicates that normal forages have greater ND than do abnormal forages (81 vs. 72%) when both of them contain no AD-insoluble N/N. Markedly different slopes (b1) were also evident be- tween the two equations with the normal forages having a greater slope than the abnormal forages (1.A7 vs. 0.96). Statistical analysis revealed that the slopes of these eqations were significantly (P< 0.05) different. Results from this study clearly demonstrated that the value of AD-insoluble N/N in predicting forage in 1132 ND will largely depend on the extent of heat damage that the forage has undergone during storage or treat- ment. b. Nitrogen digestion coefficient and other nitrogen fractions: Regressions of total nitrogen digestion coeffi- cients on various nitrogen containing fractions excluding acid detergent insoluble N/N are in Table 15. The equa- tion using acid detergent (AD) insoluble N as a percent of dry matter as the predictor resulted in the greatest r (-0.90) with the least standard error of the estimate 93 0:.o om:.I mo.mI wm.om mm sea .2 .HomaH .csd + sHmaom :m m:.o oo:.I mm.mI Hm.ms Hm 2\ooH a zImmz .omdoHoa mmz steam mm mH.o on. H:.mH so.om :m sea .aowoarHa .HomaH sons: pom mm so.w e:m.- ms.I mm.sm :m 2\OOH a z .HomaH .asa + aHmaom Hm H®.® 00m. 0N.0H 50.0H dm SMR .GowoapHc .Hom .Gwm + :Hmmom 0w :s.m cam. co.mH m:.oH rm :0 .aowoaoHa .Hom raomaoroeIoHos oH mm.m nmo. mm.H 0m.mHI :m R .Heppws haw .Hom chmom + Goadm 0H oc.s dos. no.Hm do.OH rm 2mm .aowoapHa .Hom aHmaoa + dossm SH 0m.w wow. Hm.H :m.h :m .Hoppwe haw .Hom .cwm + GHmmom 0H mm.> mm».I 0®.MHHI 0H.m® :m ASCOJJ pm Gov wGszoan ho common mH mm.s sma.- oo.- 4:.mHH :m sea .Aozov masoerpmaoo HHsz HHoo :H oo.n are. mH.mm nc.HH mm sea .2 .Hom .asa + aHmaoe + steam mH ®~.0 mow. m>.0m mm.mH mm SQ& .GowoapHc .Hom GHonm NH mr.r are. mH.H om.o mm a.zm .Hom .ada + sHmaoa + sossm HH om.c cos.I mm.mm- oo.mo mm zoa.z .HomaH .add+ aHmaoa + cossm OH sm.c aoe.I mo.cm- mm.Ho :m sea .2 .HomsH sHeaca + dosam o Hw.m w:®.I 0N.HI 00.:0 :m 2\00H K 2 .HomcH QHonQ + Goadm w no.m new. om.H mm.c mm a .aoepds has .Hom sHmaoa A as.m :m.- om.HI om.mHH :m sea .AamsvaosHe rswwsorooIoHos r Md.m w©®.I ON.HI :m.00 NN Z\OOHNZ .HOmGH Wfiwg + CHmQOQ +HG®SSm m mm.m scm.I mo.om- H:.cm mm sea .2 .HomsH aHmaoa : HS.: som.- so.HI mm.mm mm 2\00H a z .HomsH sHmaoa m oc.s soo.- mo.:- ms.ooH :m sea .sHawHH haemaorooIoHos m :m.: doo.I om.so- :s.oo :m mama wz wHomaH aaowacroeIeHos H m mam a HD on c x COprsdm a o m s . mucoHonmooo QOHmmonom IMP ! .mede> hLOpwpoan m5OHHw> wGHmD AMV mpquOHmmCoo COHpmomHQ Somoasz Hmuoe maprEHpmm How mGOHmmonem HonHH .mH OHDwB 9L1 “mod vs u o “Hod ya u o ”mood vs u o Hood Va u n Mmoood v .GHcmHH pcowaepeo 0Ho¢ moo an e a w m .Heppwe haw n Sam .QomompHc n zm .oHnsHomaH u .Homch .mpwgflpmm .HO ..HOchHQ ULImUHprm H mam H a .oHQdHom H .HomM .ucoHOHmmooo SOprHoHHoo H H0 .GproHoqwm n .Qwah .CQOHm NHQm .HmHQOLOHE amass u coESmH .pmeoaopQH ”09¢ mm.OH HmH.I Hm.I Hm.ms :m ame\OOH a gas om :3.0H eoH. oH.S os.o: rm somoaHHa awroa om HH.OH HHm.I mm.mI mm.HS :m 2mm .2 .Hom tops: pom mm am.o eHm.I om.I om.cs mm z\OOH a 2 .Hon sneeze Hdsost em mm.o com. sm.o Ho.s: :m 2\OOH a z .HomaH horas pom cm cs.a Mos. mo.o Hm.mm :m a .aoesms sac .Hom were: pom mm H o mmm a c c a x soHsaaem mpGeHOwaeoo GOHmmoawom AoosaHraoov .mH oHnme 9S (A.5, Equation 1, Table 15) excluding AD-inSoluble N/N from the comparison (Equation 1, Table 1A). The ex- tremely small difference in r and SEE values between regressions usingeither AD-insoluble N/N or AD-insoluble N/DM was not observed by Goering gt gl.(1972). They found the former much more precise in predicting nitro- gen digestibility (ND) than the latter. Another high correlation was observed by using acid detergent lignin as the predictor (r = -0.90 Equation 2, Table 15). Again, this finding is contradictory to that of Goering gt El. (1972) who observed a relatively low r value of -O.7A and concluded that AD lignin had no predictive value for nitrogen digestibility. According to Van Soest (1965), the acid detergent insoluble N fraction is primarily associated with AD lignin in heat damaged forages. Thus a significant negative relationship between AD lignin and ND should be expected when forages have been damaged by heat. Other important nitrogen digestion coefficient predictors were pepsin insoluble nitrogen as a percent of total N (r = -.89, Equation 3), pepsin insoluble N as a percent of dry matter (r = -.86, Equation A), rumen microbial + pepsin + pancreatin insoluble N as a percent of total N (r = -.86, Equation 5), acid detergent fiber (r = -.8A, Equation 6), pepsin soluble dry matter 96 (r = .8A, Equation 7), and rumen microbial + pepsin insoluble N as a percent of total N (r = -.8A, Equation 8). Moderate degrees of precision (r from -.79 to .51) in predicting N digestion coefficients were measurements obtained from rumen microbial + pepsin + pancreatin solubility, rumen microbial + pepsin solu- bility, pepsin + pancreatin solubility values, cell walls and degree of browning (Equation 9 to 22, Table 15). Hot water solubility, mineral buffer solubility values, rumen ammonia release and total nitrogen were correlated with N digestibility so poorly that no practical pre- dictive value for these variables were obtained. Saunders gt El: (1973) obtained a good correla- tion (r = -.87) between alfalfa protein solubility in pepsin + pancreatin solutions and rat in 1312 nitrogen digestion coefficients. Following the same procedures, a correlation coefficient of -.5A (Equation 21, Table 15) was found in this study. In this procedure pepsin was incubated with the sample for only three hours which might be insufficient for pepsin to digest any significant amount of forage protein. Further investi- gations and modifications using this procedure may be desirable. Protein solubility in mineral buffer solutions 97 was highly advocated by Wohlt 32 21. (1973) as a stan- dard procedure for evaluating protein quality although they did not present any evidence that this solubility test would be significantly correlated with in 1113 nitrogen digestion coefficients. Results from this study do not give any credence to this hypothesis. The simplest laboratory procedure used to pre- diet in 3132 N digestion coefficients was a measure of the degree of browning (an extraCtion-spectrophotometric procedure), but the precision of prediction by this variable was only moderately high (r = —.73). Apparently the degree of heat damage to forages can only be approxi- mated by color development. Hot water solubility measurements have been used as a measure of silage fermentation (Waldo, 1973) with increasing amounts of hot water soluble N indicating increased silage fermentation. However, hot water pro- bably coagulates all protein regardless of its biological availability. In other words, measuring protein solubi- lity in hot water is not a sensitive method to detect protein damage by heat. Results from the present ex- periment strongly supported this assumption (see Equa- tions 22,25,26 and 28 in Table 15). The most complicated procedures used in this study were rumen microbial + pepsin and rumen microbial + pepsin + pancreatin incubations. The predictiability 98 of nitrogen digestion coefficient by these measurements were high but not as great as those obtained from much simpler procedures (e.g. pepsin solubility or acid de- tergent solubility) (Equations 1,2,3,5,8,9,10 and 11 on Table 15). Nevertheless, one must be reminded that those in 12239 rumen fermentations were designed mainly for predicting in 1132 DMD not ND. Nitrogen insolubility was believed to be positively related with degree of heat damage in forages (Goering ‘23.21. 1972) and because of this relationship the term N insolubility was used as the predictor for 22:2112 N digestibility in Equation 1, in Table 1A and Equations 3,5,8 and 21 in Table 15. However, if the regressions were calculated by using N solubility (100 - insolubility) as predictors, the b0 (i.e. estimated in gizg N digesti- bility or solubility when forage protein solubility is 0) would have biological meaning. Ideally, if an in 31322 protein digestion method had the same action on forage protein solubilization as the in yizg_protein digestion process then the regression coefficient b0 and b1 would be 0 and 1 respectively. Conversely, if an in 12232 digestion method had greater or less ability to solubilize forage protein than did the in zizg_diges- tion process, the b0 value would be a negative or a positive value respectively, and the greater the absolute 99 value, the greater the difference between in 31232 and ‘in 1119 N digestion processes. The bO values for the equations using N solubility as the predictor can be easily derived from the equations based on N insolubility and represent estimates of in 3113 N digestion coefficients when total N was 100% insoluble. These new bO values are in Table 16. Table 16. Regression Coefficients (b ) From the Re- gression Equations Using N Solubility Determined by Several Solubility Methods at the Predictor. Solubility Method bO Acid detergent solutions -l27.77 Rumen microbial + pepsin solutions —32 Rumen microbial + pepsin + pancreatin solutions -20.66 Pepsin solutions -l8.77 Pepsin + pancreatin solutions 11.53 The bO values in Table 16 suggest that the extent of N solubilization by pepsin + pancreatin solutions was near but slightly smaller than l&.XlXQ N digestion, while the other four solutions, particularly acid detergent solution were able to solubilize forage N to a much greater extent than in XEXE digestion. The reason for the large decrease in digestibility of AD insoluble N fraction could be due to the increase in fecal bacterial 100 nitrogen when forage soluble N fraction content decreases. (2) Regressions for Nitrogen Balance Regressions for nitrOgen balance on various nitro- gen containing fractions are in Table 17. Generally, nitrogen balance was not as predictable as the nitrogen digestion coefficient by these forage N fractions. Also, based on r and standard error of estimate values, the ranking for precision in predicting N balance and N digestion coefficients by these N fractions were appre- ciably different. Nitrogen balance was predicted more precisely by soluble N fractions than insoluble N fractions while the reverse situation existed for predicting N digesti- bility. For example, the two best predictors for N balance were pepsin soluble N as a percent of DM (r = .85, Equation 1) and rumen microbial + pepsin + pancrea- tin soluble N as a percent of DM ( r = .80, Equation 2). While the respective r values for the insoluble N frac— tions of these two solubility methods were -.67 and -.57 (Equations 13, and 18, in Table 17, respectively). The acid detergent insoluble N as % of total N was the best predictor for digestibility ( r = -.92, Table 1A) but N could predict N balance only with a moderate degree of precision ( r -.60). These results indicate that the proportion of N that is soluble is directly related to 101 mc.m on:.I om.omI mo.s rm Asses: as gov maHsSotp Ho ooawom mm rm.m ch. sm.o ma.o mm a .29 .Hom .saa + aHmaoa Hm s:.m orm.I HH.HHI SH.0 :m an m .z .HomsH aaomaopooIoHos om 0:.N 00m.I NH.0I 00.0 :N 20% .z .HOmGH aHonQ + cmsdm 0H :m.N 00m.I 00.0I 00.0 MN 20& .z .HomGH .ndm + chmom + meadm 0H 0m.m oHc.I o:.I 0:.0 em z\ooH a z .HomcH nsowaopooIoHos SH 0m.m are. co.m mm.mHI :m an a .z .Hom HomeschooIoHos cH m:.m rmc.I rm.HI H0.0 Hm A2\OOH a zImmzv onaoHoa mmz acesm mH mm.m asc.I mo.HI mo.mH :m sea .AHQHV stmHH rsomaoecoIoHos :H cm.m mac.- 00.0- :m.m mm sea .2 .HomaH aHmaom mH 0H.N w00.I 0N.I 0N.0H :N Z\00H N z .HOmcH chmom + CoESm NH 0m.m 000. 0m.o Ho.mHI mm a .20 .Hom .cse + chooc + cream HH NH.m mos.I cm.I nm.cH am so a .mrsoerrmaoo HHaz HHoo OH m0.N wNN. 02.0 0:.NHI :N & .aoppwe 0&0 .HOm seems pom 0 mH.m amn.- om.I oo.mH rm zxooHUH z .HomaH .csa + aHmaoa + dossm 0 Ho.m are. m:.o m0.mm- :m a .2m .Hom sHmaoa + aossm s 00.H emu. H0.0 m0.NHI :N 20% .z .Hom GHonQ + coEdm 0 oo.H ars.I sm.- 2:.o mm 2\OOH a z wHomsH aHmaom m Ho.H aSS.I Hm.I mo.cH :m sea .AaesvtonHe asomaorooIeHos : oo.H are. mm.o :H.mHI mm a .acaras awe .Hom chaoa m :0.H sow. H:.s eo.mHI rm sea .2 .Hom wean + «Hence +Hsossm N am H mama mm 0 am mH mm mama mz mHon aHmaoa H mmm H HQ on a x COprdvm a o m a . mpGoHOHmmmoo SOHmmeHmem .mous> haouwaopwH mSOHHw> Go ANV mme eonme Gemoasz How mGOHmmonem awocHH .NH mHan 102 ”moo to n o “Hoo o n o ”mooo yo n o ”Hooo vo n a $893 .opwEHpmo mo Hesse womwcdpm u mmw .mo.o.ao u o and s .Heppwe haw u 290 .Gomoo HG n p. zpH .e S cm H . ow HQ H H m Q .cHQSHomGH u .HOmQHM .pGeHeHmmooo COHpmHoaaoo n so .QproaQGmm H .mew .eQOHm ”Ham .HmHQOHOHE cosfia u oesdm .pmoooopQH Hons H so.m oom. om.H ms.H so so a .s .Hom aces: nos mm oo.m can. mo.: m:.HHI do so a .somoaaHs Haooo rm sc.m one. m:.s mc.oI mm so a .z .Hom .saa + canoes mm H o mmm a Q Q G x GOprsvm mpooHoH00000 EOHmmeowom AwesoHpcoov .NH oHan 103 the quantity of N retained in the animal body but that ration of insoluble N to total nitr0gen is not. Simple regression equations having an r value below .70 have no practical predictive value (Goering gt 21. 1972). Thus acid detergent lignin, rumen ammonia release as a percent of total nitrogen, degree of browning, hot water solubility measurements and total N etc. should not be used as predictors for estimating N balance (Table 17). (3) Regressions for Nitrogen as A Percent of Absorbed Nitrogen Some regressions for N retention as a percent of absorbed N on various laboratory measurements are in Table 18. None was sufficiently correlated to be useful for predictive purposes. The greatest relationship was between hot water soluble DM and N utilization ( r = .62). Others will not be discussed or presented. (A) Regressions for Estimating Dry Matter Digestion Coefficients Regressions for dry matter digestion coefficients on 31 laboratory measurements are in Table 19. Several laboratory determinations were able to predict DMD with a great degree of precision. Important predictors were largely measures dry matter solubilities in various solutions and only a few were measurements of nitrogen solubilities. For example, the two highest r values 10A .eHQSHOmQH H .HOmGH h .eHQSHom u .HOmm .chmeHoqu n .me H .opwEHpme Mo Hesse UHwUQmpm u mmmm .HINHQOLHOHE GQEHHIH H HIHGEHHm m .uGoHonmeoo COprHoaaoo H to mo.o.Ao u o H ”m0.0 v0 N 0 ”H0.0 vo n 0 “H00.0Avo u 90 .CQOHm H pm 0 .ooppms 0H0 n Sam .ugooaoucH n as om.o osm. oo.HH :m.mmI am so a .somoaaHs Hence mH s:.o oom.- :m.- sm.om rm s\ooH a s HHomaH scam: nos HH H:.o oHs. oo.HH Ho.HHI rm sos .s .Hom sHmooo + amass oH H:.o oHn. os.o mm.s so so& .2 .Hon woos: nos 0 om.0 om:.I mm.- :m.sm rm sos .AoosvsooHo oaowaooooIoHos o mm.o om:.- o:.- om.o: rm sos .Aosovmosoerrmdoo HHa: HHoo s H:.o ass. Ho.o om.HHI mm s .so .Hom sHmooo o mm.o cos. no.HH mo.mHI mm sos .z .Hom sHmooo m mm.o cos. sm.mH oo.mHI mm sos .z .HommHado + choo + amass : .4:.o on.I Ho.HI s:.om Hm AzxooH a s- ssv omsoHoa ss noses m os.s 6mm. mo.o oo.msI rm s.so .Hom sHmooo + saosss m om.s omo. sm.H Ho.omI :m s .oso wHon tors: nos H mam H HQ on Q N COHumddm o o o a . mpceHoHMHooo GOHmmoowem mSOHaw> no AMV UoQHOmQ< mo R mm Gomoasz UoGHmpom .meSHw> haouwoonmq pom mGOHmmeamom awoGHH .mH oHnt 105 os.m ooo.- sm.moI om.os :m Asaoss as ooo maHszoao oo cosmoo mm mo.m oso.I mo.mI :m.oo Hm s\ooH a sImss omaoHoa mss amass Hm mm.m aso. mo.:H oo.mH :m sow .z .Hom osowaoacoIoHos om oo.m aHs. :H.H oH.oH no so a .Hom topaz nos oH :0.: wmw.I mm.NNI :H.:0 :N 20R .2 .HOmGH GHmmom + coesm 0H 00.: mi. m0. 8.: mm so R .Hom .83 + onooo RH 00.: wm0.I :0.HNI mH.00 NN 20R .2 .HOmCH .me + choom + Cosdm 0H om.: oss.I so.omI :m.ms :m sow .s .HomsH raomaopooIoHos mH 0N.: 000.I 00.0HI 00.H0 0N SQR .z .HOmGH chmom :H NN.: mH0. 00.H 00.:HI 2N R .20 .Hom GHmmoQ + Goesm 0H SH.: an. oo.SH oH.mH :m soR .s .Hom aHmooo + amass mH oo.: smo.I om.HI om.ms :m Z\ooH a z .HomaH roomsooooIoHos HH oo.m cso.I so.I oo.ss rm s\ooHs s .HomaH anooo + scans oH so.m amo. oo.oH mm.:H mm sos .2 .Hon .sao + aHmooo + amass o sm.m sso.I HH.H- m:.oo :m soR .Aoosv sHamHH oaomaooooIoHos o oo.m dso.I mo.I mm.Ho mm Z\ooHsz .HomaH .sao + sHmooo + amass s om.m moo. oo.oH om.mH mm soR .s .Hom sHmooo o ::.m doo.I os.I mo.mo rm sos .Aosoo mrsosoHonaoo HHa: HHoo m o:.m aom.- ms.I sm.ms mm s\ooH a s HHomaH sHmooo : 0N.m w00. 00. 00.m :N R20 .HOm wnwm + CHmmom + mcoadm m :m.m aso. oo. oo.mH mm R .aoroas see o.Hom aHmooo m ms.m m:o.- oo.I os.mo so oso R .AoosvaooHo odomaooooIoHos H o mam a HQ on c N COHdevm o o o a . mpGoHOHNQeoo COHmmeamom .mode> shepmooan mfiOHam> Go AMV mponoHMMooo GOHpmomHQ soups: ham mom mGOHmmoomom HonHH .0H oHQwB 106 00.0 v@ .I. 0 “H00 v0 .HGHDOQOHE 30089 H HHQEHHm .oHnsHom u . m 0m H o .moo so u .H w o “mooovo n 6 “Hood vo u o “mooo yo .I. m .aeppmfi how u 20m .omeHpmo mo scope vswwqwum N mmmm .erHOHmmeoo GOprHooooo n so .oHQSHomQH H .HOmaHh .eQOHm HHQm .Gprosoch H .meH .pmoooopGH non< m .o osm. :m.o om.sm :m sowoaoHs mm m .o om:.