I/IB R.AFLY OF THE UNIVERSITY OF ILLINOIS G30.T Co p . Z AGRICIH.IURE NON CIRCULATING CHECK FOR UNBOUND CIRCULATING COPY UNIVERSITY OF ILLINOIS Agricultural Experiment Station BULLETIN No. 303 THE DIGESTIBILITY AND METABOLIZABLE ENERGY OF SOYBEAN PRODUCTS FOR SHEEP BY T. S. HAMILTON, H. H. MITCHELL, AND W. G. KAMMLADE URBANA, ILLINOIS, MARCH, 1928 CONTENTS PAGE REVIEW OF PREVIOUS WORK 239 Critical Consideration of Results of Previous Investigations 241 INVESTIGATION OF 1923 244 Plan of Investigation 244 Experimental Results Obtained in 1923 245 INVESTIGATION OF 1925 255 Plan of Investigation 255 Experimental Results Obtained in 1925 257 GENERAL DISCUSSION OF EXPERIMENTAL RESULTS 263 The Digestibility of Soybean Products 266 Significance of Refused Feed in Digestion Experiments 268 Significance of Indirectly Calculated Coefficients of Digestibility 271 Metabolizable Energy of Soybean Products Determined for First Time 276 SUMMARY AND CONCLUSIONS 2SO LITERATURE CITED 231 APPENDIX.. . 283 THE DIGESTIBILITY AND METABOLIZABLE ENERGY OF SOYBEAN PRODUCTS FOR SHEEP BY T. S. HAMILTON, H. H. MITCHELL, AND W. G. KAMMLADE* Since the introduction of soybeans into Europe in 1875 there has been an increasing interest in the adaptation of this valuable legume to the European and American systems of agriculture. The plant has been used as a leguminous crop in the corn belt only comparatively recently, but it is rapidly becoming more popular, not only because of its value as a means of maintaining soil fertility, but also because of the high percentage of costly nutrients it contains. With the increased supply of soybeans there have developed new uses for the products obtained from them. Some of these products are rapidly taking the place of more expensive feeds in the rations of farm animals and are being put to new uses in the industries. While the beans from the soybean plant form a standard article of diet for hu- mans in the Orient, none of the various dishes which may be prepared from them have found great favor in this country. However, with the increasing use of soybeans and soybean products for livestock, a de- mand has arisen for information concerning their nutritive value. The investigations reported herein were undertaken, therefore, in order to determine the digestibility and metabolizable energy of soybean hay, soybean straw, whole soybeans, and soybean oil meal. In order to obviate what was thought to be one of the greatest faults with most previous investigations along these lines, that is, the use of too few experimental animals, it was decided to determine the digestibility and metabolizable energy for each feed on each of 12 sheep, a number three times as large as has heretofore been used in any single digestion experiment with soybean products. REVIEW OF PREVIOUS WORK Investigations concerned with the feeding value of the various parts of the soybean plant were reported by Weiske, Dehmel, and Schulze 1 * in 1879. These investigators determined the digestibility of the nutrients in soybean straw and in soybean pods for farm animals. In 1883 the digestibility of soybean straw was again determined by Weiske, Kennepohl, and Schulze, 2 * and in 1886 Kellner 3 * determined the digestibility of soybean seed and of soybean hay. *T. S. HAMILTON, Associate in Animal Nutrition; H. H. MITCHELL, Chief in Animal Nutrition; W. G. KAMMLADE, Assistant Chief in Sheep Husbandry. 239 240 . BULLETIN No. 303 [March, In America there have been comparatively few digestion experi- ments with soybean products. No digestion experiments with soybean oil meal or with soybean straw have been reported. Of three digestion trials on soybean hay, the first was that of Sturtevant 4 * in 1884. In this experiment the hay was fed, combined with corn meal and wheat bran, to two cows during a six-day trial. The feed intake was con- stant for only three of the six days and the feces were collected for one day only. The digestibility of the hay was calculated indirectly from the "nutrients digested from the combined ration and from aver- age coefficients of digestibility for corn meal and wheat bran obtained from German tables. The second American trial was reported in 1894 by Emery and Kilgore, 5 * who fed soybean hay alone to a cow and a goat. The collection period, which followed a suitable preliminary- period, was of five days' duration. The third and last digestion trial with soybean hay was that of Lindsey, Beals, and Smith, 6 * who de- termined the digestibility of the hay when fed with English hay to two sheep. The digestibility of soybean seed has been determined five times with ruminants and once with swine. In all, 5 swine and 17 ruminants (16 sheep and 1 steer) were used, and six different samples of beans were tested. Lindsey, Smith, and Holland 7 * in 1894 determined the digestibility of the nutrients of ground soybeans when fed with English hay, using two sheep. In 1895 the digestibility of soybeans fed with timothy rowen hay was determined in two trials, with four sheep in each, by Phelps and Woods. 8 * The only experiment concerned with the digestibility of soybeans by steers was reported in 1901 by Willard and Clothier. 9 * In this trial soybean meal was fed with kafir corn stover to a steer. Lindsey in 1903 10 * and again in 1904 11 * obtained digestion coefficients on ground soybeans when fed with English hay, using two sheep in each of these studies. The digestibility of soybeans by swine has been determined by Forbes, Beegle, Fritz, and Men- sching. 12 * The soybeans were fed with corn to 5 swine. The digestibility of soybean forage when fed green to sheep has been determined on seven different samples reported in five investiga- tions. Three of these investigations were carried out at the Connecti- cut (Storrs) Station and the other two at the Massachusetts (Hatch) Station. In the Connecticut investigations reported by Phelps and his coworkers, 8 ' 13 - 14 * 12 sheep were used in the determination of the digesti- bility of the green soybean forage, fed alone, at five different stages of growth from early bloom until seeds formed. In the two investiga- tions by Lindsey and his coworkers 11 ' 15 * at the Massachusetts Station, 6 sheep were used and the forage was fed with English hay. The digestibility of soybean silage has been the subject of three investigations. Emery and Kilgore 16 * at the North Carolina Station in 1892, using two goats, were the first to carry out such an experiment. 1928] SOYBEAN PRODUCTS FOR SHEEP 241 In 1895 the digestibility of soybean silage by four steers was studied by Hopkins 17 * at the Illinois Station and in 1903 Mooers 18 * of the Tennessee Station made a similar study using 3 steers. Rations containing soybean products have also been studied in a few cases without any attempt to interpret them with regard to the digestibility of the soybean product alone. In 1895 Phelps and Woods 8 * reported the results of two investigations, in each of which 4 sheep were used. In the first the ration contained .5 pound soybean meal and 1 pound timothy rowen hay, and in the second .75 pound soybean meal and 1.5 pounds hay. In 1896 the Massachusetts (Hatch) Sta- tion 19 * carried out two investigations with sheep. In one of these the di- gestibility of corn and soybean silage in the ratio of 2 to 1 was studied, and in a second, the digestibility of millet and soybean silage. In 1904 Bartlett 20 * reported, among a large number of digestion experiments with sheep and steers, three trials with rations containing soybean silage. In one trial two sheep were used and soybean and corn silage (9:14) was fed alone, one sheep receiving 3,000 grams daily and the other 2,000 grams. In a second trial the same silage was fed to a steer, and in a third trial the silage was fed with some hay to 2 steers. The details of all the above mentioned investigations will be found in the Appendix, Table 22, which gives a compilation of all American digestion experiments on soybean products to date, including those reported in this bulletin. No reports of investigations on the metabolizable energy of soy- bean products have been found in the literature. Critical Consideration of Results of Previous Investigations With the possible exception of those experiments in which green soybean fodder and soybean silage were used, most of these experi- ments are not satisfactory. With what is perhaps the most important soybean product, viz., soybean oil meal, no study has been made. A brief examination of almost any of the trials in which two or more animals were used will show quite wide variations among the coefficients obtained, this being especially the case when the coeffi- cients have been calculated indirectly, i.e., from those obtained with a mixed ration. Such results indicate the necessity for the use of a larger number of experimental animals and longer collection periods. Collection periods of 5 or 6 days are too short, especially for steers. Naturally, the coefficients obtained for the same feed by different in- vestigators or by the same investigator at different times will vary no less than do those of any one trial. Average coefficients compiled from these data are to be found in several standard publications and are widely used. The value of such averages in the planning of rations cannot be considered satisfactory in view of the objections mentioned, while the application of coefficients obtained from one species of 242 BULLETIN No. 303 [March, ruminant to another has not hitherto been shown to be justifiable. One example will be considered in detail to illustrate the questionable value of such average results. A well-known American compilation of the digestibility of farm feeds 21 * gives the following coefficients for soybean oil meal for rumi- nants: crude protein 92 percent, nitrogen-free extract 100 percent, ether extract 68 percent, and crude fiber 99 percent. These values are given as the averages of two digestion trials and are taken from Mentzel and Lengerke's "Landwirtschaftliche Kalender," 22 * which is itself a compilation. While no literature references are given in this German publication, the original investigation from which these values were obtained may be identified as one which Honcamp 23 * car- ried out in 1909 at the Rostock Agricultural Experiment Station. After first determining the digestibility of a basal feed of clover hay for 2 wethers, this investigator fed each of the wethers a mixed ration of 760 grams of clover hay and 240 grams of "pressed" soybean meal in one period. In a second period one of the two wethers received a ration of 600 grams of clover hay and 200 grams of a sample of "extracted" soybean meal, and in a third period a third wether was fed 700 grams of clover hay and 200 grams of a second sample of "extracted" soybean meal. The results of this investigation are tabu- lated in Table 1. TABLE 1. COEFFICIENTS OF DIGESTIBILITY OF "PRESSED" AND "EXTRACTED' SOYBEAN MEAL (HONCAMP M *) "Pressed" soybean meal "Extracted" soybean meal Period I Period II Period III Average Wether 1 Wether 2 Average Wether 2 Wether 4 Organic matter 90.3 91.8 97.9 93.6 16.8 96.8 93.8 107 . 5 92.0 55.4 93.6 92.8 102.7 92.8 36.1 96.9 91.9 112.9 29.6 68.3 96.4 91.8 98.0 105.5 130.4 96.6 91.9 105.4 67.6 99.3 Crude protein Crude fat Crude fiber The values above quoted are obviously the averages, in whole numbers, of Honcamp's coefficients for the "extracted" soybean meal, coefficients above 100 being quoted as 100. The average coefficients for crude fat and crude fiber, i.e., 68 and 99 respectively, are of little significance since each is the average of two widely varying figures, 68 being the average of 29.6 and 105.5, and 99 being the average of 68.3 and 130.4. The average coefficient for the nitrogen-free extract is 105.4, the individual results being 112.9 and 98.0. Digestion coefficients above 100 result from the fact that the admixture of nitrogenous con- centrates, such as soybean oil meal, often increases the digestibility of the nitrogen-free extract of the basal ration, so that the apparent digestibility of this nutrient as computed indirectly for the supple- 1928} SOYBEAN PRODUCTS FOR SHEEP 243 mentary feed becomes over 100 percent. Such an impossible result is thus due to the fallacy upon which the method of computation is based, and cannot be used with confidence in computing the digestible nutrients in any other combination of feeds. The substitution of a co- efficient of 100 for a larger value thus indirectly computed cannot be considered a solution of the difficulty, since no evidence can be offered that lower coefficients may not be nearer the truth. The coefficients for crude protein only are in good agreement and, fortunately, crude protein is not only the most abundant but also the most important nutrient in soybean oil meal. The excellent checks for the coefficients of digestibility of organic matter are a fortunate coincidence only. There can be little criticism of the length of collection periods (10 days) in this experiment, but if the average coefficients obtained are to be used in the computation of rations for other ruminants than the sheep, the possibility of other errors must be considered. Just how great an error is likely to be incurred by using for steers or dairy cows digestion coefficients obtained with sheep is impossible to estimate at the present time. Probably this error would not be great, since in digestion experiments there often are found as great variations between different sheep as between sheep and steers. However, as Bartlett 20 * says (page 204), after a- study of the digestibility of a variety of feeds by both sheep and steers, "If sheep are to be used to determine coeffi- cients for bovines, great care should be taken to select strong animals that are good feeders and will eat coarse fodders readily, otherwise results which are too low are likely to be obtained." Finally, attention should be called to the fact that the digestion coefficients of Honcamp's experiment were obtained on two samples of specially extracted soybean meal. Honcamp's analyses show 1.40 per- cent crude fat in one sample and 1.80 percent in the second sample, an average of 1.60 percent crude fat. The average percentage of crude fat in 11 samples of extracted soybean oil meal analyzed in this laboratory- is 4.64. The average of four analyses of pressed soybean oil meal is 7.31 percent crude fat, making an average of 5.35 percent crude fat in soybean oil meal purchased in Illinois. Henry and Morrison give 6.6 percent crude fat as an average for soybean oil meal, and Lindsey and associates 24 * quote 8.6 percent fat in a sample of soybean oil cake. It is therefore obvious that the two German samples of soybean oil meal were considerably different in fat content from any obtainable in America, and it is quite probable that their digestibility would not be the same as the digestibility of soybean oil meal obtainable in this country. In confirmation of this belief, the differences in the digesti- bility of pressed and extracted soybean oil meal as determined in Honcamp's experiment may be cited (Table 1). 244 BULLETIN No. 303 [March, INVESTIGATION OF 1923 Plan of Investigation Twelve western lambs weighing from 80 to 100 pounds were used- as experimental animals in the first experiment with sheep in 1923. With each sheep the digestibility and metabolizable energy of the fol- lowing five feeds or rations, corresponding to the five successive periods of the experiment, were determined: (1) soybean hay, (2) soybean straw, (3) oat straw and whole soybeans, (4) oat straw, and (5) soy- bean straw and soybean oil meal. By making use of the digestion co- efficients and the metabolizable energy values for oat straw and for soy- bean straw obtained in Periods IV and II respectively, the digestibility and metabolizable energy of whole soybeans and of soybean oil meal were calculated in the usual indirect manner from the data obtained in Periods III and V. The digestion trial for each feed or combination of feeds consisted of a 10-day preliminary period, during which the sheep received the exact ration which they were later to receive during the collection period, and a period of 8 days in the metabolism crates during the last 7 days of which the collections were made. The equipment for making the collections consisted of three Forbes' metabolism crates for swine. 25 * The three sheep grouped together in the tables were in the metabolism crates at the same time. In Period I soybean hay was fed in the following amounts daily: 910 grams to Sheep 130, 131, and 132; 1,025 grams to Sheep 133, 134, and 135; 1,135 grams to Sheep 137, 138, and 139; and the same to Sheep 136, 140, and 141. In Period II each of the 12 sheep received 1,135 grams of soybean straw. In Period III Sheep 130 to 133 inclu- sive received 910 grams and Sheep 134 and 135, 811 grams of oat straw, together with 230 grams of whole soybeans; and Sheep 136 to 141 in- clusive each received 680 grams of oat straw and 225 grams of whole soybeans. In Period IV each of the 12 sheep received 910 grams of oat straw, with the exception of Sheep 135, which received 779 grams. In Period V each of the 12 sheep was fed 910 grams of soybean straw and 225 grams of soybean oil meal. The sheep were fed twice daily: in the morning immediately after the collections were made and late in the afternoon. During the period in which whole soybeans were fed, the beans were given first and the sheep allowed to clean these up completely before the oat straw was fed. In the period in which soybean straw and soybean oil meal were fed, the meal was sprinkled over the straw in an attempt to induce the sheep to eat the straw more readily. The same purchase of Midwest soybeans and of extracted soy- bean meal served thruout the entire experiment. A sufficient amount of each of the roughages was secured for the entire experiment, but no attempt was made to mix the entire supply of either. The roughages 1928} SOYBEAN PRODUCTS FOR SHEEP 245 were weighed to the nearest 10 grams and the concentrates to the nearest 5 grams. Distilled water was kept before the sheep at all times. Salt was available only during the preliminary periods. Sam- ples of feed for analysis were obtained by removing a small quantity of each feed at each weighing. Collections of excreta were made daily in the morning. The method of collection was very similar to that described by Forbes for the management of these crates. 25 * The feces were not aliquoted, each daily excretion being added to tightly covered 50-pound lard cans until the end of the collection period, when they were ground, mixed, and sampled for analysis. The feces were ground in the fresh con- dition by putting them once thru a revolving-hammer type of mill with a large-sized screen. This method of preparation completely disintegrated the feces without any grinding or heating effect, so that they could be satisfactorily analyzed in the fresh condition. The daily collections of urine were aliquoted and the aliquots composited for the period. All samples of feces and urine were preserved by refrigeration until analyzed. Refused feed was removed daily, the daily orts com- bined for the period, dried, and analyzed. In addition to the usual routine determinations, which were made in accordance with the official methods of the A. O. A. C., the gross energy in all samples was determined by the bomb calorimeter. In making the energy determinations on urine, 100 cc. portions were evaporated at low temperatures under a current of warm air, absorbed on cellulose blocks, and ignited with the cellulose. In this determina- tion no correction was made for material lost on drying. Experimental Results Obtained in 1923 Digestibility of Soybean Products. In an investigation such as this an enormous amount of data is obtained, but for the sake of economy of space much of it will be omitted or given only in sum- marized form. The average percentage composition and gross energy of the rations offered, of the orts, of the rations consumed, and of the feces of each period are given in Table 2. The digestion coefficients of the nutrients consumed in each period were calculated in the usual manner and these are tabulated for each sheep and averaged for each period in Table 3. From the data obtained on the ration of whole soybeans and oat straw (Period III) and on the ration of oat straw alone (Period IV), the coefficients of digestibility of the nutrients of the whole soybeans were calculated in the usual indirect manner. During this period the soybeans were fed first and the sheep invariably ate their allotment of beans completely before the straw was given. Hence in making the calculations it is considered that the nutrients found in the orts from this period originated from the straw alone. 246 BULLETIN No. 303 [March, ATION8 b fOOCMOO co C5 o co if* CO ^^ OC '-*5 30 OC5OCO OOINOOOO H 8 tl 6 ,j! CO CO CON CO CO CON C3 ^ ^ CO ^ W CO <" CO * CO CM i O "26 . -CO 5 :8 ICO -CM --H ~ " r - aot, - U5 . ' -CO O fe fc 00 O ll _^ H hH 11 . Ol CO PO 00 M co IN ^-"00 "3 g o 3 I 1 t*CDOOONO> i'g *g aass 03 E ~to SoSSS o O *c3co5cOCM 03 ta (N IN CM CO o 03 O Cl O3 CO O5 O ^ CO CO CO CO F^ o u X 0) O d'S 5 1 iO ^ ^. 