- om.I mo.oo :m s\ooH a s .HomsH .aso + sHmooo so ms.m com. om.mH :S.mm :m sos .s .Hom .aso + aHmooo mm H o mom a o o a s soHrmsoo meoHOHHMooo GOHmmoowom Aveschooov .0H oHnt 107 were observed for Equations 1 and 2 based on acid deter- gent insoluble dry matter (acid detergent fiber) and pepsin soluble dry matter respectively. The regression coefficient b1 of Equation 2 based on pepsin soluble dry matter was 0.99, indicating the action of the pepsin solution in solubilizing forage dry matter was remarkably similar to that of in 1132 dry matter digestion. However, the value of bO indicated that the in ziyg_digestion system was able to solubilize 12.6 percentage units more dry matter than did pepsin. This 12.6 percentage units dry matter could presumably consist of fibrous consti— tuents that rumen microbiota could solubilize but that were not solubilized by the pepsin incubation. Actually, this assumption was supported by the fact that action of in_zitrg rumen microbial + pepsin + pancreatin solutions (Equation 3) was even closer to actual in 1112 dry matter digestion processes than that of the pepsin solution (Equation 2) alone since the b 0 value was smaller for Equation 3 (5.1) than for Equation 2(12.6). Dry matter solubility in a rumen microbial + pepsin solution has been widely used as one of the most reliable predictors for in 1112 dry matter digestion coefficients of normal forages (Tilley and Terry, 1963; Oh gt 31. 1962; Van Soest, 1973) but in the present 108 study the r value was only 0.81 (Equation 13). The regression coefficient b1 was approximately 1.0 and would suggest a close similarity of action for this .in.yitrg method with that of the in yizg process. The negative value of bO indicated that rumen microbial + pepsin solutions solubilized more dry matter than occurred in the in £339 digestion processes. This di- fference probably can be explained by the fact that some undigested microbial dry matter is associated with the indigestible forage dry matter (Van Soest, 1969). When forage DM is partitioned into soluble and insoluble fractions by certain kinds of solvents, esti— mates of in ziyg digestibility of each dry matter frac- tion can be obtained from regression coefficients. For example, the acid detergent method separates forage dry matter into acid-detergent soluble dry matter (cell con- tents and hemicellulose) and insoluble dry matter (cellulose and lignin or acid detergent fiber). The regression equation for in 1312 DMD with AD insoluble DM (ADF) was calculated and given in Table 19 (Equation 1). The actual digestibility of ADF can be estimated from that prediction equation by extrapolating the level of ADF to 100. In other words, if forage dry matter was all ADF then the estimated in yiyg DMD would also be 33 £332 ADF digestibility. Using values from the present study, the in vivo ADF digestibility was 3.79% (100 x 109 0.90 (b1) - 93.79 (b0), Equation 1, Table 19). Similarly, .12.!1X2 digestibility of the AD-soluble DM can be esti- mated by using the same equation and extrapolating the ADF level to 0% and the value thus obtained for this study was 93.79 (93.79 + (-0.90)(0%) Equation 1, Table 19). Following this method of calculation, in 1112 digestibility estimates of various forage dry matter fractions were calculated and are presented in Table 20. Data obtained from other sources were also calculated and presented in Table 20. For the present study the estimated soluble forage dry matter fractions ranged from 133 (hot water soluble DM) to 86 (AD and 72% H2SOA soluble DM). Van Soest (1963) reported an estimated in yixg digestion coefficient of 98% for cell contents of normal forages, the value of 96% obtained in the present study suggests that our samples are probably normal. 0n the other hand, in 1119 cell content digestibility was estimated as only 73% by using combined samples. The reduced digestibility probably was due to heat da— mage Since a certain portion of potentially digestible protein in the cell contents would be converted to in- digestible material (e.g. acid detergent insoluble N). Van Soest (1973) has stated that the action of acid pepsin solution is probably very similar to that of neutral detergent solution, but in the present study 110 CmOHSHHoOHEom QoHpsHom SH astHo oHH Mmmwmwwwowwww aHrsoaoaao emOHSHHoo + eHmmom emOHSHHCOHEem COHpsHom HH aHsmHo mHH MMmMMWWMoMfiMM sHmooo omOHdHHeo 0Ho< : m . toast .mmhhm mmHI m- aHomHo ms to omoHsHHooHsos roomaoroo mpoepcoo HH00 0He¢ GOHpSHOm CHGNHH 2 0N : . . m0 :0 monopooo HHoo poemoopoo 0 omOHdHHoo 0Ho¢ GHCNHH QOHpsHom Nm N: 0H omOHSHHoo 00 00 monopooo HH00 powwoopoo omOHdHHoOHEem HwAHSez IIIIIIIII R IIIIIIIII IIIII R IIIIIII 002 noeoHnEoo wbmz mosespHpmmoo QUeQHQEoo wbmz mpoodepmooo GOHpowom QOHpomHH vogue: oHosHomaH oHosHom seHHHosHom .mssos HaoHanrarm so oooaermm mCOHpowam Lopes: ham emwoom mSOHaw> 0o meeHOHmmooo QOHumomHQ o>H> CH .0N oHQmB 111 .eowoapH: cHopoam Gee H 2020 .Qooflm onms UoCHanoow meoHOHmmeoo COHpmomHU o>H> mw.cwozo .om H G .AwHOmoeonV Gemoon + AchQoomH3v oomoemaow + Aoethowzv mGHooow + 002 H UoQHQEoon .mm H a .sonaoaHao cream adeooHs n omsa mmOHdHHoOHEem mUHQHH 0H GHGNHH 00H mHmNSm oHQEHm Lopez 000 emOHsHHeo czwz mGHopoom COHHSHOm emOHSHHoU GproHoch m QHGNHH 00 oMOHSHHeOHEom +QHmmom + mmOHsHHoo mpoepeoo HH00 HprooOHE ooEdm COdeHom omOHsHHoo mHI GHCNHH :0 omOHSHHooHEom QHmmom + omOHdHHeo mpoepeoo HHoo HwHQOHOHE Goadm IIIIIIII R IIIIIIIIII IIII R IIIIIII 002 UeQHnEoo pm: mpeedepcho UeQHQEoo 00: meeSHHpmcoo cOHpowam SOHpowsk venues cHosHomsH oHosHom soHHHosHom AvoSGHpooov .0N eHan digestibility of pepsin soluble DM was more than that of neutral detergent soluble DM (112 vs. 96%). Similarity in estimated in 2312 digestibility of soluble dry matter was observed among similar analytical methods. For example, digestibility for acid pepsin soluble dry matter was close to that for pepsin + pan- creatin soluble dry matter, and the digestibilities of soluble dry matter were similar between two in 21333 rumen microbial + enzyme digestion methods. Using data of the present study, estimated in 1113 digestion co- efficients for insoluble DM fractions ranged from -15 to 19%. Lignin digestibility is zero or nearly so (Van Soest, 1969) and the digestion coefficient of -3 esti- mated from the present study would support the concept that lignin is rather indigestible material. However, this negative coefficient could also be due to the effect of heat damage since lignin undergoes qualitative and quantitative changes when forage is heat damaged (Cym- b31Uk.2£.El' 1973; Van Soest 1965). In fact, the es- timated lignin digestibility was -123% using combined sources of samples which include a large number of heat damaged forages. The digestion coefficients estimated from the present study for cell walls and lignocellulose were 17 and A (Table 20). However, the actual deter- mined digestibilities for cell walls (neutral detergent 113 insoluble DM) and lignocellulose (acid detergent insolu- ble DM) were about 50% which is much higher than the value estimated by statistical means (Table 20). This difference suggests that in the 12;X1X2 condition, the negative effect of lignocellulose on dry matter diges- tibility probably is not a linear type response. Dry matter digestibility was more related to acid detergent insoluble N as a percent of total N ( r = -.82, Equation 11, Table 19) than to the term acid detergent insoluble N as a percent of dry matter (r = -.77, Equation 15). This is opposite to that in the report of Goering £3 21. (1972) where acid deter- gent insoluble N as a percent of dry matter was superior to acid detergent insoluble N as a percent of total N ( r = -.90 vs. -.70) for predicting energy digestibility. Furthermore, they found that the pepsin insoluble N fraction was less related to energy digestibility than was the acid detergent insoluble N fraction (r = -.79 vs. -.90) which is opposite to that of this study (r = -.88, Equation A vs. -.82, Equation 11, Table 19). Results from the present study indicated that in yiyg dry matter digestibility is primarily correlated with dry matter solubility determined either by chemical methods (acid-detergent solution), or in yitrg enzyme digestion methods (acid pepsin solution), or in vitro 11A rumen microbial + enzyme methods (rumen microbial + pepsin + pancreatin). Nitrogen solubility determined by acid pepsin is also significantly correlated with .in‘yiyg dry matter digestibility. Considering the values of r and standard error of estimate the best predictor for in 1113 dry matter digestibility was acid detergent insoluble dry matter (ADF). (5) Regressions of Maximum Dry Matter Intakes Linear regressions and correlations for maximum dry matter intakes with various laboratory measurements are in Table 21. From the total of 31 variables only 11 are shown in Table 21 and only rumen microbial + pepsin soluble dry matter showed a reasonably high correlation with maximum dry matter intake ( r = .82, Equation 1).. Presumably other factors not evaluated in this study (e.g. dry matter content, leaf-stem ratio, organic acids, pH etc.) were more closely related to voluntary intake of forage dry matter by ruminants than were the 31 items studied. IV. Relations Among Nitrogen Containing Fractions Correlatation coefficients among nitrogen con- taining fractions are in Table 22. In general, high correlations were observed among fractions obtained from similar analytical methods. For example, high correla- tion coefficients were noted between the following 115 .ooppwe how u 2Q o .eHQSHOmGH n .HomGHM .OHQSHom u .Homm .HwHQOHOHE CeESa n Geesm .GmmoopHc n z m h .opwEHpmo mo moose cawwowpm u mmmm .Qproooewm H .qwm H .peoHOHmmeoo GOHuwHooaoo H 90 .m0.0.Am n m ”m0.0_vm n o .eQOHm ”Hem “Ho.o_vo u o umooo vo u o o “H000 vm u o. “m000.0 vm H mm 1.309.3ch n as. so.o coo.I oo.I 0H.o do so0 .Aooso acoHo rcomoopooIoHos HH :o.o ooo.I mm.I m0.m :m soo .Aoosv chwHH osomaopooIoHos oH :o.o oHo.I so.I oH.m rm s\ooH a s .HonaH sHmooo + amass o mo.o omo. oo. mm.I mm so R .Hom .aao + sHmooo o mc.o mo. 0:.H mH.I :m soR .s .Hom aHnooo + amass s mo.o moo.I oo.I Hs.o Sm soR .AosovmoaosoHomsoo HHsz HHoo o om.o ooo.- so.I ::.m mm s\ooHas wHomaH .sdo + sHmooo + dossm m Ho.o poo. H:.H :o.- mm soR .s .Hom sHmooo : sm.o woo. mo.H 0m.I mm soR .Hs .Hom weao + sHmooo + coeds m Hm.o ass. :H. oo.HI :m R .so .Hom hood: oos m o:.o amo. mH. or.:. :m R wso oHom sHmooo + masses H mmm o HQ on a x SOH adv o o o s .o o mpoeH0H0mooo QOHmmoomom .moSHw> keepmooan mSOHom> so ANV Amooflm no oocHEHopoQ ozmHoz 000m RV magma ICH soups: ham ESEHNwE Mo mpGeHoHMRooo COprHoaooo 02w mQOHmmoomom amoeHH .HN eHQwB 116 m-oH H-oH 0-0 0-0 m-0 H-0 0-0 m-0 H-0 R 0 m 0 H m H 00.. 00.0 00.0 00.- 00.- 00.0 R0.0 00.- H0.- 00.- m0.- 00.- H0.. m:.- 00.0 00.0 20R .2 .Hom .ddo+deooo+cossm s-HH Rm.0 m0.- 00.- 00.0 H0.0 H0.- 0H.I 00.0 ms.0 H0.0 00.0 00.0 00.0 00.0 00.0 mm.0 soR.z .HoacH .cao+chooo+coess m-HH 00.0 ms.- 00.- 00.0 00.0 rs.- 00.- 00.0 00.0 00.0 00.0 0.0 00.0 00.0 :m.0 00.- z\00sz.HomcH .Cao+:HmoooIcosss m-HH 00.- No.0 HH.0 00.- 00.- 00.0 m0.o H0.- 00.- ms.- 00.- 0.- mm.- 00.- mm.- 00.0 soR .Hoa .asoIcHaooo+:osam H-HH 0H.I s0.0 00.0 00.- 00.- HH.0 H0.0 0m.- os.- 00.- s0.- 00.- 00.- 00.- 00.0 HH.0 soR .z .Hoa somooo I 00200 s-mH 00.0 00.- 00.- 00.0 m0.o 00.- .m.- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0H.0 20R .2 .HoacH camoao + doses H-mH 00.0 H0.- 00.- 00.0 00.0 00.- 0 .- 00.0 H0.0. H0.0 00.0 00.0 R0.0 00.0 00.0 HH.- z\oonz .HoncH chneo + steam m-NH om.- HH.0 0R.0 ms.- 00.- 00.0 00.0 00.- 00.- m0.- 00.- 00.- 00.- 00.- 00.0 00.0 20R .Hon sonata +ocosss H-mo H:.0 0H.I HH.0 HH.0 :m.o 00.0 00.0 0.0 0H.0 0H.0 HH.0 0H.- 00.0 0H.0 00.0 m0.0 20R .2 .Hoa toad: pom 0IHH 00.- H0.o 00.0 0H.- Hm.I 00.0 00.0 H.- 00.- 00.- 00.. NH.- Rm.- RH.- m0.- 0H.- 20R .2 .HoncH teas: 0oz HIHH 00.0 00.0 0.- 0H.- sm.- 00.0 0m.- 0.- mm.- mm.- 0m.- 00.0 00.- H0.- 00.- 00.- z\oon z .HomaH toad: nos m-HH 0H.- 00.0 0s.o 00.- 00.- HH.0 00.0 mm.- 00.- 0.- 00.- 00.- 0H.- Hm.- 00.0 00.0 R .20 .Hon nous: 0oz H-HH 0a.- RR.0 H0.n 0H.0 00.- 00.0 00.0 H.- s0.- s0.- H0.I HR.- 0H.- 00.- 00.- H0.0 20R .2 .Hoa .aao I condos s-oH 00.0 00.- 00.- 0..0 m .0 0H.- 0o.0 00.0 HH.0 00.0 mm.o 0.0 00.0 0H.0 00.0 00.0 20R .2 .HoacH .000 + canoes H-oH 00.H ms.- sm.- 00.0 00.0 m .- 0H.I 00.0 Hs.0 00.0 00.0 00.0 03.0 HH.0 00.0 00.0 z\00Hsz .HoncH .000 I :Hmoco mIoH 0,.H 00.0 00.- Hm.- 00.0 00.0 00.- H0.- 00.- 00,- 00.- 00.- 0m.- 0.- 0m.0 R.so .Hoa 0.:00 I cHnooa H-oH 00.H 00.- 00.- 00.0 00.0 e0.- 0a.- RR.I H0.I 00.- .m.- 00.- 0H.0 00.0 20R 2 .Hon chooa 0-0 00.H 00.0 00.- Rs.- 00.0 00.0 00.0 00.0 00.0 0.0 e .0 0H.0 00.- zxoon z .HoacH cHaooo 0-0 00.H 00.- 00.- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 No.0 0H.0 soR .z .HomcH :Haooa m-0 00.H 00.0 00.- m0.- 00.- 00.- 00.- mg.- 0m.- 00.- 00.0 R .20 .Hom cHwooa H-0 00.H 00.- 00.- 00.- 00.- R0.- 0H.- 0H.I 0m.0 00.0 soR .z .Hom ocomnooooIoHo< 0-0 00.H 00.0 00.0 H0. 00.0 m0.0 R0.o 0H.0 00.- soR .z .HoncH acomaoooe-eoos m-0 00.H 00.0 00.0 00.0 s-.0 Hs.0 00.0 mm.- z\00sz 0HonaH ocowtoooo-0Ho< H-0 00.H 00.0 00.0 0.0 Hm.0 0H.0 0H.I soR .cacon snowsoooo-0H6< s 00.H H0.0 00.0 00.0 0H.0 00.- soR .aano acomtoooo-oHo< 0 00.H HH.0 RH.0 0H.I 0m.- soR .mocosoHoacoo HHa: HHoo m 00.H 0m.0 0N.0 mN.0 Assoid ovwchzoan mo seamen 0 00.H 00.0 0H.0 A2\00sz- mzoouaoHoa 02 cases 0 00.H 0H.0 z\00sz 0Hon Loosen HatoaH: m 00.H ssoR .aHoooao costo H uCoHoHuhooo coHusHoLLoo oHQEHm :oHWMMWMHWHocH QBMWO .0oHvzum mquEoLSmaoz Hm ecu wooE< mcoHuoHoapoo oHQEHm .NN oHnma 117 .Hwanopoae nosdh u coEdmo .cfiudonoqaa u .ndn 0 .oHQSHOmcH n .Homcao .oanzaom ".0009 .Loupus haw u :00 0-00 0-00 0-00 0-00 0-00 0-02 0.02 0.00 0-00 0-00 0.00 0.20 0-00 0-00 00.H 00.- 00.- 00.0 00.0 00.- 00.- 00.0 00.0 00.0 00.- 00.0 00.0 0H.- 200.2 .000 .cda+000000+:ms:2 0-0a 00.H 00.0 00.- 00.- 00.0 20.0 00.- 00.0 m0.- 00.- 00,- 00.- 00.0 200.2 .00020 .qdaIchmoaIsoesm 0-00 00.0 00.- 00.- 00.0 00.0 00.- 00.0 00.- 00.- H0.- 00.- 00.0 2\00022.Homcd .camIcfimaoaIcossm m-0H 00.0 00.0 00.- 00.- 00.0 00.- 00.0 00.0 00.0 00.0 00.- 200 .Hom .qaaIcfluaoaIcossm 0-00 00.H 00.- 00.- 00.0 00. HH.0 00.- 00.0 00.0 00.- 200 .000 camaogI 005:2 0-00 00.H 00.0 00.- 00.0 00.- 00.- 00.- 00.- 00.0 200.2 .0000“ 200000 I 005:2 0-mH 00.0 00.- 02.0 00.- 00.- 00.- 00.- 00.0 2x00022.00020 000000 I 20202 m-ma 00.0 00.0 00.0 00.- 00.0 00.0 00.- 200 .000 220000 I 205:2 H-NH 00.2 00.- 00.- 00.0 00.0 00.0 200.2 .000 20002 002 0-00 00.H 00.0 0 .- 00.0 00.- 200.2 .0002“ 20003 002 0-00 00.H 00.- 00.0 00.- 2x00022 .0000“ 2000: 002 0.00 00.0 00.0 No.0 «.20 .000 20003 002 0-00 00.H m0.- 200 .2 .000 .anI 200000 0-00 00.0 200.2.00000 .caQI 000002 0-00 coaumofimfipconH .oc udmaoammooo copraohnoo oHQEHm oanafipd> Qsopc 20002000000 .00 00000 118 analytical fractions: Hot water insoluble or soluble N and mineral buffer soluble N ( r = > -.82), rumen microbial + pepsin insoluble or soluble N and rumen microbial + pepsin + pancreatin insoluble or soluble N (r = .55 to .97). Nevertheless, there were some high correlation coefficients among fractions derived from completely different analytical methods. For example, a r value of .90 was found between pepsin insoluble N and acid detergent insoluble N both expressed as a per- cent of dry matter; a correlation coefficient of .93 was found between pepsin insoluble N and the degree of brow- ning (brown color); and a correlation coefficient of .95 was found between rumen microbial + pepsin insoluble N as a percent of dry matter and the degree of browning (Table 22). Goering EE.EL° (1972) also observed high correla- tion coefficient ( r = .92 ) between pepsin insoluble N as a percent of total N and acid detergent insoluble N as percent of total N but the regression coefficients (b = 17.02, b = 1.00) from their study were markedly O 1 different from that calculated from the present study (bO = .18, b1 = 1.72). Their study indicated equal increases of both with additional increments of heat damage and that about 17% of the nitrogen was pepsin insoluble when acid detergent insoluble N content was 0. 119 They concluded that there was a major fraction of N in- soluble with pepsin but soluble in acid detergent.p Re- gression coefficients of the present study suggested that both analyses measured similar fractions (bO = .18) and that amount of pepsin insoluble N would increase 72% (b1: 1.72) more than would acid detergent insoluble N with every additional increment of heat damage. The correlation coefficients observed between the following variables: AD lignin and AD insoluble N, r = .85; ADF and AD insoluble N, r = .71; and cell wall constituents and AD insoluble N, r = .63 confirmed the concept of Van Soest (1965) that the additional N in the acid detergent insoluble fraction of heat damaged forages is primarily associated with lignin but is less associa- ted with acid detergent fiber or cell wall constituents. V. Relations Among Various Dry Matter Solubility Measurements Correlation coefficients among various dry matter solubility measurements are scattered throughout in Table 22. Generally, dry matter solubilities measured by different methods were rather well correlated with each other (r ranged from .9A to .59). Dry matter solubility in pepsin was highly corre- lated with acid detergent soluble DM (r = .9M) and neu- tral detergent soluble DM (r = .90) but the linear 120 regression coefficients indicated that the function of pepsin more resembled that of neutral detergent ( b = O 9.hu, b1 = 1.06) than that of acid detergent (bO = 12.50, b1 = 1.0a). Similar statements were made by Van Soest (1973). The correlation between acid detergent soluble DM and rumen microbial + pepsin + pancreatin soluble DM was .89. The regression equation had b0 and b1 values of h-99 and .97 respectively indicating that the func- tions of these two different analytical methods were similar. This may be related to the high negative corre- lation coefficient between acid detergent insoluble DM (ADF) and in vivo DM digestibility. VI. Regressions of Five In Vivo Parameters on Thirteen Selected Laboratory Determinations Grouped from An Operational Standpoint Although a total of 31 different laboratory ana- lyses were performed on each forage sample, some ana- lytical values were grouped together based on labora- tory operations. For example, values for pepsin soluble dry matter, soluble N % DM, insoluble N %DM, and in- soluble N x 100/N can be obtained from one analysis scheme on one sample. Regression coefficients for the five in vivo parameters using each of these 13 groups are in Table 23. Because the number of variables in 121 00.- 00.0- 00.0- 00.0- 00.0- :00 .0000-0 00.0 00.00 00.00 00.00 00.000 00000000 000.0 000.0 000.0 000.0 0000-0. 00 00 0 00.- 00.- 00.- 00.- 00.0- :00 .0000-0 00.0 00.00 00.00 00.00 00.000 00000000 000.0 000.0 000.0 000.0 000.0 00 00 0 00.- 00.- 00.- 00.- 00.0- 200 .0030-0 00.0 00.00 00.00 00.00 00.000 00000000 000.0 000.0 000.0 000.0 000.0 00 00 0 00.0- 00.00- 00.00- 00.00- 00.000- 000000 00 000 .mcficzonn mo oomon-N 00.0 00.00 00.0 00.00 00.00 00000000 000.0 000.0 000.0 000.0 000.0 00 00 0 Az\0m0x z- 0020 00.0- 00.0- 00.0- 00.0- 00.0- 0000000 02 00000-0 00.0 00.00 00.0 00.00 00.00 00000000 000.0 000.0 000.0 000.0 0000.0 00 00 0 00.0 00.0 00.0 00.- 00.- 2\00000000m0000000-0 00.0 00.0 00.0 00.00 00.00 00000000 000.0 000.0 000.0 000.0 000.0 00 00 0 00.0 00.0 00.0 00.0 00.0 200.0000000 00000-0 00.0 00.00- 00.00- 00.00 00.00 00000000 000 0 000 0 000 0 000 0 000 0 0m 00 0 000000 00000000 2 0 . . 00002 00 00000000 0 00000000 0 0000000 20 0000000 0 0 0000 me000000000 QOHmmmpmmm 0 .0000000000 - 00:000000Qo :0 8000 vamSOLG 000 £00£3 0Q0000c0enopom 0000000900 m0 so mmmnm Eonm woq0enopma whopmfiwawm o>0> mm.o>0m Mo mpflo0o0k%ooo GOHmmopmom .mm 00909 122 000. 000.-00m. - 0000 000.-000.00- 00.0 00.0.00.0 00.0 000.00 0.0 0000 000.0000.0 :0 .z.0o. .0ua+.aoa+0onzm 0-00 00000 000.000 .0 00000 000.000.000 00000 000.-000.0- 00.0- 000.-0 0.00- 00.00- 000.-000.00- :0 .z.0o.00.000+.000+0oesm 0-00 00.- 000.-000.- 0>00 000.-000.0- 000. 000.0000.0 .00 000.0000.0 000 000.-0 0.- z\z.0o-00.000+.000+0oasm 0-00 00. 00.-000.- 000 000.0000.0 :00 000.-000.- 00. 000.0000.0 00.0 000.0. .0 :00.000 .0ua+.aon0coasm 0-00 00.0 00.00 00.000 00.00- 00.00- 00.00 00.00 00.00 00.00 00000000 000.0 0000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 mm 00 00 000 000.-000.0- 000 000.-000.00- 00.0 000.0000.0 00.00 000.000.0 00.00 000.0000.00 :0 .z .000 .000+ 00200 0-00 00000 000.0000.0 00000 000.0000.000 000 000.-000.0- 000 000.-000.0- 00.00- 00.-000.00- :0 .z.00000 .000+ 00000 0-00 0000 A00.-000.- 0000 000.-000.0- 000. 000.0000.0 0000 000.-000.- 000 000.-000.- z\z.00000 .000+ 002:0 0-00 00.0 000.000.0 00.0 000.0000.0 0000. 000.0000.0 00.0 000.0000.0 0000 000.000.0 .100 .000 .aog+mcoesm 0-00 00.0- mm.b 00.00- 0. 0 00.00- 00.00- 00.00- 00.00 00.00 00.0: 0:000:00 000.0 0 .0 000.0 0 .0 000.0 000.0 000.0 000.0 000.0 000.0 00 00 00 0000. 000.-000.0- 0000 000,-000.00- 000 000.-000.00- 00.0- 000.-00.00- 0000 000.-000.0- :0 .z .000 0000: 000 0-00 000 000.000.0 000 000.0000.00 00000 000.0000.0 0000 000.0000.0 00.00 000.0000.0 :0 .z.0on00 0000: 000 0-00 0000 000.-.00.- 00000 00.-000.0- 0000 000.-000.- 000 000.-000.- 00. 0000.-v00.- zxz .00000 0000: 000 0-00 00.0 000.0000.0 00.0 000.0000.0 00.0 000.0000.0 00.0 000.0000.0 00.0 000.000 .0 :00 .000 000.: 000 0-00 00.0- 00.0 00.00- 00.00 00.00- 00.00 00.00 00.00 00.00- 0 .0- 00000000 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 00 00 000 000.-000.- 00V 000.-000.00- 00. 000.0000.0 0000 000.0000.00 000 000.0000.00 :00.z .000 .000+ .000 0-00 00000 000.0000.0 00000 000.0000.000 00000 000.0000.0 000 000.-000.00- 00000 000.-000.00- :0u.z.00000 .000+ .000 0-00 0000 000.-000.- 0000 000.-000.0- 0000 000.-000.- 00000 00000000.0 0000 000.0000.0 z\z.00000 .000+ .000 0-00 00.0 000.0000.0 0000 000.0000.0 00.0 000.0000.0 00.0 000.000.0 00.0 000.0000.0 :00 ..000 0000 + .000 0-00 00.- 00.0- 00.00 00.0- 00.00- 00.00 00.00- 00.0 00.00- 00000000 000 0 000 0 000 0 000 0 000 0 000 0 000 0 000 0 000 0 00 00 00 00.0 000.000.0 000 000.-V00.m0- 00.0 000.0000.0 000 000.-000.- 0000 000.-000.00- :0u.z .000 .000 0-0 00000 000.-v00.- 0000 :00.-000. - 000 000.0000.0 00.- 000.-000.0- 00.0- A00.-000.0- :\000xz.00000 .000 0-0 000 000.0000.0 00000 000.0000.000 0000 000.-000.00- 0000 000.0000.00 000 000.-000.00 :00 .2 .00000 .000 0-0 .000 000.0000.0 0000 000.0000.0 00000 000.0000.0 00.0 000.0000.0 00.0 000.0000.0 :ou .0000.000 0-0 000.0 00.0 00.00 00.00- 00.00- 00.00 00.00 00.00 00.000 00000000 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 00 0 00.0 000.0000.0 00000 000.0000.00 000.0000.00 00.0 000.0000. 0000 000.0000.0 :00 .2 .000 00 0-0 0000 000.-000.0- 000 000.-000.000- 000.-000.00- 00.00- 00.-.00. 0- 000 000.-000.00- :00 .: .00000 00 0-0 000 000.000.0 0000 000.0000.0 000.0000.0 0000 000.-000.0- 00.0- 2000.-000.0- z\000xz 000000 0-0 00.0- 00.0- 00.00- 00.00- 00.00 00.00 00.00 00.00 0 00000 000.0 000.0 000.0 000.0 000.0 000.0 000.0 000.0 00 00 0 CO0uoaoc CO000000 £0000000 :0000000 CO0uoHoU c0000000 £00uoaou :0000000 CO0uoacv Ceauoaov LeuMK- ohmhwm Leuu< onouwm Leuum oaouom Louu< onouem Louu< opouom nonpomna R 00 000000 200000: 00000000 00000000 0000000 00000002 0000000000000 :0 0000000z 0000000000 nuC00o0uaoou co0maomwdm : nacho. AUOSCOucoov .mN 00906 123 .HwfinohOfiS Cmesn n coEdmm .Cfiuwmnoqwm n .med .mo.o Swap pmpwmhm mm: monwoflMHanm menu can? .QOHuwSUo scammopmmn map 8099 wouoamo mmanwamwb mo monmdvmm map mpnmmmpgmh Hdnmnsdc nwEomm .cflmmmm n .mmmo .mpcofloHMMmoo Soapwamnmoo Hwfippmm mpw mmmmnpcmhmm Ga mmdaw>z .oHQSHOmGH u .HOmnHS .pq®MpoumU Uflow n Q mm pmmzmfls map mmw cmqaamowqd mmSHwbm .Gficmfifi pcmmmmpmc @How H ammH moqmoflmficmfimw .mansfiom u .Homm .mmgwdn Hwhoqflzm .Apnwflmz %Uon RV oxwucfi pmppwe haw Edefixmz .Ahwv\z wv monHwn Emmopuflz Q o .R .muaoflofimmmoo Gowpmmmflc Ampuwe hum ..& “mpcmfioHMMmoo Goapmmmflv GmmOLsz< m Aumsqfipaoov .mm magma 12h each of these 13 groups ranged from one to four, both simple linear and multiple regression equations were obtained. In addition, least square stepwise deletion multiple regression was computed so that variables not significantly (P< 0.05) regressed with in zivg_para- meter were removed from the final equation. The great- est correlation for N digestibility was obtained with lignin as a percent of dry matter (Group 7). Slightly lower R2 values were noted for Group 9 - the pepsin so- lubility measurements (R2 = .87) and Group 8 - the acid detergent solubility measurements (R2 = .86). Protein content (Group 1), protein solubility in mineral buffer (Group 2) and EE;XEEZE rumen ammonia release had no value for predicting the in Kilo N digestibility of a forage (Table 23). The greatest R2 values for DM digestion coefficient and N balance were for the pepsin solubility measurements (R2 = .95 and .78 respectively). None of the 13 groups was able to produce a high R2 for the regression of N retention as a percent of absorbed although.the highest R2 (.u9) was found for rumen microbial + pepsin solu- bility measurements (Group 12). This group also gave the highest R2 for the regression of maximum intake (R2 = .73). Among these 13 groups of determinations only three 125 (rumen microbial + pepsin solubility measurements, acid detergent fiber and hot water solubility measurements- # 12,6 and 11) had reasonable correlations with all five ig_vilg_responses. Four groups (pepsin solubility, rumen microbial + pepsin + pancreatin solubility, acid detergent solubility measurements and acid detergent lignin) were highly correlated with the three more pre- dictable in 1119 parameters of N digestibility, DM di- gestibility and N balance. Partial correlation coefficients for variables in each group containing more than two variables are also given in Table 23. A variable could have a high partial correlation coefficient with one in 1112 res- ponse but have a rather low coefficient for another in viva response. For example, among four pepsin solubility measurements (Group 9) pepsin soluble dry matter possessed the greatest partial correlation coefficient for in XEKQ N and DM digestion coefficients but not for N balance. Pepsin soluble N% DM was more related to N balance than was pepsin soluble DM. Least square stepwise deletion programs were used to delete out insignificant (P<(.OS) variables in those groups having more than two variables. Generally, in- soluble N as a percent of forage total N was relatively unimportant in predicting in vivo parameters as compared 1 I‘() with other solubility measurements (Groups 10,11,12 and 13) except for AD/N and pepsin soluble N/N for DN. In most groups, two out of four variables were deleted and the sequences of deletion were negatively related to the partial correlation coefficients. For example, when DM digestibility was regressed on the four pepsin solubility values (Group 9), pepsin soluble N, %DM was the first variable deleted and it also had the lowest partial correlation coefficient. Pepsin insoluble N, %DM was the second variable deleted (Table 23). The new equation included only two variables and possessed essentially the same R2 as the equation using four variables (.95) and this was also true for many other groups (such as acid detergent solubility Group 8, and rumen microbial + pepsin solubility measurements - Group 12). Thus in Kilo responses can be predicted by using fewer variables with about the same degree of precision with these equations from the least square deletion program as when using all variables. In practical situations the selection of labora— tory measurements to predict in vizg_forage nutritive value will be based not only on precision of prediction but also on other factors such as time, specific equip- ment and labor required. When all these factors are 127 taken into consideration measurements for acid detergent solubility (ADF, Group 6) and pepsin solubility (Group 9) bec0me the most desirable items to predict N and DM di- gestibility and N balance of forages. VII. Multiple Regression of Five In Vivo Parameters With Measurements of Two Laboratory Determinations A second sequential process to select the most appropriate multiple regression equations is to randomly use combined measurements from any two of the 13 groups of laboratory measurements and then to use a least square deletion program. In this process a total of 78 regression equations were generated but only important regression equations will be presented and discussed. Table 2h gives three equations which gave the highest R2 values for predicting i_n_ v_iv3 N digestibility. Equation 1 and 2 used variables of pepsin solubility (Group 9) plus either hot water solubility (Group 11) or rumen microbial + pepsin solubility (Group 12) mea- surements and gave remarkably high R2 values of .96 and .95. These multiple regressions improved R2 by only about .08 units when comparing these R2 values to the highest R2 (.88) obtained from regressions using only one group of measurements (AD-lignin, Table 23). Re- gressions in Table 2h are certainly much more complicated using many more variables and laboratory analyses. 128 - Aea.ovom.oa 2mm .2 .Hom teem: pom :-HH Aew.ovma.ma Am:.ovoa.ma 29m .2 .Homce teem: pom m-HH - Aeo.ov0H.o 2\00sz .Homce topaz pom m-HH Aom.ovem.a Ame.ovoe.o 2mm .Hom tees: pom H-HH lee.-vee.em- Awe.-voe.ae- Asses: eevmeeezoee e0 eeewoe : No.0H :H.m- eqepmqoo :m.m He.o em.m me.o mam one mm m - Amm.-vmm.:a- mm .2 .Hom needed + steam :-mH Ame.-vam.oea- Ape.-vem.:mfi- .z .Homae seemed + steam m-mH Amm.oveo.: Abm.ovma.: Z\ooawz.H0mqe summed + steam m-mH Amm.-vem.- Amm.-vmm.- 2mm .Hem seeded +scessm H-mH Aom.-vm~.m- Ame.-vea.m- z\OOsz .Homqe senate m-o Aoe.ovom.ema Ame.ove:.wmm zam.z .Homce gamete m-e Ame.ovmo.a Aam.oveo.a Rzm .HOm gamete H-o mm.om HH.0NH endemcoo :m.m mo.o om.m mm.o mmm was mm m - AOH.ovmm.e zo&.z .Hom twee: pom :-HH Amm.ovom.o Amm.ovmm.o z\oonz .HOmce teem: pom m-HH - Aofi.ovom.o 2mg .Hom teem: pom H-HH - Ana.-vem.m- see .2 .HOm senate :-e flew.-vmo.m- Ame.-v:m.m- 2\00sz .HOmee esteem m-e Ate.oveo.me Aem.ovom.ee saw e.z mflomce gamete m-e - nxda.ovoa.o w2m&.e.H0m cemeom H-e ee.oe m:.em eqepmcoo eme.m e:e.o :e.m 00.0 emmm one mm H ufloaoammmoo pdoaoammmoo .on .on QGOHmmmamom QOflmmmLMmm adopw Coauwddm .AHN H H: mpCQHOHQMmoo Goapmomflm mowompflz o>H> mm.aom mmSHw> mm nmam obww page memansmwoz %aonwaonmq mo mQSOhm m mqflmb hp vopmadoawo mQOHmmommom mamfluadz oomne .dm oHQwB 129 .HwfinoaOHE mossy H Goad . I mx nomoapfla zn .oHQSHomcH H .Homcfla .mpcmfloflmmmoo cOflpmHoppoo Hmflpawm mam mommanoawm CH mozaw>£. .Aoppwe haw u Sew .oHndHom n .Hom.H .Ewhmopa Soapoaoc pop%w Goapwswo oSp pom flame .Ewpmomm Goepmaoo popmw Goapwsvo onp AOM mmp .mpweapmo Mo ponse Upwocmpm n HMmo .mpCoHofimmooo aoflummMHU z o>fi>.mw.£pflz vommoAan Amo.o.AQ v hapcwoflmflcmflm pom ohms p30 Umpoaoo moanwflaw> .Ewpmoaa Coapmaoc momwsvm umwoa mflp Song mcflpadmmn Coapwdqo on» now who Edaoo menu CH Gm>flm mpnofiofimmooo Goammopmomn .mm manta as steam we coeeeseoece eoaeeeee .mmdoam ma ounfl cocHQEoo oaoz mucoEmAdmon %MOpwmoan ”.0G adonww Awesceecoov .zm wanes 130 Equation 3 (Table 2h) is less complicated than equation 1 and 2 requiring only three laboratory measurements (degree of browning, hot water soluble DM and insoluble N, % DM), but this prediction equation still has to be considered the second choice when compared with the equation using only one laboratory measurement — acid detergent lignin (Group 7, Table 23). The partial correlation coefficients shown in Table 2h indicated the relative importance of variables in that equation. Three multiple regression equations with high R2 for in 2113 dry matter digestibility are presented in Table 25. Although all three equations gave extremely 2 high H (.97), they represented only slight improvements as compared to the R2 of .9u obtained by using only two of the pepsin solubility measurements (Group 9, Table 23). Pepsin solubility measurements are apparently important for predicting in give dry matter digestibility since all three equations in Table 25 included this set of measurements. Furthermore, a combination of the variables pepsin solubility and hot water solubility gave the highest R2 for in 3113 DM digestibility (Equation 1, Table 25) and also for in 3112 N digestibility (Equation 1, Table 2h). Therefore, when predicting both these in vivo responses pepsin and hot water measurements should be those of choice. 131 Hoe.ovmm.om HHm.ovmo.o: :mm.z .HomeH ecemeeeoe-eHe¢ m-m Aoz.-voH.H- AHm.-vme.- z\z.HomeH ecemeeeme-eHe< H-w oe.Hm mH.mm eceemcoo me.H 00.0 m>.H 00.0 mam one mm m u Amo.ovdo.a 20R.z .Hom .cwm+chmmm+ Gmsdm :an Abe.ova.Hm AbH.ova.wm zom.z .HomcH .cee+cheee+ steam m-MH Rom.-va.- AjH.-V:m.- z\OOHmz .Hoqu.aee+:Hmeee+ steam m-mH Amm.ovmm.o AHm.ovmm.o 2mm .Hom H.aee+aHmeem+ steam H-MH - HmH.ovow.o z\OOsz .HOmaH meeom m-o Aom.-vmm.0H- “NH.-Voo.~m- zom.z .HOmcH queem m-e - AH:.ovmm.o 2mm .Hom cheom H-e mo.mm om.:m endemcoo Ho.H oo.o em.H No.0 mam one mm m Awe.-vbb.oH- Aem.-ve:.mH- zam.z .Hom teem: pom :-HH Amm.-V~:.- Amm.-VHm.- 2\00sz .HomaH noes: pom m-HH Hew.ovmw.o Azm.ovmm.o sax .How goes: pom H-HH Hoe.ovm:.mH Amm.ov:e.o 2mm .2 .Hom cheem 2-0 - Amo.-vom.- z\OOsz .HOmaH chmom m-o I A%.Ovmm.m SQR mz WHOmGH CHMMQM NIM - .o . . so .HOm ch mm H- mm.m: SA vma.md w Rm pawpmcoo one.H eoe.o oe.H 00.0 emmm one mm H pQGHOHMhooo pQOHOH.H.HmOU .OG .OC GOflmmoamom coammmhwom meopw GOHpmdqm D .ll .AHN n a; szoflOHmmooo Coapmomflm Aoppwz ham o>H> CH pom mozaw> mm anm o>ww pane meoEonSmwoS hLOmeoan mo manopw N mcamb hp Umpdadoawo mcowmmoawom bananas: moans .mm oanwe 132 .Gfipwmhoqwm n .cmg H .:m eHnwe eem x was ”.H.s.w.e.e.e.e.n.w - AHo.-Veo.- z\QOHaz .HomaH cheem m-o Ame.-VOH.0H- Ame.-v:m.:H- zag.z .HOmaH cheem m-o Aem.ovoe.o Azm.ovwb.o sax .Hom nHmmom H-e - Azo.ovmm.H zow.z .Hom enemteeee-eHe¢ m-m pflmaoammooo pcofloammooo .oc .on coammonmom Goammoawom Adomw GOprSdm heeseHecoov .mm eHewa 133 Three multiple regressions having high R2 values for in 1312 N balance are given in Table 26. Equation 1 used variables of acid detergent solubility (Group 8) and pepsin solubility (Group 9) measurements and pro- duced the highest R2 (.88) among all equations. After a least squares deletion of insignificant (PJ>.OS) variables the R2 was reduced only by .06 units for Equation 1. No reduction in the R2 value was observed for Equation 2 after the least square deletion program. However, Equation 3 (after deletion) is the simplest equation with respect to the laboratory analyses required (pepsin solubility measurements and rumen ammonia re- lease). Since pepsin solubility measurements would also be required to predict in 3112 dry matter and N diges- tibility (Equation 1 of Table 2h and 25), only one addi- tional laboratory analysis would be required to predict N balance (in vitae rumen ammonia release). (Two multiple regressions having relatively high R2 for in XEXQ nitrogen retention as a percent of absorbed nitrogen are given in Table 27. Maximum R2 for single group measurements was .u9 (Group 12, Table 23) and the use of two group gave an increased R2 of .75 (Equation 1, Table 27). Even the improved R2 is still unsatisfactory from a practical predictive standpoint and some of the analytical measurements required in the equation are 13h .:m oHnsa eon H was H.n.w.m.e.n.o.n.d Amw.ovoe.m Aom.ovmm.e sax.z .Hen dHneem :-e - Aeo.ovem.o 2\00sz .Hode dHneem m-o - Amo.-VHo.ev zom.z .HondH dHneom m-o - Aeo.ovmo.o zen .H m dHneem H-e Amm.-VHe.- Am:.-veo.