00 00 1^ 00 1 r~t i 1 CD o9 co 10 w r. r /.-. > t^OOINO 2 B o SooooS.S V ^^2-0 M I co -< i-* Period I " M "" 1 CO 00 O * i-l CO CO ^H r-l OO o o fcl PL, E PL, 09 J "J" Tjl us o CO O> ^ CD 00 CO 9 o J3 1 OSITIOK fSUMED W| j^OO O O (^ O O O OObOOTl< Period I (N 00 i IN -H --H -./---. 0500510 I s 1 i-H 00 1-1 N CM >O O OC C3 C5 O TJ* SO W O TABLE 2. AVERAGE PERCI Soybean hay , Orts Ration consumed Feces Soybean straw Orts Ration consumed Feces Oat straw Whole soybeans Ration offered Orts Ration consumed Feces Oat straw Orts Ration consumed Feces Soybean straw Soybean oil meal Ration offered Orts Ration consumed Feces 1928} SOYBEAN PRODUCTS FOR SHEEP 247 Similarly, the coefficients of digestibility of the nutrients of soy- bean oil meal were computed from the data obtained on the ration of soybean oil meal and soybean straw (Period V) and on the ration of soybean straw alone (Period II). In this period the soybean oil meal was sprinkled over the soybean straw in an attempt to induce a greater consumption of straw. This procedure did not obviate the TABLE 3. SUMMARY OF DIGESTION COEFFICIENTS OF RATIONS CONSUMED: INVESTIGATION OF 1923 Sheep No. Dry matter Crude protein N-free extract Ether extract Crude fiber Period I Soybean hay, 1,051 grams 130. . . 52 69 62 67 33 131 56 69 67 69 34 132 58 73 65 63 41 133. . . 56 68 65 51 42 134 52 65 64 39 36 135 55 70 67 52 35 137 49 64 62 63 33 138 55 69 66 63 39 139 54 68 67 69 33 136 53 70 63 67 24 140 54 70 61 69 32 141 56 73 61 71 35 Average 54.1 69.0 64.1 61.9 34.7 Period II Soybpan straw, 1,135 grams 130. . . 41 9 48 -4 43 131 42 -7 52 15 43 132 39 18 44 12 41 133... 41 26 58 14 25 134 42 16 55 47 33 135 45 30 61 25 30 136. . . 36 12 51 3 27 137 41 20 56 13 30 138 40 11 54 21 32 139 40 10 54 3 31 140 38 5 56 26 141 40 24 57 26 21 Average 40.4 14.5 53.8 14.6 . 31.8 Period III Oat straw, 779 grams, and whole soybeans, 228 grams 130. . . 46 66 48 63 38 131 47 70 46 67 40 132 45 63 46 64 35 133. . . 47 69 37 63 49 134 42 69 28 70 42 135 47 73 40 58 41 136 .. 46 73 41 ' 72 38 137 48 72 45 59 42 138 49 76 41 65 40 139 ; 51 73 44 63 50 140 54 74 50 77 48 141 53 73 46 79 51 Average 47.9 70.9 42.7 66.7 42.8 248 BULLETIN No. 303 TABLE 3. Concluded [March, Sheep No. Dry matter Crude protein N-free extract Ether extract Crude fiber Period IV Oat straw, 900 grams 130. . , 36 47 34 11 49 131 43 18 35 26 60 132 41 19 41 32 50 133. . . 48 14 49 31 56 134 48 11 45 34 61 135 49 25 48 35 56 136. . . 53 -3 53 26 65 137 50- 6 46 21 68 138 56 10 54 37 67 139 45 17 45 15 55 140 39 13 42 9 47 141 46 6 49 3 53 Average 46.2 14.3 45.1 23.3 57.3 Period V Soybean straw, 910 grams, and soybean oil meal, 225 grams 130. . . 53 66 63 51 38 131 54 68 60 57 43 132 56 70 61 60 46 133 55 69 65 71 41 134 54 66 66 42 40 135 55 68 69 47 36 136 57 66 67 63 44 137 59 70 71 46 46 138 59 69 66 69 50 139 52 68 58 68 41 140 55 70 61 70 44 141 56 70 62 65 46 Average 55.4 68.3 64.1 59.1 42.9 possibility that the nutrients in the orts were derived both from the soybean oil meal and from the soybean straw, but from the similarity in appearance of the orts to the coarser portions of the straw, and also from the similarity in composition, it is assumed that all of the nutrients in the orts are derived from the straw. The individual coeffi- cients of digestibility obtained from the data on each sheep and the averages for whole soybeans and for soybean oil meal are given in Table 4. Metabolizable Energy of Soybean Products. The metabolizable energy of each ration and of each feed was calculated in the usual manner, i.e., by subtracting from the gross energy of the feed con- sumed the gross energy of the solid, liquid (corrected to nitrogen equilibrium), and gaseous excreta. The gross energy of the feeds and of the solid and liquid excreta was determined directly in a calori- metric bomb. It may safely be assumed that all of the energy lost in the gaseous excreta of ruminants is in the methane produced by the fermentation of carbohydrates. According to Armsby 26 * steers pro- duce 4.5 grams of methane for each 100 grams of digestible car- 1928} SOYBEAN PRODUCTS FOR SHEEP 249 TABLE 4. SUMMARY OF DIGESTION COEFFICIENTS OF WHOLE SOYBEANS AND OF SOYBEAN OIL MEAL, CALCULATED INDIRECTLY FROM PERIODS III AND V: INVESTIGATION OF 1923 Sheep No. Dry matter Crude protein N-free extract Ether extract Crude fiber Period III Whole Soybeans 130. . . 48 85 68 82 247 131 51 88 54 86 215 132 40 82 53 83 307 133... 49 90 16" 81 58 134 32 84 62 a 87 113 135 49 89 17 70 88 136... 47 94 20 92 166 137 52 92 43 73 110 138 54 97 26 81 101 139... 63 94 39 77 40 140 75 94 74 96 61 141 74 93 53 100 33 Average 52.8 90.1 44.7 84.0 -128.3 Period V Soybean oil meal 130. . . 102 86 109 80 178 131 107 88 90 92 323 132 110 89 94 92 326 133... 115 89 124 118 258 134 111 85 126 64 247 135 110 87 137 72 132 136. .. 120 81 127 97 393 137 132 86 149 67 436 138 130 84 124 107 534 J39 101 92 77 117 285 140 109 93 89 120 333 141 121 96 96 112 431 Average 114.0 88.0 111.8 94.8 323.0 Omitted from average. bohydrates in the ration and each gram of methane has a calorific value of 13.34 calories. The energy lost in the gaseous excreta was calculated using these factors. Unless the animal is neither gaining nor losing body protein, the difference between the energy consumed and the energy lost in the solid, liquid, and gaseous excreta does not indicate the true metaboliz- able energy value of the feed. A correction for the gain or loss of body protein was made according to the factor suggested by Rubner. 27 * For each gram of urinary nitrogen derived from the catabolism of body proteins (equal to negative nitrogen balance) 7.45 calories are sub- tracted from the urinary energy, and for each gram of nitrogen stored in the body (equal to the positive nitrogen balance) 7.45 calories are added to the urinary energy. The nitrogen balance of each animal in each period is given in the Appendix, Table 20. The data concerning the gross energy, of feed and excreta, and the metabolizable energy of each of the rations fed during the five periods, 250 BULLETIN No. 303 [March, TABLE 5. THE GROSS ENERGY, ITS LOSSES, ITS PERCENTAGE DISTRIBUTION, AND THE METABOLIZABLE ENERGY OF THE RATIONS FED: INVESTIGATION OF 1923 Metabolizable energy ti_3l Period I Soybean hay co Jj (fl-g "S S I* Percentage losses e S m 1-1 s -! I "S a *S i 3 o. E c W sl Hi N ' I 3 -1 | Dry matter eaten per day and head I 1 ll S 6 jjtococo Soo Sto 'C I - 'T 8CNCO C35 O CO Tf cocoa 83S CO ^ CO C.t-0, s 4: co co co CO CO CO CO CO CO cococo co lOCOCO cococo COtNCN CO CO CO CO gcscoo toco t~ o> i oogco Sgg g to-*o *IOGO S5K S38 $$3 t- .< co co co CO CO CO CO-* CO co co co co to co co cococo COCNCN co co co co : ooo oo 2S COOCN to toco t- .ocg co o to oot-t- ... CCCN-* tO COO t- a. coco t- - - - 55? CO t CO CO COCOrf" * *coco co co co . t CN CNCNCN O5O COCNCO 8 "^ CO 00 Oi ^ CNcSoO co o; co o CN Tf tO 00 gcogj to coo O to CO 00 CO ooiom co CN *iOtO tO to tO *iom >OtO to "0 ^.O! lO oo-* oo cot- oo 1 CN to to OCO-* co-cSS CNt-CJ 22 C5 *>OiO #n*> 0-xJ"* rftoto * '"' "^ _* tOCO >OCO CO 28 O ooo oocoS toeo to cc co CO CO CO at^ei- CO X CO O CO O ss * owo O5OCO tOtO CN -*CO ^H 1 im 9QDOB OCOO 00 O i i to C C35 to tot oo a CNCNO too to c; c^i c^ S;O ?i SCO oo to COCNCN 1 -""" - i CN ^^ lilS CNt-tO * III III 1 e e a 000 c c a ooo OOO __ c c a ooo I jriBI ss N t-OO 000000 f- IIS t-CNOO t-t-CO COIOC35 SCO CO t- CO *xs CO> **"* O IS _ 1 ll -t^> o3 "o ^ & T3 C V iS-2 C3 T ^H ^ !_ 2 tt Q to H a 3 a < H d a S l o III Oal "o T3 a i 1 o O S f^ _o o, >> ai i HH g g w 3 o H CT3 c /. 11 go *J g.3 " g O " 51 O"" II o a o o ^^ ecOOiO I .": CD'J'Tjl CNCOCO WO CO Tjl 10O fflCO'-H .-I-*-* >o CO -e 01 CO 01 *-*iO COM-* CO CO CO CO eoi ono SOTHOJ (NiOO t^coco COOt^ OOCO t^-*o> loosoo 00 <: co co co Tjurj co co-*-* CO CO CO * iO^in 0005CN ccroco *^cc t~iO(M IN io>O * Oi **** MtNCO co cc co *** CO ..Ht^OI gOO 00 01 ot- HM S83 ooco^-i OOCOIM * O5 ^COO)-* *MCN CO TjH T}< *coeo CO -o 2-*oco Hcooo I. COU5-* 00"O>O >-<*> oot^ t^iOco t^t i .-i COO oo -p-> t t t CNCNCN '-Ir-'tN .-lIM'-l 1-1 |-*MO go^ooo CO 00-* t^.COiO SIOOO r^ >o o; O O CM oo |t^3"O j; CON oo TOTfcO oo-* m 00 (N "O COO5O2 t~ooco I-l * t^ IN 00 rS ' ' 1S22 t. CO IO CO O5CCO1 O5 -< O ^Ht> iO-*iO CO h-O5 lOTjIl-H ooo --I t^O COCO CO CO SINININ CNINrH IM (N O4 (NCMIN IN 2o-* 5 t~c:oo h rococo COW-* t~"* co>no5 CV COcDt^ t^00t~ *co-* *** O 00 iO "5 1 O--0 co co co CO CO CO co co co co co co co CO CMi-< COCO>O O5CO-* (NiOi-H oo to-* toot~ CO "5 00 00 0500 to co co co co co co co co co CO CO CO CO CO COO O2OCN OOOO COCN^H *OOO OtO-* Ot^lN t^>OlN IO t- 00-* IN IN CO CO "5 00-* co co co COO51O *CO-* r-cooo co co co c OCOCD t~-*OiO CO CO CD to tot>. irj-*iO IO INO3CO t^CC-* oo--ir- --* 10 IN 1OCN CO IN CO r~ oo O co * *oo*-i coiOco CO CO CO iO>OiO 00>-liO *iO-* "OO5-* 1OOK5 >o >o >-l-*lN coot- IN-*CO COiOO lot^oo *>oco OOiOiN oo coo oot^o COO"O "Ooooo to-* tD * t^- 1O OCO-* OCO(N IN IN IN ooo>o IO O >O IN - t-CtO O--IO OOO500 IO O5 IO ooo o t^oco *^HO (>.>OcO >-H iNiO -IOOS COIN CM -*t~O -!* IN .-UNO * CO * * t^. CO CNIN(N IN IN IN (NC4N OXNOI 01 t-H O i ' iCCOcO (NlNlN CDCOt^. '-icqco i-i>-iO CO>-l- O5COO3 -*iOO r-.t-.io h- tOCO ^OlT)( to co to *** O"5 O5 >OiO CO to ^ grt-oo CO CO CO O5O-H CO'*"* < 252 BULLETIN No. 303 [March, M .23 3|.I C^*(NOi OI^-iN CO^*O COOOO Ol j^OOdC i-iT}<O COCOO CCCOCO 1OIOCD CO j < < I 1 . ^,, ass ggj. 22; oo j ! p "3 R IN 1 g*CO(N INO^O OCOIN ^*INI^ CO rOO^OS TTCOiO CNCOIN COCOO c S ,88i 888 SS8 SSS S a O 03 >> 2? II .COOO t--*-* OSOO t^CC--l t~ CsOO '^ O O O CO t^ t CO 1-H Tj* I t t*- P o. O a M | S ) r 4 gggg gsg g5 1 1928] SOYBEAN PRODUCTS FOR SHEEP 253 are tabulated in Table 5. The metabolizable energy of each ration is expressed in therms per kilogram of dry matter and per kilogram of digestible organic matter, according to the method followed by Armsby, and also per kilogram of total digestible nutrients, in the calculation of which allowance is made for the superior energy content of fat. The metabolizable energy of the whole soybeans and of the soy- bean oil meal is calculated by an indirect method exactly analogous to the indirect method for the calculation of the coefficients of digesti- bility of these concentrates. The calculated metabolizable energy con- tent of these two concentrates for each sheep is given in Table 6 and the average metabolizable energy of each ration and each feed alone is given in Table 7. TABLE 6. METABOLIZABLE ENERGY OP WHOLE SOYBEANS AND OF SOYBEAN OIL MEAL*: INVESTIGATION OF 1923 Sheep No. Whole soybeans (fed with oat straw) Soybean oil meal (fed with soybean straw) Per kg. dry matter Per kg. digestible organic matter Per kg. total digestible nutrients'" Per kg. dry matter Per kg. digestible organic matter Per kg. total digestible nutrients'" 130... therms .70 1.52 .82 3.73 3.44 5.51" 2.37 2.61 3.51 2.16 3.23 3.05 2.47 2.82 thermx 1.36 2.88 1.92 8.17 12.28 11.38' 4.59 4.60 6.19 3.46 4.40 4.32 4.19 4.60 therms 1.02 2.14 1.36 5.84 7.22 8.59" 3.21 3.52 4.61 2.65 3.33 3.21 3.09 3.42 therms 3.44 3.65 3.88 2.91 3.61 3.44 4.47 5.26 6.04" 3.59 3.75 4.40 3.71 therms 3.67 3.69 3.85 2.68 3.48 3.40 3.94 4.20 4.S9" 3.79 3.65 3.90 3.66 thermx 3.47 3.48 3.63 2.50 3.34 3.25 3.73 4.06 4.63= 3.50 3 39 3.66 3.46 131 132 133... 