- 2\00Hx z- mz cease m mm.w- H>.mH- edsendeo em.H mm.o m:.H mm.o men and mm m - Aeooo.vmo.o sma.z .HdmsH sees: pom :-HH - AHo.oVHo.o 2\OOHaz .HondH need: new m-HH Ame.ovm:.o Aoe.ovn:.o sax .Hon need: new H-HH Ase.ovoo.0H Ace.oVHm.0H 2mm .2 .Hdn edomsoeee-eHe< m-m Ase.-vHo.:m- Ann.-vmm.:m- 2m .2 .HOndH snowsoeee-eHe¢ m-m Ame.ovdm.m nee.ovem.m Z\00sz .HdnnH edewsoeee-eHe< H-m om.H:- m:.m:- ecsendoo mm.H mm.o ::.H mw.o new use mm m - Aoo.ovmm.o 2\00Hsz .HonnH dHnmom m-o Ame.-vnm.HH- AoH.-von.wH- sea .2 .HondH dendm m-o - Amm.ovom.o see .Hon dHnmem H-o Aom.ovH:.m Am:.ova.HH 26w .2 ”Hen edemseeee-eHe< m-w - Adm.-vme.em- mama .z .HoncH edemseeen-eHe« m-w AHn.ovmm.o snmm.ovmn.H Hz\OOan mHOndH edemseeee-eHes H-m me.wH- mm.o:- . . essencoo e::.H emm.o wN.H mw.o omen end mm H pflmflOHh.H®OO pflmflOHMMQOO .02 .0Q QQOHmmoaMmm coammommom wgbomc GoaprUE .AHNHGV mofldamm Gomoppfiz o>H> WW mom modaw> mm nmfim obwo pane meoEoMSmwoS haouwhoan mo mmsohw m mmflmb hp wmpwHSOHwo mcoflmmoawom oHQHpHSE moans .0N oHQmB 13S .cfipwmaocwa n .quH .:m ersa men a and n.H.a.m.e.e.e.e.p.s Amn.ovew.neH AHe.ov:e.m:m 29m.z .HdndH cheee + cease m-mH not.-vse.m- Ame.-ven.n- 2\oonz .HdndH dedod + cease m-mH Aen.ovo:.H Ann.ovm:.H zen .Hdn dHneee + dream H-mH - Amm.-va.mH- 2mm.z .Hon edomeoeee-eHe< m-m non.-v::.mom- Ame.-vme.eom- zma.z .HOndH edomseeee-eHe< m-m Amr.ovem.OH Ame.ovom.o 2\z .HdndH edemseeoe-eHe« H-m 0:.nm- om.mH- essenqoo mo.e oe.o . 30.0 mn.o new new mm m “do.-va.mo- Ame.-va.mo- zm&.z .Hon .ssd+dHnnnm+qmszm :-MH Ame.ov:e.Hmm Ame.ovmm.mmm saw.z .HondH .dsn+cHneoe+dessm m-MH no».-vmm.MH- AHe.-Vme.MH- 2\00sz HHOndH .qse+cHneee+deesm m-mH Ame.-voo.m- Awe.-vo:.m- mEQK.WH0n qum+chdma+ steam H-mH - snow.-vsm.o H2\00sz-mmzadeesm m mm.:mm oa.mom . pcwumcoo eem.m eme.o mm.m mn.o emmm eds mm H uCoHoammmoo pdoaoammmoo . .on .ofl nQonmoamom moammopmom wmdoaw GoapmSUm .bwmnnv Gmwoppflz UQQAOmQ< wo unmoaom < mm GOHpcopom Gomoapflz o>H> :H 90% m Qmfim haobfipwamm m>wu pane meoEoQSmwoz %LOpwaonmq mo manopo m mo mmmnwflpw> mcflmb he UmpwHSoawo mGOHmmmhwom mamflpasz 039 .Nm manna 136 rather time-consuming and complicated. Two multiple regressions gave relatively high R2 values for maximum dry matter intake and are in Table 28. Both equations included in XEEEQ rumen microbial +pepsin solubility measurements, particularly, the so- luble dry matter fraction. This may suggest that dry matter solubility in the rumen is related to the forage dry matter intake. In other words, forage voluntary intake of ruminants is probably somewhat regulated by the physical capacity of the rumen. This was in agree- ment with the concept of Crampton (1957) and Weston (1966, 1967 and 1968) that voluntary intake of forage dry matter by ruminants was limited primarily by rate of cellulose and hemicellulose digestion in the rumen. Species differences with regard to forage voluntary intake has been reported (Conrad, 1966), thus, the pre- diction equations developed from the present study pro- bably should be used with caution for other ruminant species. VIII. Multiple Regressions of Five In Vivo Parameters Using Selected Variables That Produced High R2 Values Fourteen combinations of selected variables obtained from all 13 laboratory measurements were used in a multiple regression least square deletion analysis. .em stse sss a eds .n.H.d.m.d.s.e.e.d.s 137 - Adm.-vem.m- saw.z .Hdm stdsd+dsadm :-mH - Amm.oveo.: 2mm.z .HosdH stdsd+dsddm m-mH - Amm.-vem.- Z\00Hdz .HOsdH stdsd+dsddm m-mH Ade.ova.o Amm.ov0H.o an. .Hos stdsd+dsddm H-mH - AmH.ovmm.H zms.z .HdsdH does: pom m-HH Aem.-vmo.- AoH.-veo.- 2\00Hdz .HOsdH does: pom m-HH - Aeo.ovHo.o mam .Hds sees: pom H-HH Hm.m- oe.e edsesdoe om.o H:.o me.o mmm eds mm m - Aem.ova.m zaw.z .HdsdH dHndsd+ dsadm m-mH nee.-vmo.- Adm.-vmm.- 2\ooHdz .HdndH stdsd+ dssdm m-mH AHe.ova.o Amw.oV:H.o and .Hos stdsd+sdsadm H-mH - Ado.-veH.- 2m& .2 .HOn edswdspse-eHed m-m - AHm.-vmo.:- wzmo.z .HmsdH edsmdsese-eHed m-m Ame.ova.o dAmm.ovem.o 2\00erZmH0sdH edsmdsese-eHed H-m mme- eH.m.. . . edsesdoe sem.o eoe.o em.o me.o emmm eds mm H edsHeHudsoe pdsHeHensoe .od .od GoemmoAMmm Qoammopmom dadoao Goepwdvm Q .AHN n av oxprH poppwz ham SSSHKmE mom modaw> mm swam ham>flpwaom obww page mquEoLSmwoz hLOpmaonmq Mo madcao m mme b ha Umpmadoawo mcoemmmawom bananas: 039 .mm manna 138 Selection of combinations were generally those with greatest r values. However, factors such as time, techniques and instruments, involved in practical opera- tions and inter-correlations were also considered. These 1h combinations gave 70 multiple regression equations but only a few representative equations will be presented and discussed. Three multiple regression equations giving high R2 values for in 1222 ND are given in Table 29. An extremely high R2 of .99 was noted for Equation 1 which used seven variables from six analytical groups. The standard error estimate was also smaller for Equation 1 than for the other two equations. Partial correlation coefficients of Equation 1 indicated that crude protein (Group 1) and rumen microbial + pepsin soluble DM (Group 12-1) were negatively correlated with in XEXE nitrogen digestibility. Opposite relationships were found for these variables in simple correlation analyses. These complete complete changes in relationships suggested the existence of interrelationships among variables that were not revealed in simple correlation analyses. Par— tial correlation coefficients of Equation 1 also in- dicated that variables obtained from chemical, pepsin and in 33:39 rumen microbial + pepsin digestions were all highly correlated with in zizg ND although pepsin solubility measurements had slightly lower partial 139 Table 29. Multiple Regressions Calculated by Using2 Selected Variables That Produced High R for £2 Vivo Nitrogen Digestion Coefficients (n=21). Equation Groupa Regression no. no. Coefficients l R2 and SEEb 0.99 1.38 Constant c 259.06 d 1 Crude protein,%DM -3.95(-.76) 6 Acid-detergent fiber,%DM -.52(-.82) 7 Acid-detergent lignin,%DMf -2.70(-.85) 8—1 Acid-detergent insol? N/N -.99(-.86) 9-2 Pepsin insol. N,%DM 120.oo(o.75) 9-3 Pepsin insol. N 00/N -h.08(-.76) 12-1 Rumen8+pepsin sol. DM,% -.78(-.87) 2 R2 and SEE 0.98 1.71 Constant 38.30 3 Rumen NH3-leOO/N 1.72(O.80) 9-1 Pepsin sol. %DM o.51(o.82) 9-3 Pepsin insol. leOO/N -.88(-.95) 11—3 Hot water insol. N,%DM 7.91(o.89) 3 32 and SE 0.97 1.97 Constant 178.91 6 Acid-detergent fiber,%DM -.h6(-.68) 7 Acid-detergent lignin, %DM -3.22(—.87) 8-1 Acid—detergent insol.N/N -l.02(-.79) 12—1 Rumen+pepsin sol. %DM -.81(-.75) aGroup no.: Laboratory measurements were combined into 13 groups, detailed information is given in Table 23. bSEE = standard error of estimate. CDM = dry matter. d coefficients. einsol. = insoluble. f _ h N - nitrogen. sol.= Values in parentheses are partial correlation gRumen microbial. soluble. 1&0 correlation coefficients. Equation 2 required measure— ments from only three analytical groups (Group 3, rumen NH3—N x 100/N; Group 9, pepsin soluble DM plus pepsin insoluble N x 100/N; and Group 11, hot water insoluble N, % DM), yet had a very high R2 of .98. Equation 3 used variables from four analytical groups and these variables result from common analysis presently per- formed in many laboratories. For example, acid deter- gent fiber, acid detergent lignin and in £3333 rumen microbial + pepsin soluble dry matter have been routine analyses for forage samples in many university labora- tories for many years. Recently, the analysis of acid detergent insoluble N x 100/N has also become a routine analyses. Thus, Equation 3 could be used for practical predictive purposes. Partial correlation coefficients of Equation 3 indicated that microbial + pepsin soluble dry matter was again negatively correlated with in zizg ND. Three multiple regression equations producing high R2 values for in zizg_dry matter digestion coeffi- cients are given in Table 30. Multiple regressions computed by using selected variables slightly improved the R2 and standard error of estimate (SEE) compared with the values obtained by using variables from only two 2 groups of measurements (R = .98, SEE = 1.07, Table 30 vs. R2 = .96, SEE = 1.61, Table 25). Equations listed 1&1 Table 30. Multiple Regressions Calculated by Using2 Selected Variables That Produced High R for 'In Vivo Dry Matter Digestibility (n = 21), Equation Groupa Regression no. no. Coefficients 1 R2 and SEEb 0.98 1.25 Constant c 6u.30 5 Cell wall constituents,%DM -.36(-.72) 8-1 Acid-detergent insol? N/Nf 0.h8(0.62) 9-1 Pepsin so1.8 %DM 0.3o(0.59) 9—3 Pepsin insol. leOO/N -.51(-.87) 2 R2 and SEE 0.98 1.07 Constant 51.15 8 Degree of browning(0Duu0) 93.h8(0.82) 9-1 Pepsin sol. %DM 0.37(0.71) 9-3 Pepsin insol. leOO/N -.6h(-.78) 12-3 Rumenh+pepsin insol. N,%DM -2l.36(-.81) 3 R2 and SEE 0.98 1.26 Constant 85.01 5 Cell wall constituents, DM -.68(-.9h) 7 Acid-detergent lignin, DM 1.09(0.52) 8-1 Acid-detergent insol. leOO/N 0.55(O.66) 9-3 Pepsin insol. leOO/N -.76(-.90) 11-3 Hot water insol. N,%DM 2.81(0.60) d a,b,c,d,e and f see Table 29. 8 sol. soluble. hRumen rumen microbial. 1M2 in Table 30 have the same R2 value but Equation 2 is the least complicated in terms of laboratory operations, and the standard error of estimate is least for this equation. Acid detergent insoluble N as a percent of total N (Group 8-1) was used in both equation 1 and 3. In both equations a positive relationship between this variable and the in 3239 dry matter digestion coeffi- cient was indicated by the partial correlation coeffi- cients. This finding is definitely contradictory to the idea that AD insoluble N is highly negatively co- rrelated with £2.21X2 dry matter or energy digestibility (Goering gt 31. 1972). However, one also Should be reminded that the samples used in the present study contained less AD insoluble N than in the Goering 33 ‘31. (1972) study. Two multiple regression equations calculated by using selected laboratory measurements that produced relatively high R2 for in 1312 nitrogen balance are in Table 31. Variables used in these two equations are practically the same except that Equation 1 includes the additional variable, crude protein content. The R2 for multiple regression equations in Table 31 were markedly above those for Simple regressions obtained from Table 17 (.92 vs. .72) and the standard errors of estimate were also improved (1.0h vs. 1.59). More 143 Table 31. Multiple Regressions Calculated by Using2 Selected Variables That Produced High R for .In Vivo Nitrogen Balance (n = 21), Equation Groupa Regression no. no. Coefficients 2 b l R and SEE 0.92 1.0h Constant c —32.19 d 1 Crude protein, %DM l.38(0.76) 8-1 Acid-detergent insol? N§N 2.2u(0.72) 8-2 Acid-detergent insol. N,%DM -6u.29(-.67) 9-3 Pepsin insol. leOO/N -.23(-.67) ll-l Hot water 301% %DM 0.33(0.71) 2 R2 and SEE 0.91 1.03 Constant -3l.h9 8—1 Acid-detergent insol. N/N l.76(0.77) 8-2 Acid-detergent insol. N,%DM -u7.05(-.77) 9-u Pepsin sol. N,%DM 8.59(0.85) ll-l Hot water sol. %DM . 0.31(0.o9) a . Group no.: Laboratory measurements were combined into 13 groups, detailed information is given in Table 23. bSEE = standard error of estimate. 0DM = dry matter. dValues in parentheses are partial correlation coefficients. einsol. = insoluble. N = nitrogen. gsol. = soluble. 1AM analytical values will be required for multiple regre— ssions than for simple regressions. Nevertheless, N balance is a sensitive measure for differences in value of forage nitrogen for animal productivity and main- tenance (Chalmers, 1961) and animal producers and.re- search personel may more seriously consider greater emphasis on estimating the term N balance in order to more properly evaluate feeds for most efficient pro- duction. Since multiple regressions calculated using selected laboratory measurements did not result in any Significant improvement in R2 for in zigg N retention as a percent of absorbed N and maximum dry matter in- take as compared to those R2 obtained by using two groups of measurements (Tables 27 and 28), equations will not be presented and discussed. IX. Comparisons of Multiple Regressions Developed From Different Sources of Forage Samples for Estimating In Vivo Nitrogen Digestion Coefficients _Multiple regressions by the stepwise least square deletion analyses were calculated for estimating 'ln vivo ND using data from uu.samples from Dr. H.K. Goering of USDA, Maryland plus 18 samples from the Michigan State University plus some samples from Uni- versity of Wisconsin and Minnesota. Variables used in the regressions were mainly acid detergent solubility 1&5 measurements. Two representative equations are given in Table 32 along with equations using the same variables on 21 samples of the present study. When variables of crude protein, acid detergent insoluble N x 100/N, and acid detergent insoluble N, % DM remained in the re- gression analysis, a R2 of .89 was found for the com- bined samples and none of the variables were deleted in the least square deletion analysis. However, none of the variables had a high partial correlation coeffi- cient. This suggested that these variables were not highly nor directly related to the igngzg ND. 0n the other hand a Slightly lower R2 (.86) was observed for the regression based on 21 samples from the present study and two of the variables (crude protein and acid detergent insoluble N, % DM) were deleted by the analysis due to insignificant (P>'.05) regression coefficients. The standard error of estimate was somewhat smaller for the equation using MSU samples than for the equation using the combined samples. Consequently, the regre- ssion coefficients for the two resulting equations are markedly different. These inconsistencies may be due to sample size, species, type of preservations, geo- graphical locations, variation in laboratory analytical techniques and digestion trial procedures. Regression analyses of Equation 2 also indicated discrepancies between the two equations in that lignin 186 .popqu haw n Ema .cbwoapHc n z: .eHnsHomnH u .HOmGHw .mpceHoemmooo Goepwaohnoo Hewpaem can memonpcoaam Ga modae> .Aoooa .owaaoe was Hoopmvmqoap umswo camepads one ma wows meanness» one Aegean oHQEwm pom 60pm3n6fidb .SOmaon .U.Q .mm was Gammoomez mo hpemao>ficb .comcewAOb .¢.z .AQ u mnonpoo .edsHedsz .enmp .wdeooe .m.m .dm u «an: .aeasdoeHde oesdm dsdeon n mes M U H Ame.-VNH.m- - an .demHH edomdoeoe-eHo< “em.-vmo.em- AHe.-v:e.e- an .2 .HosdH edomdopoe-eHo< a A>>.ovmw.a Geopoam 06590 0H.moa o~.o: pneumcoo No.0 oe.o mm eoesdnee om.m No.0 mw.© o®.o mmm use mm Hm on d m - Hem.-ved.:m- 2mm .z .HondH odomdseoe-eHod Ame.-va.m- fies.-VH:.- z OOdez wHosdH edomdsose-eHo< - unmm.ovmo.H dHoeodd oedde 0N.Ho ma.oo pcdpmcoo mm.e om.o smm eoesdHee no.: ee.o m:.e oe.o emmm eds mm Hm on d H sz omaenpo +Qom pcoeoemmooo QOHpmomHQ cemoa»ez o>H> mm pom mcoemmoawom onHpHSS mo mGOmHLmQEoo .mm canes “+7 was deleted (P>’.05) from the combined samples but not the MSU samples. However, in both equations and both groups of samples the term AD-N as percent of N or DM had a important relationship to in 1119 nitrogen di- gestibility. SUMMARY AND CONCLUSIONS Part 1 The predictability of five animal responses was evaluated using data from 31 laboratory measurements on the forage fed. The five responses were nitrogen and dry matter digestibilities, nitrogen balance and reten- tion, maximum dry matter intake. These data were obtained from sheep fed preserved forages at Michigan State University (2u.samples) and at other experimental stations (66 samples). A large proportion of samples were heat damaged. Data from three groups of labora- tory measurements were collected for most samples. These three analytical groups were chemical determina- tions, enzymatic incubations, and sequential rumen microbial plus enzymatic incubations. Among these 31 measurements, acid detergent (AD) insoluble nitrogen (N) as a percent of total N, AD in- soluble dry matter (ADF), pepsin soluble N as a per- cent of dry matter, hot water soluble dry matter and rumen microbial plus pepsin soluble dry matter were the best single predictors (had the highest correlation coefficient(r) and the lowest standard error of estimate) 1H8 1M9 for in £112 N or dry matter digestibilities, N balance, N retention as a percent of absorbed N, and maximum dry matter intake. However, the r values for the latter three parameters were lower than that for the former two (r = .80 vs. .93). Forage N content was insigni- ficantly (P 9%) presumably due to heat damage. When the 31 laboratory measurements were combined into 13 groups based on analytical schemes and used as multiple predictors for lg £113 responses, those mea- surements obtained with pepsin incubations had the 150 greatest precision. Some improvements were found in predicting animal responses when variables of the two groups were jointly used. Finally, multiple regressions with extremely 2) .98) were obtained for high degree of precision (R ig_zlgg N and dry matter digestibilities using selected variables (no restrictions as to analytical groups). Regression coefficients along with partial cOrre- lation coefficients were useful not only in defining relative importance of variables in predicting animal responses but also in revealing the true (unconfounded) relationships among variables and animal responses. Large differences were observed for variables obtained from different sources of samples with regard to the relative importance in predicting lg vivo N digestibility. 151 Part 2. Haylage Preservation With Various Chemicals I. Effect of Chemicals on Alfalfa Haylage Temperatures (A) Effect of Chemicals on Haylage Temperatures During Storage During the entire h2-day storage period, each 3110 was equiped with four thermocouples at four different levels. Temperature was measured ten times per day in an attempt to discover the diurnal pattern of haylage temperature. Representative hourly temperature of haylages and ambient temperature during the storage period are given in Table 33. According to the ten hourly readings of this study, the lowest and the high- est ambient temperatures were generally observed near daily hours 3 or u and hours 18 to 20, while no con- sistant changes were observed in hourly haylage tem- peratures. For example, on June 30, haylage tempera- tures of silo 3,6 and 7 (treatments of .8% propionic acid, control and .5% AIB) were Slightly higher during hour 1h to 18 than during hour 0 to 6 but this trend was not clearly observed for the other three silos. Haylage temperatures were generally related to the ambient temperature on July 10, but no clear re- lationships were observed on July 15 to 20. These in- consistant findings indicated that no clear diurnal pattern was found in haylage temperatures. Thus, ten 152 0m m.33 e.m3 3.03 3.0m m.m3 m.em 0 am 0.m3 m.33 H.H3 m.am a.H3 5.0m 0 am 0.m3 3.33 0.03 m.em e.H3 a.em 3 mm 0.03 o.m3 0.03 m.am e.H3 p.0m m 3m m.33 3.m3 0.m3 H.0m e.H3 n.0m 0 am 0.m3 0.m3 m.0m o.mm H.H3 H.0m 0m em 0.m3 m.m3 m.m3 e.mm H.H3 H.em 0H 0H o.m3 e.m3 m.m3 o.mm H.H3 H.0m 0H 0H 0.m3 0.m3 m.m3 e.em e.H3 5.0m 3H om 0.m3 0.m3 H.H3 m.0m m.m3 3.0m mH mxe mm o.m3 m.0m a.H3 0.mm m.m3 0.mm 0 0m 0.03 m.0m m.0m e.mm 0.03 0.0m e Hm 3.33 0.03 m.em m.mm H.H3 3.0m 3 0H 3.33 5.0m m.0m m.mm H.H3 3.0m m 0H 3.33 0.03 m.0m m.mm H.H3 e.0m 0 0H 3.33 0.03 e.H3 m.mm H.H3 0.wm 0m mH 3.33 0.03 e.H3 m.mm 0.03 m.mm 0H mH 0.03 0.03 3.0m S.Hm 0.03 0.mm 0H mH m.m3 0.03 3.0m 0.00 0.03 0.mm 3H mH m.m3 0.0m 3.0m 0.0m 0.03 m.mm NH 0 3 m a m e0 0mxe oomem+ LmH 7am.0 nww0 e3.0 edoHodd Hodedoe Aam.0vas dee eHos oHdonode Hmwmmv amnesn oaflm use momwawwa mo mpGoEpmoae one open )1'1 l1 .AmeoH .m deemed on 3m oddevseoHdom oMmLOpm one mnHaSQ onmoadpoaoQEoe pmoens< one owwahmm handom popooaom .mm manna 153 mm 0.00 m.00 0.m0 H.H0 0.00 0.00 0 H0 0.00 0.00 0.0m H.H0 0.00 0.00 0 H0 0.00 0.00 .mm H.H0 0.00 0.00 3 00 0.00 m.00 .0N H.H0 0.00 0.00 m 0H 0.00 3.00 0.H0 0.H0 0.00 0.00 0 3H 0.00 3.00 H.H0 H.H0 0.m0 0.00 om NH 0.00 0.03 0.00 H.H0 0.00 0.00 0H NH 0.00 0.m3 0.0m H.H0 0.m0 0.00 0H 0H 0.00 0.03 0.0m H.H0 0.00 0.00 3H 0H 0.00 0.03 0.H0 H.H0 0.m0 0.00 mH mH\0 0m 0.03 0.03 0.03 3.03 H.H3 0.00 0 am 0.H3 m.33 0.00 H.H3 3.00 0.00 0 0m H.H3 3.03 0.03 H.H3 3.00 0.H0 3 3m 0.03 0.m3 m.00 m.00 H.00 H.H0 0 mm 0.00 0.m3 0.00 0.00 H.00 0.H0 0 0H 0.00 0.H3 0.00 0.03 H.00 0.H0 0m 0H 0.00 0.H3 0.00 H.03 H.00 0.00 0H 0H 0.00 0.03 0.00 0.00 0.00 0.00 0H 30 0.03 0.03 0.03 0.03 0.03 0.00 3H 00 0.m3 H.03 H.H3 0.00 m.00 0.00 NH 0 3 m 0 0 0 0H\0 0000 + 0H 00.0 00.0 R3H0 ederde Hodedoe Ham.00mH< de eHos oHdonodd Hmwmmv Aeneas 0000 was momwahwn mo mucoepmoae can ease AeoddHedoev .00 oHese 15a .aoQESG oaemp .AdodeHos 000 s do amm.H- 0 oeadoeHsddod u 0000o .opwahHSQOmH EdHcoeaw n mHoH ASO% Seam ecoaoooa moAdpwaomEop name one mucomoamoa ensueAOQEop nowmo .nonssa oaem .GOHpSHom 00m 0 Mo me.a : oehnoeawshomw 0 0 m0 000 H0H 00m 00 30H 0HH ommmm nooawom 0 m0 H3 00 00 mm H3 00 “MMMs mm 03 00 03 00 00 00 dsoz H0. 00 03 00 00 30 30 00 de0 .. a a a e H H s a... Hm m 0 d ms. m3 H30 mm Mm mm .03 H0 edm mm H3 0 00 em 00 H3 03 edm H0 00 03 00 H3 30 H3 o00 psH ................................... o --------------------------------- Assozv eoHdom A00 H30 Hmv A00 A00 H00 edoHddd 0mm: Hodedoe somem 0H .m.0 00.0 03.0 use: + 0WH¢ .dm 00909 160 nitrogen digestion coefficients when the haylage tem- peratures were calculated as "degree days above 35 C" than as mean storage temperature. Degree days above 35 C were calculated by summation of daily haylage tempera~ tures which were above 35 C during storage. These temperatures were calculated for the Six haylages and are in the bottom line of Table 3h. Although the ranking of these degree-days above 35 C was the same as that for the mean haylage temperatures i.e. control>'1% AIB treated?>.5% AIB + formaldehyde (1.25% of a 37% solution) treated> .8% or .1198 propionate treated> .5% AIB treated haylages, the magnitude of difference in temperature was much greater for the expression of degree-days above 35 C than that of mean temperatures. For example, the mean storage temperature of control haylage was only 1.08 fold higher than that of haylage treated with .5% AIB while the former was h.88 fold higher in degree-days above 35 C than the latter. Weekly mean temperatures were also calculated on the position of the thermocouple in the silos (Table 35). Five thermocouples were placed in each silo. Their distribution was two near the bottom; one in the lower- middle; one in the upper middle; and one more than 1.33 m below the top of the silo. Both the overall mean and maximum temperatures were positively related to thermo- couple position in the Silo. Haylage temperature was 161 Table 35. Mean Weekly Temperatures of Haylages Ensiled in Concrete Silos for A h2-Day Storage Period. Temperatures Are Presented Based on the Level or Position of Thermocouple in the Silo. Position of Thermocouple Bottoma Middle-1b Middle-2C Top Period ------------------------- c .................. (Week) . 1st 36e 36f 38 AS - 39 2nd ‘3? 3g 02 E8 1:: 3rd 3 ‘3_ 9 2 nth 32 33 11% 1.19 39 5th 30 32 39 11.8 37 6th 29 30 37 A9 36 Mean 33 31; 1L0 ’45 Degree- days above 21 A2 217 sac 35 C ' aApproximately 1.2 m above ground. bApproximately 3 m above ground. cApproximately 5 m above ground. dApproximately 7 m above ground. eValues underlined represent maximum temperatures. fEach temperature represents the mean temperatures recorded ten times per day and seven days per week. 162 always the greatest at the upper—most thermocouple. The top area had 26 times greater sum for degree-days above 35 C than did the bottom portion and even the upper middle are (> 3 m. below the surface) had 10 times great- er sum for degree—days above 35 C than did the bottom portion. This relationship was probably caused by the degree of compaction and of exposure of haylage mass to the air. Gordon 33 El° (1961) did not observe this type of relationship between the position in the silo and haylage temperature when the top was covered by plastic Sheets in similar Silo structures. Temperature in bottom area of silo remained the same during the first 3 weeks while maximum temperature was attained during week 3 for other Silo heights. (B) Effect of Chemicals on Haylage Surface Temperature After Silos Were Open for Feeding After the silos were opened for feeding, haylage temperatures were measured by mercury (Hg) therometer inserted 25 cm below the surface about two times per week. Mean and maximum haylage surface temperatures measured during the feeding period of A9 days are pre- sented in Table 36. Maximum ambient temperature as well as haylage surface temperatures were observed during the first 10 days of the feeding period (July 11 to 23). With the exception of the .h% propionic acid treatment, the other four treatments reduced mean and maximum 163 .doHedHos 000 s do 0mm.H - seedoeHsddoe n .Aonesc 000mm 30 am as Z 03 3 mm sdfimdammmse 0m 03 H3 00 03 00 33 oddemmwmaoa IIIIIIIIIIIIIIIIIIIIIIIIIIII U IIIIIIIIIIIIIIIIIIIIIIIIIIII H00 H30 wmv .va H00 sH00 edoHddd Hodedoe d0mem adew oesa Ham 00.0 300 Ham.000Hs How HCOH a nomH Edecoesd U. o. . o m .nomo one: moaem doflz when 00 we UoHaom < mneASQ momwahem mo mommaSm one soaom Se mm UoASmwoS moasewaomeoe SSSeNdS one See: .om oHQwB 168 surface temperatures by approximately 5 and 10 C respec- tively. Treatment with .h% propionic acid had practically no effect on surface temperature of the resulting hay- lage. Contrary to the findings described in the pre- vious section, chemicals applied at higher rates generally had increased effectiveness in reducing surface tempera- tures. No sound explanations are evident. (C) Effect of Chemicals on Haylage Temperature During Refermentation Approximately 30 kg of haylage taken from the lower middle portion of each Silo were loosely packed into 55 gal. barrels. Temperatures during a 50-day refermentation period were measured about three times per week by thermocouple and are shown in Figure 6. A sharp increase in temperature was found in control hay- lage on the second day of the refermentation period. Maximum temperature (59 C) was reached by day three then haylage temperature declined rapidly. A second peak was observed by about day nine (Sept. 23), but the second maximum temperature was only about 37 C. Haylages treated with .8% propionic acid and .5% AIB + 1.25% formaldehyde did not develop a high temperature until day Sixteen with a maximum temperature of 57 C attained on day eighteen. Temperatures then declined gradually. Treat- ment with .5% or 1% AIB was able to retard refermentation until day twenty when the rate of temperature increase 165 .eHod oHdoHdodm 03.0 .opwnhpSQOmH SSHGoEEd &m.o .Hoppcoo .opwahpSDOmH Edecoss< RH .oohflooawsaom RmN.H + onwahHSQOmH SSHCoEE< Rm.o .eHod oHdoHdodd 00.0 II II II II aniUvor~q1 .mfiwoefioflo mSO0as> QpHB newness omwahwm mo deepepnosaomom mceazm mpdoEQoaoon endpwaoQEoB .0 madmem 65 166 a? 13 . (0 )allniVUBdWSL 45 40 30 25 15 10 DAYS Figure 6. 167 was rather slow. About 10 days elapsed between initial increase and maximum temperature (about M3 C) after which temperature declined gradually. The most striking finding was that haylage treated with .8% propionic acid did not develop any significant amount of heat (i.e. above 35 C) for the entire 50 days of refermentation. In fact, this haylage was all good at day 50 when the hay- lage drums were emptied and separated into good and moldy portions. All other haylages were completely moldy when emptied at this time. Results obtained from this experiment were not in complete agreeable with observa- tions reported in the previous two sections where treat- ment with .h% propionic acid ranked inferior to .5% AIB treatment in reducing haylage temperature during storage and surface temperature during emptying time. Samples used in the refermentation experiment represent only one area of the silo and thus, mighy not be representive. In any event, results from all these measurements suggested that the AIB solutions (.5 and 1%) were su- perior to propionic acid in reduction of haylage tem— perature. An opposite ranking has been reported by Goering and Gordon (1973) when both chemicals were evaluated under laboratory and pilot conditions. 168 II. Effect of Chemicals on Haylage Dry Matter Losses During Storagp Amount of top spoilage for the six haylages is shown in Table 37. With no addition of chemicals top spoilage amounted to h.u%. This amount of spoilage is normal when low moisture crops are ensiled (Gordon 33. El. 1965). No significant improvement in reduction of top spoilage was found for the two AIB and the .h% pro- pionic acid treatments and only Slightly improvement for the AIB + formaldehyde treatment. The most effective treatment in reducing top spoilage was .8% propionic acid. This treatment reduced the extent of top spoilage to about 1.8%. Zero top spilage has previously been re- ported for this treatment (Yu _e_t_ _a_l. 1973). Silos with the greatest amount of top spoilage were not those with the greatest temperature. In fact, extremely low correlation coefficients were found be- tween extent of top spilage and temperature for upper Silo thermocouple (r = -.13) or days above 35 C for the entire silo (r = -.u3). However, top spoilage was some- what but not Significantly (P) .05) correlated with de- gree—days above 35 C for the upper thermocouple (r = .61). Extent of top spoilage was positively but not highly related to measured temperature during storage. The term "other spoilage" in Table 37 implies amount of spoilage other than top portion. Haylage 169 .0000 8000 0&0» 0&0000 no 0000009 00 0000 on 050 005000 203p a0waw0 00 0000» 00:0 20 0000 000000m Show 009 .0w000090 000 R u 0M000000 nonpo 0 u 000 0&00003 000w R u 000 H 00m 00 #000 A0ppws 000 00000000 .000008 909 n Ego .doHedHos 000 s 00 00N.H- oehdoeHsadoa0 .L0QESG 000m0 00 9m 0m :3 mm 0m 00000000 SD 00 0000 0300000 00 00 N0 03 H0 00 Heodoes 20 00 eom swsHhsm eooe 3H.0 030.0 H0H.0 300.0 000.0 00H.0 A20 000 esm omsHasm eooe HH.0 00.H 00.H 00.3 00.0 00.H Aeodoes 20 00 owsHHod0 dsdeo 00m 03m 000 000 00m 000 A20 000 smsHHod0 dodeo 03.3 00.0 03.0 00.3 00.H 00.3 Aesdoes 20 00 omsHHod0 doe 000 000 H00 300 000 000 Head 000 owsHHod0 doe H00 H3w va wmv H00 sA00 00 0 o . .0 s . o . Hododoe 0 + AmHevsesdmedp 00 0 0.0 ARm.ova< 1000 ES0QOEE< 000< 00G00 0&9 .00m00hwm 0000009 0mm 0000Goo 90 0000m 09000Goo G0 000000 0mmaOpm hmmumd 00000 0000omm 000 0000>oe0m 000002 909 00 undead .0m 00309 170 treated with .5% AIB had a greater portion of other spoilage than did control (h.02 vs. 2.11%) while the other four treated haylages had comparable or slightly smaller portions of other spoilage than did control. Amount of "good" haylage fed is given in Table 37 how- ever the value does not represent the exact amount of good haylage since haylage taken from silos after the feeding trial was not accurately recorded. Similarly, the value for percent of dry matter lost as gas which is calculated by difference is given in the bottom of Table 37 but only for approximation purposes. Furthermore, the amount of dry matter lost as gas is not necessarily related to the amount of dry matter lost due to t0p spoilage (Gordon 22 2l° 1965; Yu.§£.ag, 1973). Under normal haylage processes, the extent of dry matter lost as gas ranges from 1 to 12% (Gordon 32 El' 1963, 1965). When haylage was treated with .8% propionic acid, Yu gt El° (1973) observed a 0% gaseous loss. Unfortunately, the effect of chemicals on gaseous loss or dry matter can not be accurately evaluated in the present study. III. Effect of Chemicals on Haylage Characteristics When silos were opened for feeding, haylages were sampled three times per week and composited biweekly. Thus, four composites were made during the entire feed- ing experiment (42 days). These four composites were 171 also considered as representative samples from four different positions in the silo. Average values of several determinations obtained from composites are presented in Table 38. All six haylages had similar pH values of near 5 which is a typical value for low moisture silages (Gordon 32 2l° 1963, 1965; Thomas 33 21' 1969; Owen and Senel, 1963). Concentrations of volatile fatty acids (VFA) in these composites were rather low but were in the usual range found for low moisture silages (Gordon gt El' 1963, 1965). The propionic acid and AIB treated haylages contained appreciably greater amounts of these respective acids than did control haylage. The application rates of these two chemicals were generally reflected by the relative concentrations in the resulting haylages. Concentrations of butyric acid were extremely low and isovaleric acid was absent in all haylages. This may be considered desirable since high levels of butyric acid have been related to high levels of protein degra— dation in silage and to low silage intake by ruminants (Murdoch, 1966). Some lactic acid was produced in the control haylage (1.8% DM), but little or no lactic acid was found in treated haylages. Lactic acid is normally the major organic acid in silages. Results from the present study indicated that propionic acid and AIB greatly influenced the silage fermentation pattern by 172 .000 N 0000008 000 0w0000 £0000 M\0E00 00000000 00 0m0000 00000 00 0000000 00008000 wv.m000000e h00 0w00h0£ 0000000 M\0w00h0£ 0000000 00 00008000 m 0w00h0£ 0000000 w\0w00h0£ 0000000 00 00008000 my n 00000008000 00 h00>000m .oo0ps0om 000 0 0o 000.0- 00000000eno0 0 .00o0 00000 o00000o> u <0>o .000850 00000 n 0: mm Mm mm on 000000008000 00 0000000m . . . . . . o000000 0 0000 0000o0oo 0 3 00 3 00 0 0 0 0 0 0 0 0 000050\wo mpwsoo 000o0 000.0 030.0 000.0 0 0 0 00cm 000000 000.0 000.0 030.0 000.0 000.0 000.0 0:000o00> 000o0 0 000.0 0 0 0 0 00o0 o00o00> 000.0 000.0 030.0 000.0 000.0 030.0 00o0 0000000 030.0 003.0 000.0 030.0 000.0 000.0 00ow o000psno00 300.0 000.0 000.0 000.0 000.0 000.0 0000 o0co00o00 000.0 000.0 000.0 000.0 030.0 003.0 00o0 o00oo< ASQRV 00000 0000m0o 00.0 00.0 00.3 30.0 00.3 00.3 00 30.00 0.03 0.03 0.00 0.00 0.30 000 000 00 000002 000 000 030 000 000 000 0000 0000000 omom R0 Rm.o 00.0 00.0 n + Am0oo0m 000 000500 00mcdm .00000 0000m0o .mm.0000000 000002 009 .00m00h0m o0 00>00000< .00 00000 173 depressing lactic acid production but having no effect on acetic acid production (Table 38). The antifungal pr0perties of these chemical addi- tives were evaluated based on mold counts. The control haylage contained u.5 x 106 fungi per gram of fresh hay- lage. Marked reductions in total fungi counts were observed for propionic acid and AIB treated haylages. Low levels were not as effective as high levels for re— ducing number of fungi. Although formaldehyde solution is generally used as an antifungal agent, no reduction in mold counts was observed when 1.25% formaldehyde solution was applied with AIB. The major genera of fungi isolated from haylages were Saccharomyces (yeast), Geotrichum, Penicillium, Scopularigpsis and Mucor. Genus Geotrichum was not found in composites of .8% propionic acid treated haylage while other genera were frequently isolated and not related to treatments. Percent recovery of chemicals were calculated based on the concentration of chemical used at ensiling and the concentration in the resulting haylage (Table 38). The percent recovery of propionic acid ranged from 52 to 70% which are somewhat lower than the re- covery of AIB (82 - 85%). When AIB was applied with formaldehyde, the recovery of AIB was no% which equals only about half the recovery when AIB was used singly. These relatively low recoveries suggest that certain 17h amounts of chemicals were degraded or lost. Several values for each composite of each haylage are in Table 39. There were considerable variations in dry matter content within each treatment (silo), pro- bably due to variation in dry matter content of various loads when harvested as well as moisture migration. Some variations in concentration of organic acids were found within treatment but these variations were not consistently observed among treatments. For example, concentration of total VFA was greater in composite three than in composite one, for the .8% propionic acid treated haylage but the reversed situation was found in the .h% propionic acid treated haylage. Silages containing dry matter content above 30% have a negative relationship between dry matter content and concentrations of various organic acids (Gordon gt 21° 1965). This negative re- lationship was also apparent for acetic acid concentration of the present study. The correlation coefficient was -.83 (P‘(.O1). Within each treatment, large variations in fungal counts were evident among composites. These variations were probably due to insufficient mixing of additives, samples and variable sized clumps of fungi formed in silages. 