134 135 136. .. 137 138 139 140 141 Aver, (all) Aver, last 6 sheep The metabolizable energy of the roughage in the ration is calculated by multiplying the kilo- grams of total digestible nutrients in the roughage consumed, by the average therms of metabolizable energy per kilogram of total digestible nutrients in that roughage as found in a previous trial on the roughage alone, namely, Period IV for oat straw and Period II for soybean straw. b Total digestible nutrients are equal to the sum of digestible crude protein, digestible carbo- hydrates, and 2.25 times digestible ether extract. c These values are higher than the values for corresponding gross energy per unit of weight indicated and are therefore omitted from the averages. Discussion. The digestion coefficients obtained for each ration are fairly concordant, as may be seen by reference to Table 3. The variations among individual coefficients are, in general, small for those nutrients present in the ration in considerable concentration. As was to be expected, considerable variations were obtained with reference to the nutrients present in only small amounts. This is especially notice- able in case of the protein and ether extract of soybean straw and of oat straw; the soybean straw contained but 4.0 percent of protein and 254 BULLETIN No. 303 [March, TABLE 7. SUMMARY OP AVERAGE METABOLIZABLE ENERGY VALUES OF THE FEEDS AND RATIONS FED: INVESTIGATION OF 1923 Feed or ration Number aver- aged Per kg. dry matter Per kg. digestible organic matter Per kg. total digestible nutrients Soybean hay, 1,051 grams 12 therms 1.90 therms 3.73 therms 3.49 12 1.40 3.57 3.55 Oat straw, 779 grams, and whole soybeans, 228 12 1.89 4.18 3.65 Oat straw, 900 grams 12 1.57 3.61 3.57 Soybean straw, 910 grams, and soybean oil meal, 225 grams 12 1.92 3.69 3.59 Whole soybeans, 228 grams (fed with oat straw, 779 grams) 6 2.82 4.60 3.42 Soybean oil meal, 225 grams (fed with soybean straw, 910 grams) 12 3.71 3.66 3.46 1.0 percent of fat and the oat straw 4.8 percent of protein and 2.1 percent of fat. While, in general, fairly concordant results were obtained for the digestion coefficients of the rations as fed, quite the opposite was the case when coefficients of digestibility of whole soybeans and of soybean oil meal w r ere calculated in the usual indirect manner, as may be seen from Table 4. The digestion coefficients computed for whole soybeans vary for dry substance from 32 to 75 ; for crude protein, from 82 to 97; for nitrogen- free extract, from 62 to 74; for ether extract, from 70 to 100; and for crude fiber, from 307 to 33. For soybean oil meal, the digestion coefficients vary for dry substance from 101 to 132 ; for crude protein from 81 to 96; for nitrogen-free extract, from 77 to 149; for ether extract, from 64 to 120; and for crude fiber, from 132 to 534. These variable and frequently impossible results indicate that the assumptions upon which the indirect method of calculation is based are in considerable error. Variations in the coefficients obtained in a digestion trial are due chiefly to three causes: variations in the digestive capacity of the ani- mal, variations in the character of the feed, and technical errors of the test itself. The technical errors of the test may be due to avoid- able causes, such as errors in the taking of weights or in the analysis of feeds and excreta, or to unavoidable causes, such as irregularities in the voiding of excreta, the presence of metabolic products in the feces, and a variable consumption of feed by the experimental animals. In this investigation variations in results due to variable per- formance of the animals were reduced to a minimum by the use of 12 experimental animals. There should be only inconsiderable variations traceable to the character of the feeds, since the total supply of each kind w r as obtained at the same time, from the same place. As to varia- tions caused by technical errors, the effect of irregular excretion w r as probably not large, since a collection period of 7 days, following a 10- day preliminary period, was used thruout. There is no apparent reason 1928] SOYBEAN PRODUCTS FOR SHEEP 255 for suspecting the accuracy of the weights and the chemical analyses obtained. There remain, therefore, two possible sources of error, caused by the presence of metabolic products in the feces and by the irregular consumption of the experimental rations. The former error is inherent in all digestion trials and is unavoidable until some satisfactory method is devised for the separation of food residues from so-called metabolic products in the feces. The effect of these products upon the coefficients of digestibility computed by the ordinary method is to underestimate the true digestibility. Probably the most serious source of error in this experiment was the rather large percentage of refused feed in practically every period. Of the total dry substance offered, the following average percentages were refused: soybean hay 13.9 percent, soybean straw 31.1 percent, oat straw and whole soybeans 22.3 percent, oat straw alone 22.2 per- cent, and soybean straw and soybean oil meal 6.9 percent. The partial refusal of feed by experimental animals is unavoid- able in so far as it is due to the inherent character of the feed rather than to mere fickleness in the appetites of the animals or to the use of inferior samples of feed. It is unfortunately true that soybean rough- ages of good grade are rarely consumed completely when offered to farm animals, because of the woody stems always present. The sam- ples of roughages used in this experiment were of this character and it is questionable whether a repetition of the experiment on the rough- ages alone would have given any more satisfactory results. A study of Table 2 shows that the feed consumed did not differ markedly in com- position from the feed offered, but nevertheless the use of the digestion coefficients in connection with weights of roughages offered to animals would give an exaggerated estimate of the digestible nutrients con- sumed. In so far as the orts are actually inedible, the nutrients con- tained in them are in the same category, for all practical purposes, as the nutrients contained in the feces. With the soybean concentrates, the case is different. The variable and frequently impossible coefficients calculated indirectly for these feeds in the foregoing experiments evidently are of no practical sig- nificance. Digestion experiments were therefore again undertaken with whole soybeans and soybean oil meal in 1925, using alfalfa hay as the basal ration, instead of the less palatable oat straw and soybean roughages. INVESTIGATION OF 1925 Plan of Investigation The 1925 investigation was concerned with the redetermination of the digestibility and metabolizable energy of whole soybeans and of soybean oil meal. Six sheep were used thruout this part of the investi- gation and the collection periods were lengthened from 7 to 10 days. 256 BULLETIN No. 303 [March, The rations and their order of sequence were as follows: in Period VI, 800 grams of alfalfa hay; in Period VII, 800 grams of alfalfa hay and 225 grams of whole soybeans; in Period VIII, 800 grams of alfalfa hay and 112 grams of soybean oil meal; in Period IX, 800 grams of alfalfa hay and 340 grams of soybean oil meal ; and in Period X, 450 grams of soybean oil meal alone. Each sheep received 10 grams of salt daily. With the exception of Period X, data were collected for each period with each of the six sheep. In Period X, only Sheep R3 consumed the 450 grams of soybean oil meal daily for the necessary 20 days. An attempt was made also to feed whole soybeans alone, but this was unsuccessful, due to the scouring effect after the first few days on this ration. Considering alfalfa hay as the basal ration, the digestibility and the metabolizable energy of the whole soybeans and of the soybean oil meal, the latter at two different levels, were calculated in the usual indirect manner. Considering 800 grams of alfalfa hay and 112 grams of soybean oil meal, the ration fed in Period VIII, as the basal ration in Period IX, the digestibility and the metabolizable energy of the additional 228 grams of soybean oil meal fed in the latter period w r as computed. This was done with the thought that any associative ef- fects of combining soybean oil meal with alfalfa might be partially or wholly overcome by the 112 grams of soybean oil meal, and there- fore the computed digestibility of the nutrients and the metabolizable energy of the additional 228 grams might be more nearly representa- tive of the true values of this concentrate when fed alone. Finally, these values obtained by the indirect method were compared with those directly obtained for one sheep when the concentrate was fed alone in Period X. The digestion trials of the 1925 investigation were managed in a manner very similar to those of the 1923 investigation. The errors in- curred in the latter investigation by the comparatively large amounts of refused feeds were completely obviated in the second investigation by feeding such amounts of a very palatable feed as would be readily consumed and at the same time would adequately cover the main- tenance requirements of the sheep. All sheep either maintained their weights or gained slightly during each trial and with one exception (Sheep R3 in Period VIII had a negative nitrogen balance of .15 gram daily) all sheep were in positive nitrogen balance in all periods. There was not more than a handful of refused feed from any one sheep at the end of any period. The sheep used in this part of the investigation were Western yearling wethers varying in weight from 80 to 110 pounds, and were again run in groups of three. Two changes were made in each