175 . . . . . . . . m 0Haa moacoaoo w o m o o N m H m o o o o m o m waq%a\mo mdeoo Hmpoe 0 0 0 0 0 0 0 0 0H00 oases: 00.0 mm.m 00.H mm.H Hm.m m0.H HH.N mH.m 0 Hapoe 0 0 0 0 0 0 0 0 0H00 0H00H0> :0.0 0 00.0 :0.0 00.0 No.0 00.0 :0.0 0a00 oaphpsm 00.0 No.0 00.0 H0.0 :0.0 No.0 m0.0 m0.0 ewes oapmpsnomH m0.0 00.0 00.0 00.0 H0.0 Hm.0 s:.0 ::.0 0a00 oacoamopm mm.a m:.H mH.H 00.0 m:.H m0.H mm.H :0.H ewes oapmo: Azm&v mUHom oanwwpo 0.: 0.: 0.: 0.: 0.: m 0.: m mm 0.mm :.0m 0.00 0.:0 0.0m 0.:0 0.0m 0.0: A00 000 ms sesame ham 3 m N a : m N H 0.02 opfimogsoo Amy 0A0: 00.0 Azmm.0 UHO< oHQOHmOLm .momwahwm popwome was Hoppdoo Mo mpsdoo 0H0: paw mUHo¢ oanwmao .mm .pnopmoo hounds ham .om manwe 176 0.0 0.0 0.0 0.0 :.0 0.0 m.0 0.0 Aw\000a0v 00000000 mendw mo mandoo H0009 0 00.0 mm.0 0 0 0 0 0 0000 000000 00.0 00.m 00.0 0:.m 00.0 mo.m 00.0 0m.m 00> 00000 0 0 0 0 0 0 0 0 0000 000000> No.0 0 00.0 00.0 0 0 00.0 00.0 0000 0000000 00.0 0:.0 mm.0 00.0 00.0 00.0 00.0 m0.0 0000 0000000000 00.0 00.0 m0.0 00.0 00.0 m0.0 m0.0 No.0 0000 000000000 ::.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0000 000000 0mm &meHow o0fl0mao 0.: 0.: 0.: 0.: 0.: 0.m 0.m m mm 0.00 0.0m 0.0: 0.0: 0.00 :.0m 0.0: 0.0m :00 000 00 000000 000 d m ,N H d m N H .oz opHmOQSoo Amy 000 00 0m.0 803 000.0305 000 03000005. A0000000000 .00 00000 .00000 00000 00000000 n 0000 .Eopuon n d .900 u H 00H0m 030 G0 HoboH Ome .0Hafiwm owthwa 000momEoo mo .020 .00000000 00m 0 00 0mm.0 - 000000000000 9 .pmnfidfl OHHm0w 177 Aw\QOHHH v mmHGOHoo 00.0 00.0 0.00 00.0 00.m 00.0 00.0 0m.m 000000 00 000000 00000 00.0 No.0 00.0 00.0 00.0 m0.0 0 0 0000 000000 mm.0 00.0 00.0 m0.0 00.0 0m.m 00.0 mm.0 000 00000 0 0 0 0 0 0 0 00.0 0000 0000000 00.0 m0.0 00.0 :0.0 :0.0 00.0 00.0 00.0 0000 0000000 :0.0 00.0 m0.0 00.0 00.0 m0.0 m0.0 00.0 0000 0000000000 00.0 :0.0 00.0 :0.0 m0.0 00.0 00.0 00.0 0000 000000000 00.0 :0.0 0:.0 m0.0 :0.0 0:.0 mm.0 00.0 0000 000000 Asa &Vmc0ow 00G0w00 0.m m.m 0.: 0.m 0.m 0.: m.m 0.m 00 m.0m :.mm m.0: 0.0: m.mm 0.0: 0.m: 0.m: 000 000 00 000000 000 d m N H d m N H .02 000momEoo 000 0:0 0000000 00000 + 00m.00000 1" 00000000000 .00 00000 178 IV. Effect of Treatments on Chemical Composition of Haylage Representative haylage samples taken from sheep feeding trials 1 and 3 were analyzed for chemical and fibrous components. These samples were not composites but relatively small portions of haylages near the top (sheep trial 1) and the bottom (sheep trial 3) of the silo. Proximate chemical composition of these samples are in Table 00. Generally, chemical composition was similar among haylages although AIB treated haylage contained slightly higher levels of nitrogen than other haylages. The AIB contains the NH; moiety which would increase the nitrogen content of the treated haylages. Bottom area(sample of period 3) were generally the lowest in crude fiber content and consequently values of other constituents particularly nitrogen free extract (NFE) and crude protein were increased. All haylages were similare in NFE content indicating similar degrees of fermentation since NFE is normally the primary source of fermentable carbohydrates. However, proper evalua- tion of changes in chemical composition for each treat- ment are possible only when the composition of the original crop is known and compared with that of the silage on a quantitative basis. Unfortunately, insuffi- cient and non-representative samples of original forages were collected and no analyses were performed. Thus, 179 .00000000 00m 0 00 0mm.0 - 0000000000000 .mnwoe 000 mmmmspnmpwm :0 mmusbo .mpmphpSQOmH E50doss¢ 0 .H0000 a00000w00 @0030 G0 0000000 .cmwonp020 000.000 . 00.m0v 000.mv 000.000 0m0.0v mm.0m mm.mm 00.00 00.m0 00.m 00.0 00.0m 00.0m 00.0 00.0 000 0000000 000.mmv 000.000 000.00 000.mmv 000.00 . mm.0m 00.00 00.00 00.00 m0.m 00.0 00.00 0m.0m mm.0 m0.0 000 00000+000.00000 000.000 000.000 000.00 000.000 000.00 :0.0m 00.00 0m.0m 00.00 00.m mm.m :0.0m mo.mm 00.0 mw.0 Amy 00 000.000 00m.00v 000.00 000.000 0m0.0v 00.00 0m.0m mm.00 00.00 om.m 0m.0 00.mm m0.0m 0m.0 00.0 000 0m.o 0000 00m.0zv 000.000 00m.mv 00m.mmv 000.00 00.0: 00.0: 00.00 0:.m0 00.0 mm.m 0m.mm m0.mm 00.0 00.0 Amy 00.0 000.0mv 000.000 0m0.0v 00m.0mv 0000.00 0m.0m 0m.0m 0m.00 0m.00 0m.m mm.0 00.mm 00.0m 00.0 0m.0 000 00.0 0000 00a00mm00 ............................. 00 0 --------------------------- m0 00 m0 00 m0 00 m0 00 m0 00 powppxm £000000 powppmm Hmmwm IIIMIII .oWQMH0w omammz 00500 Amflpm 06:00 a 0 pGoSpmeB .mphflmvafipom paw mummhdeOmH Ed0QoEE< mo 00:0N0S paw onwahpsnomH ESHGoSS< .0000 00n009000 £003 0000009 mmehwm paw Honpnoo 0o Sowp0moQEoo HmoHSmQU .od mHan 180 data in Table no can only be interpreted from a quali- tative aspect. To obtain quantitative changes in com- position, dry matter recoveries must also be accurately determined.‘ For each haylage, fibrous consituents and acid detergent insoluble nitrogen (AD-N) were analyzed on all three samples taken during each of three sheep feeding trials. Again, these samples represented approximately the three positions in the silo (top, middle and bottom). Values in Table A1 are comparable among treatments. Again, quantitative effects of chemical additives on fibrous constituents could not be accurately determined due to lack of analyses on original forage. Under laboratory and farm conditions, acid-detergent insoluble nitrogen as a percent of dry matter and as a percent of total nitrogen(AD-N/N) was increased in forages heated artificially or naturally (Van Soest, 1965; Goering .gt.al. 1972; Huber §t_al, 1972). In the present study, close relationships between haylage temperature and these acid detergent insoluble nitrogen fractions were evident (Table #2). Haylage temperature computed as mean storage temperature or as degree-days above 35 C was significantly (P< .OS) and positivaly correlated with all four analy- tical fractions although the correlation coefficients were slightly higher for degree-days above 35 C. 181 Table A1. Fibrous Constituents and Acid Detergent In— soluble Nitrogen in Control and Treated Haylages. Values are Means of Three Determi- nations From Three Composite Samples. .' . . b AIB Propionic ACld AIB + Control 0.0% 0.8% 0.5% 1% HCHOC (8)3 (3) (7) (5) (Ln (0) Cell walls (% DM) P-l Sh.2 56-h 52.8 53.9 55.6 59.3 P-2 52.7 53.2 u8.6 A7.3 u8.7 52.9 P-3 52.5 50.0 53.8 50.3 u9.1 00.0 i’ 53.1 53.3 51.7 50.5 51.1 57.u Hemicellulose ( % DM) P-l 7.8 13.0 0.7 10.0 5.2 8.0 P—2 9.0 9.5 5.7 6.7 5.9 10.0 P-3 8.3 V 11.5 8.0 0.3 5.7 12.1 ‘X 8.u 11.5 7.0 7.7 5.0 10.0 Cellulose (% DM) P-1 35.2 33.8 35.0 33.5 37.? no.2 P-2 30.2 35.1 33.0 31.7 33.3 33.1 P-3 34.8 35.7 3s.u 3h.6 33.8 38.2 i' 3h-7 3h-9 3b.? 33.3 34-9 37.2 Lignin(%DM) P-l 11.2 9.2 11.1 10.0 12.7 11.2 P-2 9.5. 8.7 9.3 8.9 9.5 9.8 P—3 9.0 9.1 9.8 9.3 9.7 9.0 i’ 10.1 9.0 10.1 9.5 10.6 10.2 ADFd(%DM) P-l u6.u u3.o 00.1 03.8 SO-h Sl-h P—2 03.7 03.8 02.9 00.0 02.7 02.9 P—3 nu.2 nu.8 AS. #000 MB. A7.9 3'6 10.7 L13.9 L110? u2.8 15.5 [NJ-L 182 Table 01. (Continued) AIB Propionic Acid AIB + Control 0.0% ‘0.8% 0.5%* l%‘ HCHO - (8) (3) (7) (S) (0) (0) ADF-Ne(%DM) P-l 0.30 0.29 0.32 0 0.02 0.50 P-2 0.32 0.29 0.29 0.3 0.32 0.37 P-3 0.27 0.20 0.30 0.38 0.30 0.33 i’ 0.32 0.27 0.30 0 39 0.03 0.00 ADF-N/Nf(%) P-l 12.7 11.7 12.1 15.0 20.3 19.8 P-2 11.3 11.0 10.0 12.8 10.8 13.7 P-3 10.0 9.0 10.1 11.2 10.2 12.8 'i 11.3 10.8 10.7 13.1 13.7 15.0 aSilo number. bAIB = Ammonium Isobutyrate. CH0H0 = Formaldehyde -1.25% of a 37% solution. dADF = Acid detergent fiber. eAcid detergent insoluble nitrogen as a percent of dry matter. fAcid detergent insoluble nitrogen as a percent of total nitrogen. 183 .Loppwe haw wo pcooaom a ma Gfidwfia pnomhopoo Uaow QQ< h .poupde haw mo pcoopom a ma honflm pcomnopop Ufiow ma¢ 0 .mo.o vnH * 30.0 E $0 .oagfiwm pnowcomoccfi mm mm omaoofimcoo ma oaflm mo Hoboa Some 8099 coxwp mmwahmn “moHQEwm mo momena u 20 .GoMOApfin HmpOp no pcoonom 8 mm Gomoapfic oHnSHOmGH pcomaopoo Ufio¢ n Z\Zuaonw mhwouooawom x :o. + :m.m u w wdm. x so. + He.m: u a .Nm. ma 0 mm 00000 when N 0H. + mm.m u m twee. x 0H. + wa.>m u.w #5:. ma 0 ohdpwaomaop owwaopm new: mqm¢ th< x Ho.+ m:.oH n % *kow. N Hooo. + cm. H a 33mm. ma 0 mm o>onw mhwwuoommoa x ma. + ::.m u w twee. x moo. + em. u w -m m. mfi 0 mm 00000 when N mm. + wa.mn n % whom. x 000. + moo.u n w khwo. wa o oedpwaoQEou owwmopm G802 :: w:: Coapwswm A Soapmamm L 02 QZ\ZIQ¢ $21Q< .mGOHpowam Hwoflphams< pooh one mopdpmammsoa owmahmm coozpom mQOHmmmpmom awonflq .m: oHan 180 Number of days above 35 C was significantly but not highly correlated with AD—N/N and ADL. Acid detergent insoluble N as a percent of total N was related with haylage temperature to a greater extent than were other analytical fractions. Acid detergent fiber was generally less affected by the storage temperature than was lignin and this was probably due to the fact that acid deter- gent insoluble N fraction is primarily associated with ADL and not ADF (Van Soest, 1965). Trends in composition among levels in the silo were inconsistent except that acid detergent lignin (ADL), AD-N and AD-N/N were greater in the upper portion than in lower portion (Table 01). Since haylage tem- perature during storage was greatest for the upper portion of the silo (Table 35) and temperature incre- ment was positively related to acid detergent lignin and insoluble N fractions (Table 02), the greater value of AD lignin found for the haylage in the upper portion was presumably due to the higher temperature developed in that portion of the silo. In fact, the correlation coefficients were highly significant (P«(.O1) between acid detergent insoluble N fractions and haylage tem- perature of haylages from the three silo levels (average of the 6 silos) (Table 03). A similar effect of 8110 levels was also evident between temperature develop— ment and acid detergent lignin or fiber. The significance 185 .mo.o em .2. “So yo item .cowoapfls u 20 .Loppwe %AU 20 .pcoMmeoo doom 3 .moafim 0 mo prOp map pom Ho>oH oawm oEmm 03» mo oSHm> owwao>w .moamewm wo nomads H mm x 0H.0 + 0~.wm H w 00. x Ho.o + 0.m: u w mu. m ponfl010< x HH.o + oo.m n m mo. x moo.o + oH.o n a memo. m cflemflfi-o¢ x mm.o + ma.- u m ttmoo. x ooo. + H:.oH n w ttooo. m z\ooH a z-o¢ X 000.0 + m0.0 H w **®0. N N000. + 0m.0 H W 0**N00. m 02D & mZnDQ< Goapmswm a Goapwdvm A c o madpwpoMEop omwpopm com: 0 mm o>opm mmwonompmom m N .mQOfipowmm Hwoflphawc< 950m one mao>oq oaflm moafle no oopsmwoz opdpmpomeoe omwahwm CooZpom mGOfimmommom awocfiq .m: canoe 186 of silo level on animal performance will be discussed in the following section. V. Effect of haylage Treatments on Digestion Coefficients and Nitrogen Utilization Data collected from three sheep digestion trials were analyzed by two way analysis of variance. An example is given in Table 00. Table 00. Table of Analysis of Variance Used to Analyze Data of Sheep Feeding Trials? Degrees of Mean F Source Freedom Square Ratio Treatment 5 28.59 28.60::9 Period (silo level) 2 202.87 30.80 Treatment x period0 10 11.05 1.55 Error 18 7.10 Total 35 aUsing analysis of variance of nitrogen diges— tibility as an example. biz-i:- P < .01 CInteraction between treatment and period. Chemical treatments had no significant (P)'.05) effects on maximum haylage DM intake or intake during sheep digestion trials although slightly higher intakes for propionic acid treated haylage and lower intakes for AIB treated haylage were found (Table 05). 187 .Am0.0 v@ V pnwoflwflnmflm ohm mpmflpomaomfim pnopowmfic Qpflz modaw> m .eowHo: moon u 3mm .popooe ego u zoo .QOdeHom &wm w mo &mm.au oohnovaEhomm .AopEdc oHHw< oomH- mo- m:H- Ho- :m- om- Amoo\mv momdogo 3m com. H QMm.w 2m. nwm. 0 am wmm.ma poedmcoo R mo oocHopoa 2 0H:. H emo. H “mm. mH new. NH oooo. o omo.mm oootomoo s we oooHopop z oo9 o 0mm. H mmoo. m eHo. H 0mm. m oHH.: Hmooxz wv ooquoo 2 m5. mm ooH. Ho 00H. oo 0H0. Ho mom. om oo.Ho ARV mpHHHoHomomHo ooHooNHHHpo comotsz %z. o: oo:. :: on. Hdo 00H. e: 00 Ho. 0: oeo.~: doeHo poomoopoo oHo< oo9 s: oomH. om 0mm. on flow. H: mom. 0: :.m: moeosoHomdoo HHoz HHoo 00H:.mm one. om 00o. om ommo. mm 0H0. :m emo.:m 000000 oHeemeo oomN.Hm m:9 o: mom. o: doe. mm “:0. mm 0:.mm 000008 00o HfinpeoHoHooooo eoHomomHo ooe.m ozm.m 0mm. m oo:.m oom.m 0mm.m mHoHno doHomomHo moHeoo om:.m 0:H.m 0mm. m o:o.m oom.m ooo.m sosHaoz m Aozmfiv oaoodHozo omonwm o o o o o 0 00000 .02 H00 H30 HWW . vao Amy «Hov omom . . Hoapcoo m + AmHma mv mHo>mH OHHm aspen mm: gonads mamfiwm .0 on : moanwamw> pom .mawflhp Goapmowfip oops» CH poms momflm mo gonads HepOp was gonads oHQEwm .m was H mofinwflpw> pom .moHQEwm mo gonads n so .pGoHonmooo Coapmowflp Gomompflu H 929 .pQ®H0flMM®Oo QOHpm®MHU honpwfi hhfi H szw ex Ho.o-em.me u w**fim.- ex Hoo.-mm.mm n w mm.- ma 0 mm m>onm memeIompmoo ex we Hm.o-mo.ee u wxxmm.- mx mo.o-ma.mm u e ma.- ma 0 mm m>owm meme mx . I . I =: .I . I . I .I ov x do 0 w: :w I M**m® x HH 0 mo pm I N om ma madpwamQEou omwAOpm awe: : mx mm.HH+Hm.am u **sm.o mx mm.m-em.ae u e *H:.- on sea .2 Hence M Na mm.a-em.:~ n e**ae.- mx o:.o-o.mm u w **::.- em 2\00sz maneaomefl ea m ax m.me-am.o® n w**mw.- ax 0.3mIee.oe u wm**oo.- om czma .mz mH95HOmeH a Soapwdvm a Coepwswm a on pmz maze x .mpcoEopSmmoz chapmAmmEme moage pew wooeuompm cowoppflz mHQSHquH pcmwampoplWHo< 039 no meoHOflmmmoo Coapmomflm Cowoepfiz paw Loppmz ham o>fl> CH LOQ mGOHmmmpmom awocfiq .0: manme 191 matter digestibility ( -.h1). This negative relation- ship must be considered abnormal (Holter and Reid, 1959). These calculations strongly support the idea that total crude protein analysis is inadequate in estimating nu— tritive value of haylage and other more specific labora- tory methods such as N solubility in acid detergent solution should be used (Goering 23 El' 1972; Thomas and Hillman, 1972; and Goering and Adams, 1973). A close relationship between haylage temperature and in XEXQ responses should be expected since tem- perature was significantly correlated with AD-insoluble N (Table A3). Nitrogen digestion coefficient was sig- nificantly'(P‘(.O1) correlated with all three expressions quantitating haylage temperature development given in Table us and no marked differences in correlations were noted among these three different methods of tempera- ture computation. Dry matter digestion coefficients were negative but insignificantly related to haylage temperature. The differential significances of these negative coefficients demonstrates the detrimental effect of heat development (i.e. above 35 C) on ND of haylage exceeds the effect on DMD. The sheep digestion trial was performed over 3 time periods of about 17 days each with 2 sheep on each haylage during each period. A significant effect of 192 period on several digestion parameters was noted (Table #7). Acid detergent fiber was significantly more di- gestible during the first period than in the second period. Nitrogen utilization parameters steadily im- proved during sequential periods. Thus nitrogen di- gestibility values were more satisfactory for the lower portions of the silo. Sheep lost significantly less body weight during the third period than during the first two periods. Improved nitrogen utilization should partially explain the changes in sheep body weight. Previous data presented (Table A3 and he) have indicated that haylage temperatures were positivdy correlated with level in the silo and with amount of AD-N or AD-N/N and to storage temperatures. Values involved in these re- lationships are summerized in Table AB. Total protein content of the three levels is also shown in this table and attempts to illustrate the poor relationship of total N to in 1219 N digestion coefficients and other parameters closely related to N digestion coefficients. Figure 7 summarizes relationships among several variables considered to be important in haylage and animal production along with their correlation coeffi- cients from this trial. Degree-days above 35 C for the top silo level was positively correlated but 193 .Gmwoapflc H z m .Amo.o vm V powOHMHGmHm mam mpmflpompmdim pcoaommflp Spas mmsfim>a om.om- o:.mHHI m:.mma- Aeme\wv newness pnwaoz eeom w 9 Q mam.ma . nom.: nem.m eossmeoo 2\ooa x_eoeaepma z maa.dm nmo.