of the metabolism crates during this year. The cloth on the lower screen for recovering the feces was replaced by copper screening permanently 1928} SOYBEAN PRODUCTS FOR SHEEP 257 attached to a metal frame, and windows were placed on both sides of the crate instead of on one only, thus permitting the sheep to see sheep in adjoining crates. Both windows were fitted with %-inch iron bars placed vertically 3 inches apart. After a year's use of the modified crates, the authors were satisfied that the changes were well worth while. Experimental Results Obtained in 1925 Digestibility. The calculation of the coefficients of digestibility of the nutrients in each of the five rations fed was made in the usual direct manner and, considering the alfalfa hay as the basal ration in Periods VII, VIII, and IX, the digestion coefficients of the nutrients in whole soybeans and in soybean oil meal, the latter at the two different levels, were calculated according to the usual indirect fashion. Consid- ering the 800 grams of alfalfa hay and the 112 grams of soybean oil meal, the ration fed in Period VIII, as the basal ration in Period IX, TABLE 8. AVERAGE PERCENTAGE COMPOSITION AND GROSS ENERGY CONTENT OF FEEDS AND RATIONS CONSUMED AND OF FECES EXCRETED: INVESTIGATION OF 1925 Dry matter Crude protein N-free extract Ether extract Ash Crude fiber Gross energy per kg. Period VI Alfalfa hay, 800 grams Alfalfa hay pet. 94.4 pet. 19.4 pet. 35 8 pet. 2 76 pet. 9 55 pet. 26 9 therms 4 24 Feces 71.8 9.28 19.0 3.83 10.2 30 4 3 20 Period VII Alfalfa hay, 800 grams, and whole soybeans, 225 grams Alfalfa hay 95.5 20.8 36.1 2 66 10 4 25 5 4 19 Soybeans 94.0 32.6 29.1 19.5 5 94 6 90 5 21 Ration consumed 95.5 23 34 5 6 35 21 4 42 Feces 69.2 10.3 19 2 3 03 9 51 27 22 3 19 Period VIII Alfalfa hay, 800 grams, and soybean oil meal, 112 grams Alfalfa hay 1 91.0 19.0 36 3 2 58 9 36 23 8 4 02 Soybean oil meal 90.7 38.8 30.5 6.18 6.57 8.67 4 39 91 22 7 35 5 3 23 21 4 07 Feces. . . 71.4 10.5 18.5 2.96 10.7 28.8 3.08 Period IX Alfalfa hay, 800 grams, and soybean oil meal. 340 grams Alfalfa hay 87.5 14.3 34.7 1.71 7.49 29.4 3.84 Soybean oil meal 91.2 40.4 31.1 5.04 6.48 8.16 4 35 Rations consumed 93.0 23.3 34.8 2.80 23.0 3.95 Feces.. . 60.0 9.39 14.2 1.59 7.22 27.6 2.69 Period X Soybean oil meal, 450 grams Soybean oil meal 89.6 39.2 30.3 5.20 6.38 8.51 4.28 Feces. . . 45.1 9.66 13.8 3.22 8.37 10.0 2.21 The chemical analyses of the alfalfa hay fed to the two groups of sheep in this period were: for sheep Bl, B2, and B3, dry substance 91.47 percent, crude protein 24.06, N-free extract 36.01, crude fiber 17.68, ether extract 3.34, ash 10:38, and gross energy 4,066 calories per gram; for sheep Rl, R2, and R3. dry substance 90.90 percent, crude protein 16.53 percent, N-free extract 36.4, crude fiber 26.79, ether extract 2.20, ash 8.86, and gross energy 3,995 calories per gram. 258 BULLETIN No. 303 [March, the digestibility of the nutrients in the additional 228 grams of soy- bean oil meal fed in this period was also calculated. A direct determi- nation of the digestion coefficients of soybean oil meal was obtained in Period X, in which one sheep ate soybean oil meal alone. The average percentage composition and the gross energy content of the feeds and rations consumed and of the feces of each period are tabulated in Table 8. The digestion coefficients of each ration and each feed, obtained directly or calculated by current methods, are given in Table 9. TABLE 9. SUMMARY OF DIGESTION COEFFICIENTS: INVESTIGATION OF 1925 Sheep No. Dry matter Crude protein N-free extract Ether extract Crude fiber Period VI Alfalfa hay, 800 grams Bl.. 66 66 66 63 65 63 65 78 80 81 76 78 77 78 77 77 77 74 75 75 76 63 63 58 45 40 34 51 48 48 48 47 51 48 48 B2 B3 Rl.. R2 R3 Average Period VII Alfalfa hay, 800 grams, and whole soybeans, 225 grams Bl. . 69 70 69 67 69 72 69 80 82 81 80 81 83 81 78 76 75 76 76 79 77 78 81 78 80 80 84 80 55 51 51 37 39 46 47 B2. .. B3 Rl R2 .*. R3 Average. . . Period VIII Alfalfa hay, 800 grams, and soybean oil meal, 112 grams Bl.., 67 82 78 67 28 B2. . . 66 83 76 55 27 B3 68 83 78 64 29 Rl . 63 73 76 44 48 -R2 61 73 76 50 42 R3 58 74 73 48 37 Averaee. . . 64 78 76 55 35 Period IX Alfalfa hay, 800 grams, and soybean oil meal, 340 grams Bl. . 67 80 80 67 41 B2. . 67 80 79 71 41 B3 67 80 79 67 42 Rl. . 65 75 78 63 43 R2 62 77 77 75 30 R3 64 79 77 78 37 Average. . . 65 78 78 70 39 Period X Soybean oil meal alone, 450 grams R3. M 00 5.3 1928] SOYBEAN PRODUCTS FOR SHEEP 259 TABLE 9. Concluded Sheep No. Dry matter Crude protein N-free extract Ether extract Crude fiber Period VII Whole soybeans, 225 grams (fed with alfalfa hay, 800 grams)* Bl. . 84 82 86 89 88 B2. .. 87 89 77 93 94 B3 . . 84 88 74 89 88 Bl.., 75 85 79 95 90 R2... 80 90 79 95 77 R3 94 95 94 102 10 Average. . . 84 88 82 94 19 Period VIII Soybean oil meal, 112 grams (fed with alfalfa hay, 800 grams)* Bl.. 80 98 97 100 270 B2. .. 73 104 76 100 280 B3 86 107 103 90 -250 Bl.. 49 55 76 17 R2 36 59 81 50 - 90 R3 10 57 53 24 -230 Average. . . 56 80 81 64 -187 Period IX Soybean oil meal, 340 grams (fed with alfalfa hay, 800 grams)* Bl. . 73 82 91 76 21 B2. .. 72 81 89 88 14 B3 72 82 87 76 - 11 Bl.. 64 72 84 72 R2 53 77 81 88 -168 R3 62 80 81 94 - 88 Average. . . 66 79 86 82 - 50 Period IX Soybean oil meal, 228 grams (fed with alfalfa hay, 800 grams, and soybean oil meal, 112 grams)* Bl. . 74 84 94 91 116 B2. . . 73 82 92 100 116 B3 72 85 92 81 127 Bl.., 62 71 86 75 135 R2 44 76 81 109 - 47 R3 57 80 81 119 71 Average. . . 64 80 88 96 86 Indirectly determined. Metabolizable Energy. As in the first investigation sufficient data were obtained for each sheep and each ration for the satisfactory cal- culation of metabolizable energy. The nitrogen balance data are tab- ulated in the Appendix, Table 21, while Table 10 presents the inter- mediate and final calculations of metabolizable energy. Table 11 gives the indirectly calculated values for the metabolizable energy of whole soybeans and of soybean oil meal. Table 12 presents a summary of the metabolizable energy of all feeds and rations fed in the investiga- tion of 1925. Discussion. In general, the 1925 trials were quite satisfactory. The coefficients of digestibility of the nutrients and the metabolizable 260 BULLETIN No. 303 [March, TABLE 10. THE GROSS ENERGY, ITS LOSSES, ITS PERCENTAGE DISTRIBUTION AND THE METABOLIZABLE ENERGY OF THE RATIONS FED: INVESTIGATION OF 1925 Metabolizable energy 05 1 1 i 4 4 1 1 1 h > 1 1 P Per kg. digestible organic matter I* Percentage losses In methane J5.1 o Energy per kilogram of dry matter Metabo- lizable 1 In methane 3 .Si 3 'o Dry matter eaten per day and bead Concen- trates |1 a Jo 2-<*"*a! COCCN a ft CO CO O) O -T t CO _ _ > r / . . 02 OCO^i (Nc 10 CN co ^CNOO coo o 1-1 00 * N Hi :O O is "-- o r~ 'H -j- IN **}< Tl- ri< 1" *** ^.r).^. Tf O!COr) Tf **a> CO ,JOOOCO .-it^CO - Tjoct>- tt) O5 CO 00 N O5 <-> U) i-irtO tceco IQ O 10 O *C *O *O ifJiOO TlT}"U5 TJ gOOCDt>- TOMO >O grams 00 00 rf h. -H o o COO=O OOt^ (N P-TJI O O Tf iO US "5 >o c^ N TfiOiO TCU51O U5 OC > i O^ 00 t- ^ O ^*OM t^COO >C 3 a . tct~-oo t- *C^iM)o <* TJI m o >o >o IM(N--5 rt COCC (NCOcO * CNCNCM PUMM CN * ' I-H 2osoo>o o>o SCOCO -*t^O i-iO^iiO '~ '-. '- "5 1 n! CN O t- O CM ^J" CO i-iOOCN CNCOCO " TfCOTji IO-*IM * 2t^t-i cococo >o 000 oooooo o> ^U5"0u; -*T}-TJI TJ< 71 ?1 71 ~- -- C~. O COCOCO COOCO CO ~'* TfHTl< ^^ ^ ^ CNCNCM CNCNM CN >< cococo t>-t-co oo 5^10 5 >O O iO h- C OOO OOOOOO TJI t~ t^ t- ITS o o : : : : : : 9 : : : g 1 -H N CO "N CO' > P5 < --oico ICNCO' > a rtsJrt < SOYBEAN PRODUCTS FOR SHEEP 261 >, _2 5 S M'a-s | fe||'S 1 ^ :gc 03 __ ,2 Jf^'H.S 3 ^ CQ OS ^ " CU S, J ^3 II CB_N 01 d a 1 S J 03 1 5-c 5 (N "a p r- 1 B "3 9 PH gl 1 'o HH gj d 03 V ~J 43 j| & "~- i ID i 11 I f || 03 o i ( s pa H dry matter In methane M >> 03 *o 03 a 3 2 I - < i3 3 1 a M 3 > a a M c u O S 1 1 'i a Pk 3 o H 81 s "c3_S o -*^ S 03 o ^ || O Dry mal per day !i a ca a o O 2 co t* co co t*- os -H SCOXJCO COOCO t^ O C-l (N CO O CO "5 00 Oi 00 C5 t>- t'* 00 *-*COCO COW* CO CO CO CO CO CO CO CO Sl^C^OO t^(NC ** 1-1 as oo ot-m CN OOO i ( O5 O) O "*'*'* ** * n 0003 M< ocooo <->&>*> 01 *S!SS8 SSS S t>. t^ ^5 CO O (N -*f 10 o 10 >~ <~ i-"; 10 ^OOCNO OOOiiH 00 oom 10 coooco *-< ^ ^ ^ iO *O *O ^O O iO irj O Tj* O IO . 00 00 O ^ *~ ' O> O3 03 a (NCOCO 10CDTJH rt ^* ^ iO CO CO CO ^ o i io--lOO OOCOOO (N 03 3 CM-HIN CMCMO) (N I 2O"5CO COOOO O gcoaoco coco>o >o V Tj< * * COCOCO *O 1 8 1OCMCN O--ICO t^ 10 oo o> oo oo !O o ' 1111 SSS S CNCNtN t~OO -" OOO O5 O O O -""*'* " .. 0,000 ooo o CO CO CO CO CO CO CO W CM C^ CO t* 1^ O5 Q.COCOCO CNCNCM CM see sii i Hi ;;; i -i tNco r33 J2 *" i-cCM CO "-KNCO (> 262 BULLETIN No. 303 [March, TABLE 11. COMPUTED METABOLIZABLE ENERGY OF WHOLE SOYBEANS AND OP SOYBEAN OIL MEAL: INVESTIGATION OF 1925 Sheep No. Per kg. dry matter Per kg. digestible organic matter Per kg. total digestible nutrients Period VII Whole soybeans. 225 grams (fed with alfalfa hay, 800 grams) Bl.. therms 4.123 therms 5.141 therms 3.995 B2 4.174 5.145 3.964 B3 3.873 4.946 3.823 Rl.. 3.292 4.950 3.607 R2... 3.575 5.087 3.762 R3 4.259 5.073 3.856 Average. . . 3.883 5.057 3.835 Period IX Soybean oil meal, 228 grams (fed with soybean oil meal, 112 grams, and alfalfa hay, 800 grams) Bl.. 3.266 3.841 3.586 B2 3.333 3.966 3.675 B3 3.005 3.555 3.317 Rl.. 3.053 3.938 3.704 R2... 2.126 3.318 2.959 R3 2.515 3.279 2.973 Average 2.883 3.650 3.369 TABLE 12. COMPUTED SUMMARY OF THE METABOLIZABLE ENERGY VALUES OF THE FEEDS AND RATIONS FED: INVESTIGATION OF 1925 Feed or ration Number aver- aged Per kg. dry matter Per kg. digestible organic matter Per kg. total digestible nutrients Alfalfa hay, 800 grams a therms 2 44 therms 4Ofi therms 3 94 Whole soybeans, 225 grams, and alfalfa hay, 800 grams g 27^ 411 3 91 Soybean oil meal, 112 grams, and alfalfa hay, 800 grams a 2 51 41^ 3 97 Soybean oil meal, 340 grams, and alfalfa hay, 800 grams 6 2 47 4 02 3 86 Soybean oil meal, 450 grams 1 3 49 4 22 3 94 Whole soybeans, 225 grams (fed with alfalfa hay, 800 grams) g 3 88 5 06 3 84 Soybean oil meal, 228 grams (fed with soybean oil meal, 112 grams, and alfalfa hay, 800 grams 6 2.88 3.65 3.37 energy values of each of the rations fed were quite generally consistent for all six sheep. While the indirectly calculated coefficients were, in general, better than those of the 1923 trials, they are still far from being completely concordant. In addition to being more consistent, the 1925 results represent more nearly the true values of the feeds as fed, since there was no appreciable amount of refused feed in any period. A modification of the ordinary method of computing indirect co- efficients of digestion was resorted to in Period IX, in which 800 grams of alfalfa and 340 grams of soybean oil meal were fed. Considering the 800 grams of alfalfa hay as a basal ration, the coefficients of digestion of the nutrients in the 340 grams of soybean oil meal were 1928] SOYBEAN* PRODUCTS FOR SHEEP 263 calculated. Then by considering the ration fed in Period VIII (800 grams of alfalfa hay and 112 grams of soybean oil meal) as the basal in Period IX, the coefficients of digestibility of the nutrients in the ad- ditional 228 grams of soybean oil meal were calculated. These figures are more consistent than those in the former calculation. An inadvertent occurrence in Period VIII explains the discrep- ancy existing between the coefficients obtained for the two groups of sheep. Sheep Bl, B2, and B3 were fed, during both the preliminary feeding period and the collection period, alfalfa hay containing an ex- cessive amount of leaves and dust, the remains of the large stack of alfalfa which had been used in the preceding periods.* This hay con- tained 24.06 percent of crude protein, which is about 5 percent higher than the hay fed in the other periods, and only 17.68 percent of crude fiber, nearly 10 percent lower. The alfalfa hay fed to Sheep Rl, R2, and R3 in this period was the same as in the other periods. GENERAL DISCUSSION OF EXPERIMENTAL RESULTS The digestion coefficients of all feeds and rations fed during both investigations are compiled in Table 13. In order to compare these results with others heretofore published, the literature was searched for all digestion experiments in which soybean products were used. There are at least two well-known compilations of average digestion coefficients, one by Lindsey, Haskins, Smith, and Beals, 24 * published in 1919, and one by Henry and Morrison, 21 * revised to 1923. These com- pilations, however, give only the average coefficients obtained for dif- ferent species of animals. Since digestion coefficients for all concen- trate feeds are indirectly obtained in the case of ruminants, it is of importance to know the basal rations to which these concentrates were added in order to discover to what extent associative effects may have operated. In 1900 Jordan and Hall 29 * of the Office of Experiment Stations, U. S. Department of Agriculture, published a compilation giving in de- tail the data on all American digestion experiments. This valuable summary has never been brought up to date. In 1925 Fraps 30 * made a compilation somewhat similar to that of Jordan and Hall, tho includ- ing less information concerning the experimental conditions of the individual trials." 'This experiment continued for a longer time than was anticipated and it was necessary at this point to purchase more alfalfa hay. b ln this compilation by Fraps two mistakes in the digestion coefficients given for soybean oil meal were noted when the original data were consulted. In the second line of Table 8, page 55, of Fraps' bulletin, the averages of 772 for protein, 74.2 for ether extract, etc., for soybean meal, are the average coefficients for a ration of .5 pound of soybean meal and 1 pound of timothy rowen hay obtained by Phelps and Woods 8 * on sheep. In line three of the same table and page, the coefficients given are similar coefficients obtained by the same (see page 266) 264 BULLETIN No. 303 [March, u cj f3 N h CO CO . 00 P 5 f 1" T c i < > f e ; ? > a 3 O 3 1 1- i a < t- o- - 1 1 ^ 8TIGATI II W S N 5 f i S S 5 S I S > if ! if b > r a 1" ^ 3 O c *~ 3 00 | 50 H g i i3 S^ if to * ! fO 0> * !* T 3 t- 1" : 2 < t > t> > a b 3 T; > e> c a 1 * PQ 55 O C w CM 2" o m a ) 00 - t- t- 1 i- a < a. 3 ) a. > t- > 01 ^ 5 C C 3 a > 01 00 O p, o 1 (4 1 of * o in a !O a < t > if 3 if 3 i if > a ; f 3 if i S m 00 w h o fc S g 55 H .2.2 ri N e^ i 00 O i n 00-" o 01 g 8 oo o 8 oo 1 S i 5"! M * . C BE* O O) 03 OS -C o3 g J3 i *" GO c3 03 % OS TABLE 13. SUMMARY OF DIGE Feed or ration Soybean hay, 1,051 grams (1923) Soybean straw, 1,135 grams (1923) Oat straw. 900 warns (1923) . . Alfalfa hay, 800 grams (1925) Whole soybeans, 228 grams, and oat straw (1923). . "3 T3 1 g M us oT s - i! Soybean oil meal, 225 grams, and soybean wains (1923). . . 1 T3 g 00* 1 > b O CQ T3 06" I O ro "oS EG 'C d E I & U do Whole soybeans, 228 grams (fed with oat warns) (1923). . . Whole soybeans, 225 grams (fed with alfa warns) (1925)... Whole soybeans, average g - B > > 5 > a X a g 1928] SOYBEAN PRODUCTS FOR SHEEP 265 o fc co t~. re u* (N CO 7 1 'S <-t ~S 03 aj -C C o to IN 00 CO o S oo wg II IN CD 00 00 *g 00 00 00 S o> - 00 T* C 'o 00 X *f 2 CO CO oo X 2 - - V - - ,2_2 to to S 'E "^ IN &* t 1 1 1 _ tN 2 d a 03 J3 J3 JsX 8 a 1 3 1 _^2 00 c8 si ^ ,-*. 5 ja 5 tSo 'is 'IS 'i "2 fT =, a 13 ^ T3 t3 09 >r._g c a or ratio i 2 M 03 B 8 grama d alfalfa grams ( 5 E 1 8 IN CO 2 o s CN a IN 3 cS ll E^Ol 1| 1* s S03 M "3 l" 5 ^ 1 'o s ~ / 3^ S2 c c C *- " en cU fl fl a ll a g IB 02 11 1" |1 02 1 z i do s 1 1 00 1 > _, X X H o > li 9 a,'S ii l| o 5 g o C 8*1 . n o o o c a a 5 266 BULLETIN No. 303 [March, Because of the absence of any recent compilation giving full de- tailed data for digestion experiments on soybean products, the original publications were examined and their essential features summarized in the Appendix, Table 22, which is intended to cover all experiments, both foreign and American, reported up to May, 1927. The table in- cludes a description of the ration fed, individual data for each animal, and the length of collection period, as well as the date and reference to the experimenter. It is therefore suited to a study of the extent of variation occurring in digestion coefficients obtained under different conditions. The Digestibility of Soybean Products The average percentage composition and the gross energy content of all feeds used in this investigation are shown in Table 14. There is also listed for comparison the percentage composition of the same feeds as found in Henry and Morrison's "Feeds and Feeding." 21 * From such a comparison, it may be considered that all feeds used were fairly representative samples of feed obtainable in the corn belt (except the alfalfa fed in the first part of Period VIII, referred to on page 263). While the investigation was concerned chiefly with a study of the soybean plant, information was obtained also for oat straw and alfalfa hay. With the exception of ether extract, which is of insignificant value in the feeding of roughages, all of the nutrients of alfalfa hay are digested to a considerably higher extent than are those of soybean hay. Both soybean straw and oat s!raw are inferior to soybean hay and alfalfa hay in digestibility, while oat straw is considerably su- perior to soybean straw. The digestibility of the nutrients in the combined rations was in- variably increased whenever alfalfa hay was substituted for either soybean straw or oat straw in a ration containing either soybean oil meal or whole soybeans. In a ration containing approximately .5 pound of whole soybeans and 1.75 pounds of oat straw, the coefficients for dry matter, crude protein, nitrogen-free extract, ether extract, and crude fiber were 48, 71, 43, 67, and 43 in the order named. The same coefficients for a ration in which the oat straw was replaced by ap- proximately the same amount of alfalfa hay were 69, 81, 77, 80, and 47. The coefficients for a ration of approximately .5 pound of soybean oil meal and 2.0 pounds of soybean straw were 55, 68, 64, 59, and 43, while for a ration containing .25 pound of soybean oil meal and 1.75 pounds of alfalfa hay the average digestion coefficients were 64, 78, 76, 55, and 35 respectively; for a ration containing .75 pound of soy- bean oil meal and 1.75 pounds of alfalfa hay, the average coefficients were, in order, 65, 78, 78, 70, and 39. investigators with sheep on a ration of .75 pound of soybean meal and 1.5 pounds of timothy rowen hay. Omitting these two misquoted series of results, the aver- age digestion coefficients for soybean meal become 90.9 for protein, 92.7 for ether extract, 67.8 for crude fiber, and 79.5 for nitrogen-free extract. 1928} SOYBEAN PRODUCTS FOR SHEEP 267 l&ff SS : fc : S-i IN * ii 000 _ _ M .CO CO a* K * S 58 ** ^IMCO OOO ooco 0-.0 _ oo o* g.0000 0500 6| 000 _ IN>OIO t~ .CO w g ft coc- 22 " "5t~ coco 31 000 00 M 00 "5 IN OOO OO3 H *ss co co co OOO5CO 00 Tf b~ _ .0- *. Q| *S3 oooo 0000 00-<1. >. <8<2 Soybean s Soybean si a a a as a a 22 &&& 00 02 CO s s a _ J:_3 && 7. V. '/. Oat straw, Oat straw, Alfalfa ha; Alfalfa ha; "c- 1 IN * -H 00 O iQOOiQTt< c a l^ * c gg T3 ^R >>0 i 2 UJOO^OCO & 1OOO"3-1< o 8 i OO>iOO5T|l CO CO^-f* 9 oj i 1 13 o | OTjlU5COt T* ri * IN "5 c o O o o ss 1 O>O"5CO(N ^(OCOtOCO 1 O " a> g'S . & '-HO IN^-HCOi-KN 03 h O 0) Jo T3 OJ) 1 | O lO ^ "5 W > I MgS J. 9 Ja a t x> X 8 i e QU a ^OSTfllNiO 1 UJOOOCO c o O : .s^' : C g u V o ^ In ft) ^ M-SJ3 03 Q. o g QO^KO 268 BULLETIN No. 303 [March, From Table 13 it appears that the average digestibility of the dry matter, crude protein, and nitrogen-free extract was very nearly the same for the four rations used in 1925. Some variations occurred in the digestibility of crude fiber, but it is doubtful whether the aver- age differences are significant, while the digestibility of the ether ex- tract varied directly with the fat content of the ration. For rations consisting of 800 grams of alfalfa hay alone, and for those containing 800 grams of alfalfa hay combined with whole soybeans (225 grams) or with soybean oil meal in varying amounts (112 and 340 grams), the average coefficients of digestion of dry matter were, in the order named, 65, 69, 64, and 65; those for crude protein, 78, 81, 78, and 78; for nitro- gen-free extract, 76, 77, 76, and 78; for crude fiber, 48, 47, 35, and 39; and for ether extract, 51, 80, 55, and 70. It seems remarkable that these four rations should possess so similar a digestibility for most of the nutrients. Significance of Refused Feed in Digestion Experiments The coefficient of digestibility of a nutrient is ordinarily expressed as that percentage of the nutrient consumed which does not appear in the feces. Obtained in this manner, however, it is normally applied to the feed as fed in the calculation of the digestible nutrients of rations. This brings up the question, Do the coefficients thus obtained repre- sent the digestibility of the feed fed under practical conditions? Under practical conditions of feeding, animals naturally do not eat the coarse, woody, and unpalatable portions that may be found in