© 9mm.o popaomnw Z\ooa N pmaflwpopmz amoa.: n::o.a nmoo.o Meme\z my mossawn ammotpaz wwm.mp £20.00 0mm.:m ARV Cowoapflz emea.e: noo.m: m:m.w: Asv ponse pummempoe eaoe H pdoaoamwmoo Goapmmmwm m UOfiaom m poahom a powmmm mSopH one hp pooCoSH%QH .Aoaflm cflp Ce Hm>oqv mEHB mo pommmm %HpcdeHMficmfim mama £0H33 wpwm moamELOMme mooflm .N: oHQwB I9u .Qowompan n 29 .Aoppws haw n saw om.om- o:.mHH- m:.mmH- Aewe\wv mmweaeo nemaoz eeom mm.mm :0.mm Hm.mm R .pfloaoammooo GOflpmomwp pounds oacmmao mm.am mw.mm Hm.om R .meaoammmoo soapmomfip popuws hum mm.me 20.00 mm.:m & .peoaoaemooo coapmmmae ammoppaz momsommoa obfib GH Hm.o mm.o mo.OH sea .eaemaa eemmnouoe eao< 00.0H om.HH :m.mH z\ooax z oHQSHomGfl ucowaouop pflo¢ mam.o Hmm.o Hm:.o saw .22 mapsflomea pemmnopoe eaoe Ne.ma m:.efi HH.»H meme .sz campoeg Hence QOflpHmomEoo HonEoflo mm em 0: o powpom wwwAOpm mcfladp oASpwaoQEop omwao>< Hm omH 03m 0 mm o>onw mhwpnmopmom moASpwaomEoB soepom mauve: doe moaflm CH mao>oq .oaam one CH mao>mq one on popmaom mpGoSmmSmwoz probom .m: mance 19S .mo.o we * .Gmmoepflc n 29 .aoppme hep n Sad .moamewm om mchQomoaQom paw monEwm o mzflpcomopmmm II II II “mmamfiwm ma wcflpcomoamom .l..lI.II..o.a .moanwfiam> 039 mg» mcflpoonoo mafia hp powwoaan we womb oaaedm mo LonEdz .mpamfiofihmooo deflpwaoaaoo hwocaq mameflm pfioze paw soapwsawbm owwahwm sfl pflwpaomSH on on popopflmcoo moapwflaw> onS¢ mQHSmCOprHom mo Goapwpsmmoam oflpwaonom .N oasmflm 196 ®N.o I Q'OIOIO'O|OIDl e as. _ a 4.4-45.2.-. 3 a. 2-395325 m .o . o m gone proa *N®.I 7Q :mwaopovuwfio¢ # _ mhwmIoommo Isms . n 41 _ . . . _ 0:.0 . and. . _ _ . _ omé- t _ m * _ . F . I h . Aaoboa mo» $33333 3 *4: o _I L n 33$:an W0 IE IT mm 26% *H:.OI hounds he Mhdouoonwo 197 insignificantly (PJ>.OS), with the extent of top spoilage ( + .61). No correlation was observed between the average of degree-days above 35 C measured from three silo levels and top spoilage. The extent of top spoilage was not significantly (P)>.OS) related with either acid deter- gent insoluble fractions or $2 gixg responses. How- ever, when heat development was quantified as degree- days above 35 C the significant (P‘<.OS) negative effect of relatively high temperatures on in 2212 N availability of resulting haylages was evident. The reduction in haylage N availability by heat was pre- sumably due to the increased formation of acid deter- gent insoluble N fraction which is indigestible by ruminents. Thomas and Hillman (1972) reported that as much as 20% of the forage N is commonly found as acid detergent insoluble N in haylage made in the state of Michigan. Analytical and sheep performance data obtained from this study were not used in the regression analyses of nitrogen digestibility on analytical fractions pre- sented in Part 1 of this thesis. For the 21 earlier MSU samples used in Part 1, ND was significantly re- gressed with acid detergent insoluble N (Table 1h). Similar results were found using data of this experi- ment. Acid detergent insoluble N as a percent of dry matter was highly correlated with nitrogen and dry matter digestion coefficients. On the other hand, 198 total nitrogen content was clearly inferior to AD— insoluble N % DM in estimating nutritive value of hay- lages for ruminants. VI. Effect of Chemicals on Haylage Consumption, Milk Production and Composition of Milk of Lactating Cows Haylage dry matter and total dry matter con- sumption by lactating cows fed the haylages for a #9- day feeding period are presented in Table M9.' During the 12—day preliminary period, all cows received re- gular herd haylage with no differences among groups. Haylage consumption (kg DM/cow/day) increased by a factor of 1.32 to 1.70 for the treated haylages with no significant differences (PZ>.OS) among groups. When haylage dry matter consumption was calculated as per— cent of body weight then significant differences were found among groups during the experimental period with cows received .8% propionic acid treated haylage con- suming significantly (pI<.5) more than cows fed the other haylages. Total dry matter consumption (haylage DM + grain DM, kg/day) was not significantly different among groups. However, the percent increase was significantly lower for the 1% AIB treatment group than for the other four treatment groups. Total DM consumption as a per- cent of body weight was not significantly (P>'.OS) 199 .mCOpr>pomno CH poms macs whee m: pme hHQo p59 need a: was oowamm pcospwoae o m .aoppwe hep u 29m .Amo.o vmV paonMHame one mpmHnomammfim pcvoHMHU QHHS oQHH one do mmzaw> .mhwp :H was pOHpmm hawcfleflaoam: .GOHpSHom Rem m mo Rmm.a .mphnmpaweaom H omom .monesn OHHm N . H emHe.o eeHe.o emmm.o memo.o emmm.o emmm.o Hememmstemeeee HszeemHez meom BmOH emmH emOH emWHVHH emmH eeeiHHH A.eHHeem\OOHs.eeeeeVemqeeo wee. m Moe. m eme.m eem. m Moe. m eon. m Ham RV eOHeee escapeete %om. m em9 m eme.m MoH. m epemm. m Mme. N am RV VeOHeem eeeeHsHHeem eweHH emmH QQOH ewmmH eemH eMmmH A. eHHeee\OOHx.HeeteVemeeeo ewe. :H eoH. eH eem. :H mm9 mH woo. 0H emm. mH Heee\meV eOHeee HeeEeeeee M09 MH emm. NH Mmm. mH Mmm. NH Me9 mH Mew. mH Heee\msVe0Htee epeeHeHHon QOHmedeoo SQ proe HwmmH emmH eemH eemH eemH eeeH A.eHHete\oon.peoteVeweeeo Mmm. H Mom. H Mmm. H Mme. H New. H nmm9 H AsmsV eOHeee peeEeeoee eeH. H Mee. 0 eem. H .eoo. H MtH. H eeH. H Ham RV VeOHtee eeeeHeHHeem e02H eHeH emmH eeeH e:0H eHeH H. eHHeem\oon.peeeeVemqeeo eme.m Mm9 e ewe. w .eHm. e eem. OH “He. e Heee\meV eOHeee HeeEpeeee eom.e Mme. m em:. 0 “mm. m “mm. o Mm9 o Heee\msV :eWHeee eteeHeHHetm mm QoHmedmcoo mzm owmawwm AeV WNW MMV amp“ AmV HAmV m o o. o. o. Hoepcoo m + AmH.OS) differences for average body weight changes were de- tected among treatment groups (Table h9). Data on milk and fat corrected milk (FCM) yields are shown in Table 50. No significant differences (P) .05) were found for average daily production during preliminary period, during treatment period, adjusted daily production (adjusted by covariance analyses) and persistency (production during treatment period x 100/ production during preliminary period). Adjusted milk yields were somewhat less for the 2 AIB treated hay- lages than for the other four groups (17.3 vs. 18.5). Ranges of persistency and number of cows having persistencies over 100 are given in Table 50. All but the 1% AIB group had at least two cows with persistencies over 100. Data of efficiency of production calculated as FCM milk yield (kg/day)/ total digestible energy (kg/day) are also shown in Table 50. None of the chemical treatments improved the gross efficiency of milk production as compared to untreated haylage. Among the five treatment groups, efficiencies were lower for .5% AIB and .8% propionic acid groups than the other three groups but no marked differences were 201 woo. Nm «OH. mm mom. mm H“8.5 moo. mo moo.Ho AmvhocopmHmHom mmH. 0H wmm. OH Mmo. :H me.:H mmH. 0H www.0H Hqu\mxv cOHHoa .HonH uopmsHo< Qowpodfiopm &HHE Umpooppoocpmm . . . . . . Hvaxmxv 299 ON H mN H @N H om H 00 H mo H \HNHU\wHVHHHs umpomppoompwm waquOHHHm ®\N ®\N ®\O ®\N ®\N ®\m.. .pmdmhmg OOH .H®>O 9500 .02 mm OHH Hm omH :N om NN NOH om 03H mo mOH ANV NoamHmHmHmm Ho omnmm wo:. mo H”ON. do H”om. mm HwON. om woj. :0 Hwoo. so NHNV NoqumHmHom MQH. 0H “mo. 0H Mmm. NH MNm. NH “Hm. wH Mmo. mH HNHU\mxv ©0Hpmm .Hmope ompmsHo< «Na. 0H “mo. mH moo. NH moo. mH MNm. NH moo. mH Ach\mxv ooHHQO cogpwmpe Moo. Hm MHo. 0H Mme. om MmH. om “Na. mH MmN. 0H ANHU\mHVuOHngmprcHeHHon :m COHpoSUOHm &HHE HwSpo< HOV Adv va OHNV Amv HHmv Ho.HpHHOO Nomom RH km 0 R@.O R+WO + AHmHHVonHszn ARm.ova< nomH ESHGOEE< UHo¢ OHGOHmon Umm mzoo %o UOHHmm prQmEHHmme 6cm %HmsHEHHon mmflmsm QOHpodvomm xHHS .mowwahmm wopwmpe cam Hoppcoo om mHQwB 202 .&m> on on UmESmmw ma: thpKflS Gawmw mo pmopcoo zme .Ummm mg» Ga pmppws OHwapo mo namcpmg w mw pownpNo hmnpm H mm .Umom map.cfl noppwe haw Mo pcoohom a ma pmppwfi oacwwno M So .mawwap mcfiuomm ammnm Bonn Umnflwpno hpflafipfipmowflo mepwe canwwho u Q20 when: Ammmumfl .Hom.sfla¢.h .mmoa .m.w .Gmopw%oqv Amm mmaooo. + Ho.vzo x 920 H zme "Boaaom ma ompmadoawo mwz mowwamwfl mo pnopnoo ZQB .hwpmao oHQHpmowHU Haven n zmew .anmmm hmmcflefiaopm msflmdw Goapodvomm\ooa N vownmm onEpwmhp mdfipdw soapodvonmw .mmmhawcw monwflpwboo hp Umpmdh©¢ o .mQOfipm>pmmno CH cams opmz mmwc m: pmma haqo p59 mmww a: was coapmm psmepwoaem AEfiovfl .pcmmmmmflo %HpQwOflMHdem mam mafia mac Ca pmflhommomdm paohoM%H© flpflz modaw>d .mhww :H was UOHLQQ %hwGHEHHmAm m .qoflpsHOm Rum a mo Rmm.fi .mohnouaprom u omomm .monadc OHHmH Acmscflpcoov .om magma 203 found among the latter groups. Adjusted means of FCM and persistencies of FCM production were lower (P)>.OS) for the AIB groups than for the other groups. Waldo (1973) commented on using daily milk pro- duction to compare the quality of silages. Daily milk production will not easily distinguish between quality of silages fed with grain, and he stated that daily body weight gain was a more sensitive criterion than daily milk production. Daily body weight gain can easily be zero with poor quality silages, and the loss of intake potential is much greater than the loss of digestibility potential with poor silages. However, the intakes of poor silages are relatively less depressed when fed with grain to milking cows then when fed alone to growing animals (Waldo, 1973). This may explain why insignifi- cant differences of milk production between treatments occurred in the present study. Value for milk solids and fat are shown in Table S1. Adjusted means for total solids and fat were slightly higher (PI>.OS) for the .8% propionic acid group than for the other five groups. Higher rates of application of propionic acid and AIB resulted in slightly higher adjusted total milk solids and milk fat than did the lower application rates. In any event, propionic acid and AIB added to alfalfa haylage at the .mGOflpwHSoawo Ga on» %Hco pan .UOHAmQ Hmpnoeflmmmkm mo mhwp o: mLHpGo .momhawcw .mhwo :H moqumw>oo hp cmpmfinw< mwz anpog hmwcfisflaoam m com: one: mhmc N: pmwa pom vophoooa mam: moan: m .GOHp2HOm mam a co Rmm.a .mescmeHmELOm u omomm .aonsdc oafima L4 mo.m mo.m mH.m :w.m Hm.m mo.m use saws empmsne< m om.m Ho.m HH.m oo.m mH.m mo.m eoacmm pcmspwope L om.m mo.m oH.m m:.m mH.m oo.m weapon hpmcaeflaopm svamm saw: om.HH 00.HH oa.mH os.HH om.mH oo.mH meaHOm HmpOp meoumshe< om.HH -.HH mo.ma mw.HH oa.ma :N.HH we0flpmg pamepwone om.mH mw.HH 00.HH mm.mfi mm.HH an.aa me upon mamaflsaamam ARV meaflom proe on Adv .Amv .AWV Amy HAmV + AmHOOO OOOO-OOOO0O Ho mo :w mm mm OOH mudpwaoQEmp COOS oowOAOpm wcflhfim COHpoSUom oASpwhomSOB HOV va RWOO Oxmv wmv HOV omom 0H .o om.o o .o n + AOHOVOOOLOOOO OHOO HOOOOOO Afim.ova< IOOH ESHQoEE< OHGOHQOAm .OOH mm Ummmmhmxm mum: omwahwm Hoppaoo pom modaw> cons meOEpOope ommahmm ow Umpwaom mpoommm OSOHMO> mo OCOOHLOQSOO .mm OHQOB 207 HOH NO OO HOH HOH OOH ANV honoumHmaOm HOH NO OO OOH HOH OOH ANOO\OOVOHOHN zOm OopmsOOO OOH HO HO NO OO OOH ANOO\OOOOAOHN OOOOOHOO HO OO :O NO :O OOH ANOO\OOOOHOHN stpoO EOOHposeomm OHH: OO: :Nm ONm Omm OHOH OOH ONOOOH N .OOHOOOOOO z NHH OOOH NHO OOO OsOH OOH OOOOOOOO N .OOHpcopOp z mmN OHm OON OOm mOO OOH ooquOn z HOOHOONHHHOO szqmmopsz OO OO OO NO mO OOH OHHmz HHOO OHH OOH HHH NHH HHH OOH OOOONOHz NO NO OOH mOH :OH OOH toppms ONO xmpSONOHmwooo aoHpOQMHQ OO HOH NO OHH OHH OOH HmzooOanwpoOH NO OOH NO HOH NHH OOH mHOpr OOHOOO Oswch NO OO OO HHH NHH OOH osach aseHsz mOOsO H ApnOHmz NOOO Ox OOH\ONVOmOch mmmflommoa o>H> QH OmOm HOHOOOOONNOOO O OHoO HOOOOOO ARm.ova< nomH ESHGOEE< ONGOHQomm AOOOOHOOOOV .Nm OHOON 208 .OOHOOHOO ONO O No &mN.H .OOOQOOHOenom n OmOm .Hm OHnwea .xHHS OopwHOAnoo pawn 2:: OHQOBH .O: OHOONH .Om OHOOBQ .moppwe haw H 299 . .O OO OHn 9O 9 .Om OHnOeo .O: OHnOes .d: OHQOBUH 2:: OHnOeH .Nm OHOOBO .O mademm .Hm OHOOBO .pmnEda OHHmO OOH hobo OOOSOQOOA o>H> SH OOH mOH NOH NOH :OH OOH OOHHOO HOOO» OOOOOHOO OOH OOH :O OOH OOH OOH paw OOOOOHOO ARV QGoHpHmomEoo xHHz OO OO :O OO :O OOH Hozaa\zOmO NaOOHOHNOO HOV Hmv ANO HOV HOV HOV OmOm NH Om.O xNO.O NHMO HonpOOO + HOHOVOOOONOOQ OHOO ARm.OVmH¢ nomH EstoEE< OHQOHmoam AOOOOHOOOOO .Nm OHOOB 209 acetic acid concentration while no effect was noted for AIB treatments. Lacic acid concentration was essentially zero for haylages treated with propionic acid (.u.and .8%) and .5% AIB. Extremely small amounts of lactic acid was found ifi haylages treated with 1% AIB and the mixture of AIB and formaldehyde. Quantities of acid detergent (AD) lignin and insoluble N fractions were lower for .5% AIB, .8% and .h% propionic acid treatments than for the other treatments. The comparison is similar to the ex- tent of temperature development during storage. “lg.1lgg response measured from sheep were generally highly related to the quantity of AD insoluble N fractions or heat development during storage. All treatments markedly im- proved nitrogen utilization values. No increased responses were obtained from lactating cows fed treated haylages but propionic acid treatments gave slightly "better" responses than did AIB treatments. Values listed in the bottom line of Table 52 represent the total number of in vlzg responses with a relative value over 100. None of the chemical treatments produced 16 improved in vivo responses but prOpionic acid treatments had 12 values above 100 while AIB treatments had six to nine. The principle effect of these acid treatments was on reducing fungal counts, storage temperature and decreasing in- soluble nitrogen fractions in the haylage and increasing 210 nitrogen utilization in the animal. Unfortunately, the relatively action of chemicals in reducing dry matter loss during fermentation could not be evaluated quanti- tatively because of incomplete recording of the "good" haylage fed. Hence, no quantitative comarisons or economic evaluations with respect to animal production can be made for these chemically treated haylages. Further trials are urgently needed. SUMMARY AND CONCLUSIONS Part 2 The value of propionic acid (.N and .8%), ammonium isobutyrate (AIB, .5 and 1%) and a mixture of AIB (.5%) and formaldehyde (1.25% of a 37% solution) in preserving the nutritive value of alfalfa haylage (50% DM) was evaluated by the following criteria: heat development during storage, extent of spoilage, total fungal counts, chemical composition, performance of sheep and lactating cows. Levels used for propionic acid and AIB were com- parable on a molar basis. All chemical treatments reduced haylage tempera- ture during storage. When heat development was quanti- tated and expressed as degree—days above 35 C, the ranking of chemicals in preventing excessive heating was as follows: .5% AIB .u% propionic acid .8% propionic acid .5% AIB + formaldehyde 1% AIB. None of the chemicals were completely effective in preventing heat development near the haylage surface when silos being emptied during the feeding period. The extent of top spoilage was somewhat positively re- lated with the heat develOpment during storage. The 211 212 lowest top spoilage was noted for .8% propionic acid treated haylage. Total fungal counts were reduced by about 30 to 70% by all treatments except AIB (.5%) + formaldehyde treatment which had no effect on total fungal counts. Chemical composition was similar among haylages except that acid detergent lignin and insoluble N frac- tions were greater in those haylages which had the greater quantity of heat development. Dry matter intake determined by using both sheep and lactating cows was slightly but consistently greater for propionic acid treated haylage than for AIB treated haylages. Digestion coefficients for dry matter were not markedly affected by treatments, but nitrogen digestibi- lity was generally improved by a factor of 1.1. Marked improvements in N utilization were observed for all treatments. Significant negative correlations were found between measures of N utilizations and both acid deter- gent insoluble N fractions as well as degree-days above 35 C during storage. Digestion coefficients for fibrous components of the haylages were somewhat reduced by the treatments, particularly AIB treatments. Milk production and composition were not significantly affected by the treatments. Propionic acid treatments, however, produced slightly better results than did AIB treatments. Results from this study indicated that propionic acid was only 213 slightly better than AIB in preserving the nutritive value of alfalfa haylage, but no marked superioety was found for higher levels of acid application. Considering the practical application situation, AIB is more acceptable than propionic acid with respect to oder and corrosive nature. AP PENDIX 21h quOHN w: hopmoEono &m.o Asz Oopdoao Non Oomomum OOHONHO NH m-OOHN OO topmosono mm.H HHH: Oopeoao OOH OoHomaO OOHONHO HH N-OOHN Om aopmosono Rm.O HHH: Oormoap Non OoHompO ONHOOHO OH O-OOHN Hm aoomosono mm.H ner OOpmoar Nan OosQOHO mNHeNHO O O-OOHN O: aoomosono mm.O HOH: Ooreoae Nee Oosomam OHHONHO O 0-00HN OO topmoEOno mm.H HHH: Ooreoap Own oesomrm ONHOOHO N N-OOHN Om maopmo2ono mm.O ans Ooemoao Ne: Oosomom OOHONHO O mmHomumOHN NO ham mhdeHE mmmmw OMHOMH< m onumOHN mm ham QHOAHO oSHN> .Aufic 30HuOMHOMH¢ : gOHn-:OHN NO New aHmarm osHm> .apad OOHH-ONHONH< m Heaao>-:OHN mO New Heaao>-aOHeNHO N pHsaao-:OHN NO Nam HHOOOO-ONHONH< H GOHPOOHprcoOH ARV GOHpmbhomoHQ .oQ OHOHNLOQOH poppme ham wo make moHoon onEmm .mpGoEHmeNm GOHHOSHN>m %pHHOdd GHOHOAm ommpom SH womb OOHQEOm mo GOHpQHhomoQ .H oHQOB NHOQomQ¢ 215 .OHoO OHpmo¢ RONHOHow OHGOHQOHQ Row "HoumoSOQO O-NOON O: topmosono ONO.H HHH: Oopaoao OOOHNOO OHHONHO :N N-NOON Om aopmo5ono KOO. HOH: Oooeoap OOOHNOO OOHOOHH ON HuHOmw w: opOCOHmonm EdHcoEEw Rmo. QpHS Oopmonp omOHhOm OMHNMH< mm NH-HONN mm OHoO oHooHOoaa NO.O HOH: Oopeoao oOmHNmm OOHOHHO HN HH-HONN mm OHom oHaOHOoaO m:.O HHH; Wooaoap OOOHOOO OHHOHHO ON OH-NONN mm o OHNOO ONHOHHO OH O-NONN m: OOOHOOO ONHOHHO OH NH-HONN mm OHoo OHaoHOoaa NO O HOH: Ooreoap oOeHNOm OHHONH< NH HH-HONN mm OHoO oHOOHaoam m:.O HOH: moreoar OOOHNOO OOHOHHO OH OH-NOHN mm oOmHNmm OHHOOHO mH O-NOHN m: OOOHNOO ONHOHHO :H OH-OOHN mO Om: Ooxoeom OOHOHHO OH GOHpOOHMHpGoOH HOHuO>Homomm .on whopwmoan. pommwa ham no make moHoomm OHQEOO ApodGHpcoov .H oHQOB KHOG09Q¢ 216 Appendix Table 2. Approximate Analyses of Samples Used in Forage Protein Quality Evaluation Experiments. fl Nitrogen Sample Crude Ether free Crude Ash no. protein extract extract fiber ------—--—-----—-- % DM --------------------- 1 19.65 2.71 h3.75 23.06 10.83 2 19.30 2.62 h2.00 25.89 10.19 3 21.28 2.h7 h2.29 23.85 10.12 A 20.70 2.56 41.73 2h.15 10.86 5 16.51 1.70 h1.79 33.h1 6.60 6 18.57 1.09 32.53 39.56 8.2a 7 18. 