roughages. An examination of the mangers in most any barn will sub- stantiate this statement. Thus Nevens 31 * found that a close relation- ship existed between the thickness of planting soybeans and the pro- portion of refused hay (stems). Feeding trials showed that while 120 pounds per ton were refused from hay grown in plots on which 1.25 bushels of soybeans had been sown, 360 pounds per ton were refused from hay taken from the plots on which .5 bushel of seed had been sown. If, under practical conditions, a certain amount of the rough- age is left uneaten, digestion coefficients based on the amounts of the nutrients offered, instead of upon the amounts of nutrients consumed, would be of considerable importance tho subject to variation depend- ing upon the condition of the feed. While the data in this investigation may not be sufficient to justify any general conclusions in this connection, a comparison of the di- gestion coefficients based upon the nutrients consumed and on the nu- trients offered is of interest. Therefore, in addition to the usual method of calculating digestion coefficients, coefficients were also calculated on the basis of the nutrients offered. The two sets of coefficients are averaged for each ration fed in the 1923 investigation in Table 15. As would be expected, all coefficients calculated on the basis of nutrients 1928} SOYBEAN PRODUCTS FOR SHEEP 269 TABLE 16. DIGESTIBLE NUTRIENTS PER 100 POUNDS OF FEED OR RATION COM- PUTED BY MEANS OF COEFFICIENTS SUMMARIZED IN TABLE 15 a Ration Dry matter Crude protein N-free extract Ether extract Crude fiber Total di- gestible nutrients Soybean hay /6s. (41.4) Ibs. (9.5) (6s. (19 . 5) Ibs. (2.1) Ib*. (7.2) Ibs. (40.9) Sovbean straw 49.1 (24.1) 11.5 (.45) 23.2 (14.5) 2.5 (.12) 8.8 (7.8) 49.1 (23.0) Oat straw and whole soybeans 36.0 (36.6) .69 (8.3) 22.0 (13.7) .18 (3.5) 12.1 (8.4) 35.2 (38.3) Oat straw 43.8 (34.0) 10.3 (.42) 16.7 (14.8) 4.2 (.33) 10.4 (16.4) 46.9 (32.4) 43.6 (47.5) .69 (6.9) 19.0 (23.8) .43 (1.0) 21.0 (13.6) 41.7 (46.6) 51.0 8.0 25.1 1.1 14.3 49.9 The figures given in parentheses were obtained by the use of the coefficients of digestion based on the quantity of feed offered. They therefore represent the amounts of digestible nutrients per 100 pounds of feed or ration. The figures not enclosed in parentheses were obtained by the use of the coefficients based on the quantity of feed consumed. They represent, therefore, the digestible nu- trients per 100 pounds of consumed (edible?) feed or ration. offered are lower than those calculated on the basis of nutrients con- sumed. The greatest, differences, of course, relate to the less palatable straws and, among the different nutrients, to the crude fiber, but it should be noted that all other nutrients are considerably affected also. These differences are, perhaps, better shown in Table 16, in which are tabulated the average number of pounds of digestible nutrients per 100 pounds of feed or ration. In Table 16 the figures set off in parentheses do not include the nutrients in the refused feed; the others do. The figures in the last three lines for alfalfa hay, whole soybeans, and soybean oil meal, how- ever, would be the same on both bases since the nutrients offered were also consumed. In calculating the digestible nutrients in soybean hay (offered), soybean straw (offered), oat straw (offered), alfalfa hay, whole soybeans, and soybean oil meal, the average percentage com- position of these feeds obtained in this investigation and the coefficients of digestion tabulated in Table 15 on the basis of nutrients offered were used. In calculating the digestible nutrients in the combined rations offered and consumed, the average percentage composition of these rations as given in Table 8 and the corresponding coefficients given in Table 15 were used. It can be seen from Table 16 that the usual method of calculating the amounts of digestible nutrients indicates a much greater utilization of the nutrients than is actually the case when allowance is made for the refused, and presumably inedible, portions. The percentage of total digestible nutrients in the soybean hay in these experiments was actually 42.2 percent, as compared with 49.1 percent as ordinarily computed, neglecting the inedible fraction. The percentage of total digestible nutrients in soybean straw was actually 23.6 percent, as 270 BULLETIN No. 303 [March, compared with the customary computation of 35.2 percent, a difference of 11.6 percent. Similar differences may be noted for the other rations. The question whether the amounts of feeds refused in this experi- ment were excessive should be considered at this point. The following were the average amounts of feeds fed and the average percentages of dry matter refused from each ration: of 2.25 pounds of soybean hay an average of 13.9 percent of the dry matter was refused; of 2.5 pounds of soybean straw, 31.1 percent was refused; of about 1.75 pounds of oat straw and .5 pound of whole soybeans, 22.3 percent of the dry matter of the oat straw was not eaten; of about 2 pounds of oat straw, 22.2 percent was refused; and of about 2 pounds of soybean straw and .5 pound of soybean oil meal, 6.9 percent of the dry matter of the soybean straw proved inedible. The feeding experiment of Nevens, 31 * in which, depending upon the coarseness of the hay, from 6 to 18 percent of soy- bean hay was refused by dairy cattle, has been mentioned. In a similar experiment with sheep Kammlade 32 * fed one lot of ewes 3.12 pounds of alfalfa hay, of which 1.7 percent was refused. A second lot receiving 3.77 pounds of soybean hay, refused an average of 15.2 per- cent. When a ration of .98 pound of grain and 3.01 pounds of alfalfa hay was fed, less than 1 percent was refused, but a ration of .87 pound of grain and 4.26 pounds of soybean hay was refused to the extent of 20.7 percent. In an experiment with soybean straw Kammlade found that more than 40 percent was refused. An examination of a number of the digestion experiments pre- viously mentioned in this bulletin showed that some feed was refused in practically all cases in which the animals were receiving approxi- mately full feeds of a ration containing a roughage. In Kellner's ex- periment on soybean hay 3 * one sheep refused 8.5 and the other 18.0 percent of the dry matter from 1 kilogram of the hay. In an experi- ment with soybean hay Emery and Kilgore 5 * reported that a goat which received 1,200 grams daily refused 28.1 percent of the dry mat- ter, and a cow which received 16 pounds daily refused 14.7 percent. From a ration of 400 grams each of soybean hay and English hay, Lindsey 6 * found 14.4 percent of the dry matter of soybean hay refused by one sheep, and in another experiment 11 * one sheep, offered 2,000 grams of green soybean fodder and 400 grams of English hay, refused 7.2 percent of the dry matter. In the experiments reported by Phelps 13 ' 14 * on soybean fodder, various but small amounts of feed up to 6.5 percent were refused. Finally, Emery and Kilgore 16 * observed that one of two goats re- fused 6.2 percent and the other 31.5 percent of the dry matter from a ration of 6 pounds of soybean silage, while in Hopkins' experiment with steers on the same feed, 17 * from 11.7 to 15.5 percent of the dry matter was refused from rations varying from 48 to 66 pounds of silage per head daily. 1928] SOYBEAN PRODUCTS FOR SHEEP 271 A consideration of these experiments reveals two facts: first, that the amounts of refused feed in the present investigation were not ab- normal for the feeds used; and second, that the digestion coefficients calculated in the usual manner may be of little value in estimating the amounts of digestible nutrients in certain feeds in so far as they con- tain inedible portions. While it may be questioned whether the refusal of oat straw by experimental animals indicates anything more than the offering of more of this unpalatable feed than the animal cares to consume, the situation with reference to soybean roughages is evi- dently different. The general refusal of parts of these roughages, no matter how little is fed, as well as the appearance of the refused por- tions, demonstrates that soybean roughages may be considered as being only partially edible. It will always remain a difficult task, therefore, to estimate the digestible nutrients in samples of these feeds by any set of digestion coefficients since the inedible fraction will vary. The nutrients in the inedible fraction obviously are as worthless to the animal as the edible but indigestible or undigested nutrients. Significance of Indirectly Calculated Coefficients of Digestibility The method of calculating the digestible nutrients of a concentrate for ruminants, by subtracting from the digestible nutrients in a mixed ration the amounts computed for the basal ration as determined by a previous trial, is based upon the assumption that the addition of the concentrate to the basal ration does not affect the digestibility of the nutrients in the basal ration and vice versa. That this assumption is not sound from the physiological point of view has long been recog- nized, and that it is not correct in practice as regards certain nutrients in certain combinations of feeds was noted by some of the early work- ers, for example, Hirschler in 1886, 33 * Gottwald in 1888, 34 * and Kuehn in 1894. 