5 1.31 36.78 36.03 7.h3 8 20. 2 0.95 33.5h 35.75 8.95 9 20.36 1.2h 36.17 3h.89 7.3h 10 20.56 - - - 8.52 11 18.76 - - — 6.96 12 20.81 - — - 8.h6 13 17.77 - - - 7.2u 1h 18.26 1.98 39.20 33.35 7.67 15 21.79 3.28 35.2h 31.u3 8.22 16 19.99 2.2 38.8h 31.82 6.91 17 21.88 1.9 36.79 31.86 7.80 18 20.75 2.39 33.51 35.82 7.3h 19 21.68 3.33 3u.35 33.10 7.60 20 19.71 3.0h 35.60 3h.32 7.27 21 16.99 2.89 39.21 3h.07 7.28 22 23.59 3.23 h0.66 23.78 8.73 23 21.70 3.11 u3.76 23.25 8.17 2h 21.90 3.19 #3.58 23.02 8.31 aDetailed sample descriptions are given in Appendix Table l. 217 Appendix Table 3. Fibrous Constituents Analysis of Samples Used in Forage Protein Quality Evaluation Experiments. w samplfi cwcb ADFC Hemicellu‘Cellulose Lignin ADL/ADFd no. lose ---------------------- % DM -------------------- 1 41.58 32.18 9.40 25.69 6.49 20.17 2 44.38 34.28 10.10 27.09 7.19 20.97 3 43.32 33.78 9.54 27.26 6.52 19.30 4 42.50 32.96 9.54 25.72 7.24 21.97 5 49.42 40.96 8.46 31.86 9.10 22.22 6 60.50 48.75 11.75 38.07 10.68 21.91 7 55.96 46.26 9.70 36.36 9.90 21.40 8 58.30 50.28 8.02 37.21 13.07 26.00 9 53.24 49.72 3.52 37.22 12.50 25.14 10 60.69 48.16 12.53 37.44 10.72 22.26 11 57.26 46.34 10.92 36.26 10.08 21.g5 12 57.74 51.30 6.44 37.54 13.76 26. 2 13 53.98 44.62 9.36 35.10 9.52 21.34 14 52.08 44.73 7.35 34.69 10.04 22.45 15 42.28 41.12 1.16 32.70 8.41 20.45 16 49.38 42.24 7.14 32.78 9.46 22.40 17 50.18 42.16 8.02 32.59 9.57 22.70 18 50.48 42.01 8.47 32.77 9.24 22.00 19 44.15 41.16 2.99 32.50 8.66 21.04 20 45.82 42.44 3.38 33.82 8.62 20.31 21 57.08 42.94 14.14 33.00 9.94 23.15 22 34.95 30.43 4.52 33.00 3.61 25.01 23 36.81 25.00 11.81 16.65 .35 33.40 24 37 86 29.19 8.67 21.75 7.44 25.49 aDetailed sample descriptions are given in Appendix Table l. bCWC = cell wall constituents. CADF = acid detergent fiber. dADL/ADF = acid detergent lignin x lOO/acid detergent fiber. 218 OO.HN O:.N OH.ON OH.OO O0.0m ON.ON HO.OO NN.H HO.N: :N OO.ON NO.O OH.ON O0.0: OO.HO OH.OO NN.OO O0.0 O0.0: ON NH.ON OO.N OO.ON OO.:m OH.OO OO.OO NO.NO H0.0 NO.N: NN :0.0N ON.m N0.0m - - NN.NO OO.NO OH.O O0.0N HN mN.ON OO.: mN.OO . . O0.00 O0.0m O0.0 OO.:m ON NN.OH Om.m NN.NO - - ON.OO OH.OO O:.: OH.Nm OH OO.NN ON.m ON.OO - - OO.NO ON.Nm O0.0 N0.0N OH H:.HN NH.m HO.Nm OO.N: O0.0: H0.0m O0.0m :O.: Om.OO NH OH.ON HO.m mm.:m O0.0: Om.N: NN.:m O0.0m Nm.: OO.mO OH NH.OH OO.O O0.00 - - OH.Nm HH.Nm O0.0 OO.O: OH ON.ON 4:.O OO.O: ON.OO ON.NO OO.NO OO.HO OO.: O0.0N :H OO.N NO.- NH.NO mm.m: N0.0: O0.0m NH.:m mO.N OO.:O OH NN.ON OO.N- O:.O: .O: ON.HO O0.0: O0.0: NO.N NN.NO NH OO.OH O0.0 NN.OO O:.N: ON.OO O0.00 OO.HO :N.N NH.OO HH N.OH OH.O OO.NO NO.H: ON.NO H0.0: NO.O: OO.N OH.ON OH .N- OO.- OO.O: OO.O: O0.0: HO.O: OH.N: m:.N OO.:N O OO.O NO.H OO.Om HO.N: HN.N: OH.O: NH.OO OO.N NO.NO O ON.H- O0.0 OO.NO O0.0m :m.mm OO.NO OO.Om HN.N NO.NO N O0.0- OO.- O0.0m NN.H: OO.HO NH.OH OO.N: HO.H :O.:N O OO.OH mO.H OO.ON mm.H: ON.OO OO.Om O0.0m :O.N ON.OO m OO.:H HO.O NN.ON .H: NO.N: ON.OO ON.HO ON.O HO.OO : OO.OH ON.: HH.NN N0.0m NN.NO OO. O OH.OO ON.: N0.00 O O0.0H O0.0 OO.:N H0.0: O0.0: OO. O ON.NO OO.: ON.OO N OO.OH OH.O ON.ON OO.N: OO.O: NN.OO OO.HO O0.0 HN.OO H a NOO\z O ............. -- N ................ Ozm m m oSHw> oxOpCH peg mm .0: OHeoHOoHonooamHen-z on OOOOO oozon QzOO maze essHamz 2O OHOemO .mpGeEHpogxm QOHpOdHO>m thHOSO GHOHOAm mmmpom CH womb OOHQEOm mo Open ooqmshommom moonm .: OHQOB NHOGomm< 219 .Oooaomoe 2\OOH a ooemHeo-z .coOoaeHa mo NOHHHOHOOOOHO u 2O .aonHH ananroO OHom mo NOHHHOHOOOOHO .mpdoOOHpmeoo HHms HHoo Oo NOHHHOHOnoOHO .moppwe OHGOwao mo thHHanmowHO .aorres Nae Oo NOHHHOHOmoOHO .OOOHos Noon u 3mm odHO> HOOHMOHOHOH.H m mQ¢QU Osan EOQQ an. ApesmeGoov .: OHQOB NHOGOQQH 220 OO.OO HO.HN ON.Oz OH.OO HO.ON N0.00 :0.00 :N NO.NO N:.ON H0.0: OO. O O0.0N HH.O: NN.OO ON O0.00 NH.ON OO.N: OO. O NO.OO ON.NO NO.NO NN HO.Nm O0.00 O0.00 NO.N: OO.Nm O0.00 OO.Nm HN m:.Om ON.ON HN.N: OO.N: OO.NO OO.NO O0.0m ON ON.O: NO.mO mN.O: 2m.O: O0.0m NN.OO OO.Om OH NN.OO OH.OO NH.OH OO.O: OO.Nm OO.:O ON.NO OH - HO.ON HN.m: NH.OH OO.NO NH.NO HO.Om NH ON.Om O0.00 H0.0: NH.H: ON.Nm O0.00 O0.0m OH ON.OO N0.00 OOH: N113 O0.0m OO.OO HH.Nm OH - O0.00 OO.HO - NN.OO ON.OO OO.HO OH OH.OO NO.OO ON.OO O0.0: O0.0m ON.ON NH.OO OH OO.N: O0.0m O0.00 OO.:O ON.O: OH.ON O0.0: NH HH.Om :mOm m0.00 OO.HO O0.0m NO.NN OmLOm HH H0.0: O0.0m OO.:O ON.OO OO.HO OO.ON N0.0: OH NH.Om H0.0m O0.0: N0.00 ON.OO NO.NO OH.N: O O0.0: HN.:m H:.OO NH.OO NN.OH :N.ON NH.OO O O0.00 N0.00 O0.0N Om.mO N.Om OO.mN O0.0m N OH.O: OO OO OO.O O . O N.Hm OO. N OO.N: O 433m HN.:O OmO: OH.N: :0.0m OHHO O0.0m m NN.Om HO.OO HO.Hm O0.0m OO.NO ON.:O ON.HO : OO.HO NO.NO ON.OO :0.0: NN.OO N0.00 OH.OO O OO.HO H0.0N NN.NO OO.N: NN.OO ON.HO ON.NO N OO.HO NO.ON O0.0m OO.N: NO.NO NN.OO OO.HO H 2O.Hom.eea+ 2O .Hom 2O.Hom Omen 2O .Hom 2O ozm o.Hom maze .oa nHmmom+.Edm QHmmom+HEdm + QHmmom nHonm .HOOOQ< popmz pom o>H> mm. OHQEOm .meoEHaomNm GOHpOSHO>m thHOSd chpoAm SH womb OOHQENm owwpom :m Mo OOHpHHHnSHom poppmz ham .m oHnOe KHOGon< 221 .OHSHM amaze H .Edmm GHpmothOQ H ammo pcemaepep OHow n DHO .moppme haw H EGO .OHQSHOO H .Homn .aoopms Nae mo OOHHHanmoOHO u mzma Apoanpdoov .m OHQOB NHOQOQQH 222 NH.NH OH.O OO.H: OO.N ON.O OO.NO NO.N OH.ON OO.O ON ON.OH OH.O O0.00 OH.N ON.O O0.00 NO.N OH.ON N0.0 ON OO.OH OO.O NN.O: NH.N NN.O OH.OO OO.N O0.0N NN.O NN HH.HN O0.0 OO.OO OO.NH O0.0 ON.NO HN.H NO.OO NN.N HN N0.0H O0.0 O0.00 O0.0 ON.O O0.00 ON.H ON.OO OH.O ON OO.NH H0.0 OH.HO OO.O :0.0 OO.NO NO.H NN.NO N0.0 OH ON.NH N0.0 O0.0: NO.HH O0.0 ON.OO ON.H ON.OO N0.0 OH OO.ON HO.O OO.O: OO.OH OH.O HO.O: OO.H OH.NO O0.0 NH OH.ON ON.O ON.NO O0.0H O:.O ON.O: OO.H OO.HO ON.O OH OO.HN ON.O O0.00 ON.OH N0.0 OO.HO NH.H O0.00 O0.0 OH - - - NO.NH N0.0 NO.HO OO.H OO.O: NO.N :H O0.0N OO.O ON.ON OO.OH H0.0 NH.NN OO.N NH.NO OO.N OH OO.N: OO.H ON.N: OO.HN ON.O OO.NO HN.H O:.O: O0.0 NH NO.OH O0.0 OO.ON O0.0H O0.0 HH.:O NO.H NN.:O O0.0 HH ON.NN OO.O OO.OO O:.OH O0.0 O0.00 ON.N OO.NO ON.O OH O0.0: NH.H ON.N: OO.OH NO.O NO.OO OO.H OO.O: ON.O O :N.OO NN.H ON.N: OO.HN ON.O OH.HO ON.H OO.OO O0.0 O OO.HN OO.O OO.OO O:.OH OO.O H:.OO NO.N OO.NO OO.N N OO.ON OO.O OO.O: OH.OH OO.O OO.NN OH.N OO.OO NO.N O NO.OH N0.0 OO.OO ON.O :N.O OO.NO NO.N OO.ON OO.N O OO.OH HO.O N0.00 N:.O HO.O ON.OO NO.N NN.ON H0.0 : O0.0H O0.0 OO.NN :O.O O0.0 OO.HN N:.N HH.NN H0.0 O HH.NH O0.0 O0.0N ON.O ON.O O0.0 NN.N MO.mN O0.0 N OO.OH OO.O OO.HO OO.HH N0.0 ON.O H .N N. N OH.O H 2\z.HomaH z .HomeH aoemao Z\z on.HomeH 2\z.HomnH zOHomeH Omz .oa aHmaom deaom OHWHoO One OOO< aoeos pom poems pom mMHH.mH_ mz OHOSOO .meoEHaomNm QOHHOSHO>M thHOSO GHOHOAO CH Ommb OOHQEOO owwpom :N wo OOHpHHHQSHom GHepopm .O OHQOB KHOcon< 223 ON.OH O0.0 O:.OO OO.H :N N0.0H OO.O OO.:O HN.H ON ON.OH OO.O NO.ON NO.O NN HH.NN :N.O ON.OO NO.O HN N.NN NN.O OO.NN OO.O ON N.NN NN.O OO.HO OO.H OH OH.OH O0.0 HO.ON OO.O OH OO.OH OO.O OO.ON OO.H NH OO ON ON.O OO.OO NO.H OH OO.ON OO.O NN.HO HH.H OH OO.ON NO.O OO.NN ON.O OH OH.NN O0.0 O0.00 HO.H OH O0.00 OO.H OO.OO OO.H NH ON.HN OO.O OO.ON HO.O HH H0.00 OO.O HO.NO ON.H OH OO.O: NO.H OO.OO HO.H O OO.:O OH.H OH.NO NO.H O OO.HN HO.O O0.00 OO.O N OO.HO OO.O HO.OO OO.O O HH.NN OO.O OO.NN HO.O O OO.ON NO.O OO.ON OO.O : NO.HH HO.O O0.0H OO.O O OO.OH HO.O OO.O ON.O N OH.OH HO.O OO.O NN.O H z .HomCH z .HomnH .ofl z z.HomcH . z z .HomGH . CHMQOQ+.Edm QHonQ+ £85m .cmm+chQom w QOQ+GHQOm OHmem ApodaHonov .O OHQOB HHOsOQQH (Continued) Appendix Table 6. Browning (ODhuOnm) Rum.NH3 N/N +pepsin+pan. +pepsin+pan. Rum. insol. N insol.N/N Rum. Sample No. 22h rfirNrNfiM-d'MQ—Mr-NNNNV-V—FVNNN OOOOOOOOOOOOOOOOOOOOOOOO Noodr®®ONeom© N Homowdo OHOJONHOJNOOM' O domeoem o o I | o o o o o o mmdmmmmwmmm: m “MO—:h—MNV QOQQOMONJNQOr OO NQPOMQP mmmomOOmOmmdh Om QONJOQN . o o o o o g o 9 0 O I O I o O | O O O O 0 O O mmodowowmwmjw HO NONOOdr mmmmmmmmdmm N mm mmmmmmm (ONNOCDOLAOVOCDOOOODN dun \OQDNrfiONOJi NNQQNrNNdflWfiM)”%OINONQNQN OOOOOK—Ox—x—x—Ov—O v-O Ox—OOOOO c-dungiuvor~d>ow3¢-dnn;iuvorNd>Ow3 QHfl41 x—s—x—s—V—a—x—t—x—NN NNN ‘— 225 11 .HOHQOLOHE mofism .QHpOvoQOm .OHQSHoO .houpme had H SQ some .mem . o H mm 0 .nonHw pcmmmopep OHow H mm< .OHDSHOOGH .thHHQHUmmmHU QQWOQpHC H Q2 O n . omQH H .o n .domoapHQ n 2O HemeeHoeooO ‘1'! .O OHQOB NHOGmmmd 226 Appendix Table 7. Supplied by Dr. H.K. Goering, USDA. Chemical Composition of Forages Source Description DMa CWb ADFC ALd Orchardgrass hay 93.1 66.9 39.9 6.39 Pelleted orchardgrass 93.5 66.3 40.1 6.66 Wafered orchardgrass hay 93.1 69.8 42.0 6.77 Grass hay 91.1 60.4 34.0 3.64 Formic acid silage 22.2 55.1 35.3 3.26 Reconstituted formic acid silage 34.1 65.1 38.3 ’4.91 Orchardgrass low-moisture silage 54.1 66.1 44.9 6.87 Grass silage (silo fired) 43.1 62.3 50.4 15.00 Grass silage (silo fired) 40.5 59.0 56.9 21.32 Oats silage 28.2 47.1 33.8 3.36 Grass silage - 38.8 44.8 19.80 Bermudagrass:corn 92.0 58.8 24.1 3.62 Bermudagrass:corn 92.2 55.1 24.3 3.78 Bermudagrass:corn-autoclaved 92.7 66.3 27.1 6.45 Bermudagrass:corn—unautoclaved 92.6 67.0 28.2 8.57 Orchardgrass silage 73.6 67.0 64.0 11.54 Alfalfa silage 75.3 54.3 44.5 9.26 Low-moisture silage 73.9 65.4 51.4 15.28 High-moisture silage 34.8 43.8 33.7 4.75 Timothy hay 93.1 68.8 41.6 5.59 Timothy(autoc1aved for 30 min) 93.5 72.0 46.5 9.12 Timothy(autoclaved for 60 min) 93.3 70.6 45.6 9.40 Alfalfa silage 28.2 40.9 34.4 8.05 Alfalfa silage 59.8 43.2 34.1 8.10 Alfalfa hay 89.7 36.7 29.1 6.51 Alfalfa silage 28.6 41.0 35.0 8.27 Alfalfa silage 59.8 44-9 37.9 10.21 Alfalfa hay 91.6 38.8 31.5 6.62 Alfalfa hay 73.8 43.6 32.3 7.09 Alfalfa hay 64.8 43.0 33.5 7.59 Alfalfa hay 46.6 50.7 41.5 10.31 Alfalfa hay 41.5 48.8 42.2 10.24 Native hay 80.8 70.1 46.5 7.20 Native hay 65.9 73.9 51.9 9.69 Native hay 56.5 76.3 55.7 11.34 Native hay 49.2 76.4 57.7 12.77 Alfalfa hay 94.3 38.0 31.3 5.92 Alfalfa hay-molded 95.0 48.6 36.8 8.11 Alfalfa hay-molded 94.8 50.7 37.4 7.94 Alfalfa hay 94.0 42.0 33.4 5.62 Alfalfa hay 95.0 41.2 33.9 6.87 Alfalfa hay~molded 95.4 46.2 35.8 8.15 Alfalfa hay ‘ 93.6 48.6 34.7 8.03 Alfalfa hay-molded 94.0 62.2 42.2 7.70 Alfalfa hay 94.7 44.5 34.3 7.58 Alfalfa hay-molded 93.9 61.4 42.9 7.32 Appendix Table 7. 227 (Continued) Source Description Pep.-N AD N N8 Orchardgrass hay 0.357 0.83 2.24 Pelleted orchardgrass 0.32 0.78 2.11 Wafered orchardgrass hay 0.39 0.82 2.04 Grass hay 0.26 0.90 3.32 Formic acid silage 0.06 0.31 3.37 Reconstituted formic acid silage 0.20 1.79 2.36 Orchardgrass low-moisture silage 0.47 1.13 1.39 Grass silage (silo fired) 0.99 1.00 1.80 Grass silage (silo fired) 1.33 1.54 2.34 Oats silage 0.05 0.47 2.78 Grass silage 1.31 1.95 3.59 Bermudagrass:corn 0.18 0.68 2.24 Bermudagrass:corn 0.18 0.68 2.20 Bermudagrass:corn-autoclaved 0.54 1.03 1.69 Bermudagrass:corn—unautoclaved 0.69 0.91 2.03 Orchardgrass silage 1.11 1.21 1.49 Alfalfa silage 0.46 0.49 2.44 Low-moisture silage 1.20 1.42 1.73 High—moisture silage 0.18 0.54 2.63 Timothy hay 0.12 0.34 1.00 Timothy(autoc1aved for 30 min) 0.42 0.49 0.98 Timothy(autoc1aved for 60 min) 0.55 0.65 1 01 Alfalfa silage 0.27 - - Alfalfa silage 0.29 0 78 3.40 Alfalfa hay 0.22 - - Alfalfa silage 0.29 - - Alfalfa silage 0.52 1.48 3.56 Alfalfa hay 0.24 0.62 3.30 Alfalfa hay 0.29 0.63 2.93 Alfalfa hay 0.36 0.62 3.10 Alfalfa hay 0.60 1.35 3.20 Alfalfa hay 0.76 1.56 2.90 Native hay 0.37 0.74 1.31 Native hay 0.61 0.94 1.31 Native hay 0.75 1.14 1.28 Native hay 0.88 1.16 1.38 Alfalfa hay 0.23 0.61 3.77 Alfalfa hay-molded 0.45 1.08 3.96 Alfalfa hay-molded 0.50 1.10 3.41 Alfalfa hay 0.20 0.52 3.22 Alfalfa hay 0.23 0.66 3.49 Alfalfa hay-molded 0.51 1 28 4.18 Alfalfa hay 0.21 - 2.74 Alfalfa hay-molded 0.27 0.64 2.35 Alfalfa hay 0.27 0.65 2.92 Alfalfa hay—molded 0.33 0.75 2.50 228 Appendix Table 7. (Continued) aDM = dry matter. bCW = cell walls. cADF = acid detergent fiber. dAL = apparent lignin. eAD—N = acid detergent insoluble nitrogen. f Pep.-N = pepsin insoluble nitrogen. 8N = total nitrogen. 229 Appendix Table 8. In Vivo Digestion Coefficients of Forages Supplled by Dr. H.K. Goering, USDA. l _ ”1‘ d Source Description pm, Nb Ene? -CW Orchardgrass hay 56.0 54.1 — - Pelleted orchardgrass 55.2 55.6 - - Wafered orchardgrass hay 57.2 52.1 - - Grass hay 66.5 69.0 63.2 76.5 Formic acid silage 74.2 77.5 73.7 85.2 Reconstituted formic acid silage64.9 55,6 _ - Orchardgrass low-moisture silage73,8 35.7 - 8%.6 Grass silage (silo fired) 54.9 13.1 - 4 .8 Grass silage (silo fired) 44.3 9.2 - 50.4 Oats silage 86.1 84.7 - 82.4 Grass silage 38.0 27.0 - - Bermudagrass:corn 73.0 69.0 — - Bermudagrass:corn 70.0 66.0 - - Bermudagrass:corn-autoclaved 65.0 47.0 - - Bermudagrass:corn—unautoclaved 63.0 35.0 - - Orchardgrass silage 55.9 6.29 - 63.3 Alfalfa silage 58.9 52.2 - 56.7 Low-moisture silage 55.0 13.8 - 62.1 High-moisture silage 65.6 59.5 - 60.2 Timothy hay 56.8 45.5 - - Timothy(autoclaved for 30 min) 56.9 37.5 - - Timothy(autoclaved for 60 min) 53.0 29.2 - - Alfalfa silage 59.5 72.8 - - Alfalfa silage 59.0 63.0 — - Alfalfa hay 64.5 77.5 - - Alfalfa silage 60.1 73.6 - - Alfalfa silage 54.0 48.9 - - Alfalfa hay 63.8 76.2 - - Alfalfa hay 60.0 63.0 55.1 55.7 Alfalfa hay 56.% 58.0 50.4 51.2 Alfalfa hay 51. 40.2 47.0 52.2 Alfalfa hay 45.8 27.2 34.1 48.1 Native hay 51.6 39.7 51.2 58.8 Native hay 47.9 20.2 46.5 58.8 Native hay 45.5 10.8 44.4 57.3 Native hay .8 6.6 42.0 56.5 Alfalfa hay 67.4 78.5 66.4 55.5 Alfalfa hay-molded 53.8 54.1 53.9 56.5 Alfalfa hay-molded 55.0 55.5 58.1 59.9 Alfalfa hay 61.5 73.3 61.5 51.4 Alfalfa hay 62.3 77.7 61.4 52.0 Alfalfa hay-molded 52.2 51.8 51.3 52.5 Alfalfa hay 63.7 74.6 61.9 58.9 Alfalfa hay-molded 55.9 61.7 62.8 70.5 Alfalfa hay 65.3 77.4 65.6 56.3 Alfalfa hay-molded 64.7 69.2 65.1 70.8 230 Appendix Table 8. (Continued) . f Source Description ADFe Hemi-C AD-Ng Orchardgrass hay - - - Pelleted orchardgrass - - - Wafered orchardgrass hay - - — Grass hay _ - - Formic acid silage _ - - Reconstituted formic acid silage - - Orchardgrass low-moisture silage 80.6 93.2 68.1 Grass silage (silo fired) 47.8 52.9 22.8 Grass silage (silo fired) 35.6 — -1.9 Oats silage 81.4 85 O 12.5 Grass silage — — 3.3 Bermudagrass:corn - - - Bermudagrass:corn - - - Bermudagrass:corn-autoclaved - - - Bermudagrass:corn-unautoclaved - - - Orchardgrass silage 66.9 -l4.4 5 . Alfalfa silage 56.2 58.8 54.6 Low-moisture silage 54.8 88.8 42.9 High-moisture silage 62.4 52.6 36. Timothy hay - - - Timothy(autoclaved for 30 min) — - - Timothy(autoc1aved for 60 min) - - - Alfalfa silage - - - Alfalfa silage - - - Alfalfa hay - - - Alfalfa silage - - - Alfalfa silage - - - Alfalfa hay - - - Alfalfa hay 47.0 - - Alfalfa hay 47.7 - - Alfalfa hay 53.7 - - Alfalfa hay 49.1 - - Native hay 49.0 — - Native hay 53.0 - - Native hay 48.0 - - Native hay 48.8 - - Alfalfa hay 54.0 62.9 42.9 Alfalfa hay-molded 52.7 68.3 34.2 Alfalfa hay-molded 56.1 70.6 53.6 Alfalfa hay 49.7 58.1 20.0 Alfalfa hay 51.5 54.0 4445 Alfalfa hay-molded 50.3 60.0 34.5 Alfalfa hay 53.5 72.4 - Alfalfa hay-molded 64.2 83.9 34.3 Alfalfa hay 53.3 66.3 - Alfalfa hay-molded 65.5 83.0 48.8 231 Appendix Table 8. (Continued) L _____ J i _( aDM = dry matter. bN = nitrogen. c Ene. = energy. d CW = cell walls. eADF = acid detergent fiber. fHemi-C = Hemi-cellulose. gAD—N = acid detergent insoluble nitrogen. 232 N.NO 0.0: O.ON N.OH 0.0 N.O: 0.00 OOOHNOO OOHOOHO N.HO 0.0N N.O 0.0H O.O 0.00 0.0: OOOHOOH OOHOOHO N.OO O.N: :.NN 0.0H - - 0.00 OOOHHO OHHOOHO 0.00 O.NN 0.0 0.0H - 1 0.00 OOOHHO OOHOOHO O.N: O.O: N.HO H.OH - - N.OO OOOHOOO OHHOOHO N.NO. O.HN O.N :.OH - - N.OO MOOHOOHNOOHOOHO . . . . . . . aoaOO On OOH: N OO O NN O O H NH O O O NO O ON OOOOOaHONOH OOHOOHO 0.00 0.0N O.O N.NH 0.0 O.NO N.ON OOH OOHOOHO O.OO O.NN N.O 0.0H O.O H.OO O.HO H eemamOmHOOHN . . . . . . . to O. O: OOH H NO N OO O OH O OH N H O O: O OO OOHOOOHOOOH OOHOHHO N.OO N.NO O.OH 0.0H 0.0 N.NO 0.00 OOH OOHOHHO N.OO N.ON N.O N.OH N.O 0.00 O.OO OOH OOHOOHO N.NO O.NO N.OH O.OH - N.HO O.ON OOOHOOH OOHOOHO N.OO O.OO N.O N.OH - O.N: O.OO OOOHOOO OOHOOHO O.HO N.HN 0.0 N.OH - O.HO 0.0: OOOHOOH OHHOOHO H.NO 0.00 0.0 O.OH - O.O: O.HN NOH OOHOOHO O.NO O.OO O.O 0.0H - N.NO O.HN OOH OOHOOHO 0.00 0.0N H.O N.OH - O.O: H.OO NOH OOHOHHO .......... O ------------- ------------200--- -- -O- cHepopm Oman oOz O2\z-O< OOnaO oHOO OOOO OzO .AOpOOQGGHz mo .>HGDV GOOQOHO .O.Q .am Odd AGHOGOOOHB Mo .>HGDV concowHOO .<.z .LQ hp OOHHQQSO memwaom %o mpGeHOHmmooo GOHpmomHQ UGO QOHpHOOQEOO .O OHQOB KHOGOOQ< 233 .OHHHHOHHOOOHO HOOOOs OOO u Ozm .OHHHHQHOOoOHO OOOoaHHe n Oz .GmwoapHG HOpop Mo uuooamm O OO ComoanG OHndHomQH pdepmpoo OHoO .anmHH pCOwHOpOO OHoO .HOQHM pcmm90pow OHoO M m H Z\ZIQ