35 * In reviewing the literature it soon becomes evident that, in gen- eral, the nutrients present in the smallest amounts are the most af- fected. Thus, fat and crude fiber are especially influenced, while nitrogen may be affected in the feeds low in this element. Dry matter and nitrogen-free extract are usually but little influenced; the latter because it is usually present in considerable amounts, and the former because it reflects both the depressing and elevating effects on the several nutrients. It is a well-known fact among investigators in animal nutrition that determinations of the availability or digestibility of the ether extract of feeds are often unsatisfactory and as a rule are quite un- reliable when obtained by indirect calculations. One of the chief con- tributory factors is the influence of the metabolic products of the feces, particularly the bile residues, while the presence of fat in only small 272 BULLETIN No. 303 [March, quantities in most concentrate feeds magnifies any influence of these disturbing factors. Armsby 36 * states: "Even with the herbivorous ani- mals, however, the presence of the so-called metabolic products of the feces may give rise to serious errors in the determination of the real digestibility of some ingredients of the food, notably fat and protein." The determination of the digestibility of ash is so notably unreliable that only occasionally does an investigator attempt to calculate it. Forbes 37 * states that we have no satisfactory means for determining the digestibility of the mineral nutrients and that it is not warrantable to speak of their digestibility. The work of Eckles, 38 * of Mumford, Grindley, Hall, and Em- mett, 39 * and of Snyder 40 * seems to indicate an inverse relationship be- tween the quantity of crude fiber in feeds and the digestibility of the nitrogen-free extract, altho Ewing and Wells, 41 * in an excellent review of the literature and a study of what they very aptly term "the asso- ciative action of feeds," found that with a decrease in the amount of crude fiber in the form of silage there was a decrease, instead of an increase, in the digestibility of the nitrogen-free extract. Ewing and Wells also found, in agreement with several early workers, that starch when fed in excessive amounts seemed to exert a depressing effect upon the digestibility of the nitrogen and crude fiber, even when the excess was not great. The depression of the digestibility of nitrogen and crude fiber was often accompanied by a rise in the digestibility of fat, which was quite noticeable in the high starch ra- tions. At the same time they noted a lowering in the digestibility of the nitrogen-free extract of 10 percent or more when starch was intro- duced into the ration in excessive amounts. Atwater, 42 * on the other hand, in connection with reports of digestion experiments with men, says: "Increasing the quantity of either carbohydrates or fat above those common in the ordinary diet had no effect upon the availability of the other nutrients of the diet, while the availability of the fat or carbohydrates in the experiment in which either was used in large quantities was as large as, or larger than in the experiments in which the quantities were more nearly like those in the ordinary diet." Other general researches or discussions of the influence of one food upon another are those of Rubner, 43 * of Atwater, 44 * of Ewing, Wells, and Smith, 45 * of Fraps, 30 * and of Forbes, Beegle, Fritz, and Mensch- ing, 12 * all of whom agree in reaching conclusions similar to that ex- pressed by Dietrich and Grindley 46 * as follows: "The coefficients of digestibility obtained for a single feed by calculation from the data obtained for a combination of feeds are not reliable, especially when the single feed has been fed in comparatively small quantities. By this method of calculation all of the difference between the values for the single and the corresponding values for the combined feeds is credited to the single feed, whereas it is probable that in the combined ration each feed exerts an influence upon the digestibility of the other." SOYBEAN PRODUCTS FOR SHEEP 273 In the last sentence of this quotation is found the explanation of the fact that coefficients of digestion obtained by the indirect calcula- tion may indicate a digestibility of less than nothing or more than 100 percent. However, such an associative effect between feeds need not result in impossible digestion coefficients, and the absence of impossible results of such indirect calculations is no indication that associative relations are not operating between the added feed and the basal ration. Fraps, 30 * in connection with his compilation of American digestion experiments, made a statistical study of the variations in the digest- ibility of all feeds on which three or more tests had been made. By dividing the "percent error" found in Table 1 of Fraps' bulletin by .6745, the coefficients of variation are obtained. This was done by the authors, and the coefficients of variation for all feeds on which the di- gestibility had been directly determined and for all feeds on which the digestibility had been indirectly determined, were averaged separately. The average coefficients of variation for the directly obtained coeffi- cients of digestion are: 19 for protein, 18 for fat, 11 for crude fiber, and 7 for nitrogen-free extract. Those for the indirectly obtained co- efficients are: 29 for protein, 21 for fat, 63 for crude fiber, and 8 for nitrogen-free extract. The errors that may be incurred in the indirect method of calculation are again clearly emphasized. The associative action of feeds was very pronounced in the present investigation. Consider the indirectly determined coefficients of diges- tion for whole soybeans in Period III, in which 12 sheep were fed an average of .5 pound of whole soybeans and 1.75 pounds of oat straw, as w r ell as in Period VII, in which 6 sheep received the same amount of whole soybeans while the oat straw was replaced by approximately the same amount of alfalfa hay. These variations are best shown in Table 17, which gives the maximum and minimum coefficients and the average for each nutrient. It appears, from a consideration of these data, that of all the nutrients of soybeans crude protein only is unaffected in estimated digestibility by the change from oat straw to alfalfa hay. With all other nutrients the digestibility, calculated indirectly, is greatly in- creased. The variations in the digestibility of any nutrient (with the possible exception of crude protein) in each period are very large, and if the two periods are considered together, the average coefficients, with the same exception, are evidently of little if any value. Undoubtedly an associative action is exerted in digestion between soybeans and the roughages with which they are fed. It is equally true that indirect calculations of the digestibility of soybeans under these conditions give purely fictitious values, with the possible exception of the di- gestion coefficients for protein. A similar tabulation of the variations in the indirectly calculated digestion coefficients for the nutrients of soybean oil meal is presented 274 BULLETIN No. 303 a rj. 1| 'oo -32 + s i 1 i" ? i M ^ 6| g 2 S .1 o *^ H O oo ^ OS w wg o IH oo S Q 55 I I I i a Q * H H g 1! i -C + -3 co Is 55 M < H f* J2J y ! ^1 O 7 t>. O 03 03 j BEAN s g'^ >. o Is o ooo MOO H IN CQ oo 00 O K) 55 3 3 1 - o fe g *i f 75 to 94 84 n 1 M g 55 H H * COEFFIC Number of Sheep o 1-1 o] 03 03 i ^ 3 =ai s i g =36 ' ? I 5 ^2 _0 1 1 II a fi ^^ .2 03 i g'S "S S 2 03 03 8 K 5 H M M I' IN N ._ 00*^ IN ^ 'a? 11 co N is i" i g-0 1 m S C DC g s -s E - O Bl 9 s 'o 'o S 's | ll" | & 15 |5 o^ >)00 & I 00 r 3 1 ^ > B H HH 19381 SOYBEAN PRODUCTS FOR SHEEP 275 in Table 18. The same sort of variations that were observed in the case of whole soybeans are also shown for soybean oil meal. The extremely wide variations exhibited by Period VIII are not entirely attributable to the method of indirect calculation, as has already been explained. The degree to which digestion coefficients may be influenced by different basal rations is well illustrated by the coefficients for crude fiber. For soybean oil meal when fed with soybean straw they vary from 132 to 534, with an average of 323, and for the same meal when fed with alfalfa hay (Period IX) they vary from to 168, with an average of 50. In other words, simply changing the basal ration changes a coefficient from positive 323 to a negative 50. As Forbes 12 * says (page 233), in referring to similar results obtained with swine, such a determination is certainly, "to borrow an expression from Thudichum, 'a ceremonious delusion'." Such values are positive evi- dence of "associative action of feeds" and demonstrate that an in- directly determined digestion coefficient may possess no significance whatever. Thus, a coefficient determined indirectly from a certain mixed ration may not be applicable with any degree of certainty to any other ration. A rather common expedient used to avoid the reporting of im- possible digestion coefficients is the procedure of reporting as 100 all coefficients obtained above this value, and of reporting all negative coefficients as zero. For example, Emery and Kilgore 5 * determined the digestibility of soybean hay, using a goat and a cow. The average di- gestion coefficient for ash was reported as 23.70, but the individual coefficients were 104 and 47.41, obviously the average of 23.70 is the average of and 47.41. Concerning such a procedure, Forbes 12 * says (page 234) : "Digestion coefficients of less than nothing, and more than 100 percent, show that the determination of digestibility of supplementary foods by difference, in the usual way, is not free from objection, since the supplement affects the digest- ibility of the basal ration, which the method assumes to be constant. It seems to us more nearly correct, however, to use the figures obtained than to call all minus coefficients zero, and to give a value of 100 percent to all those which seem to be above that figure." However, it may be argued with greater force, in our opinion, that the obtaining of such impossible coefficients is prima fade evidence that the method of computation is incorrect and that the results ob- tained with it, in such cases at least, are of no demonstrable value. In the determination of digestion coefficients by difference, two things are illustrated by Periods VIII, IX, and X. First, as the pro- portion of concentrates increases, the variations in the digestion co- efficients decrease, and the coefficients themselves approach more nearly the coefficients found when the concentrate is fed alone. Second, more consistent and apparently more reliable coefficients are obtained when the basal ration contains a small amount of the concentrate. 276 BULLETIN No. 303 [March, Thus from Period IX, when the digestibility of the 228 grams of soy- bean oil meal was calculated considering the basal ration to be 112 grams of soybean oil meal and 800 grams of alfalfa hay, the coeffi- cients were more consistent than they were when the digestibility of the 340 grams of soybean oil meal was calculated considering the basal ration to be 800 grams of alfalfa hay alone. In the determination of digestion coefficients of the nutrients of a concentrate for ruminants, the basal ration is usually a roughage alone; from the above consider- ations better results would perhaps be obtained if the basal ration con- tained a small amount of the concentrate. The not infrequent occurrence of impossible digestion coefficients when the digestibility of concentrates by ruminants is determined in- directly on the assumption that it, as well as that of the basal ration, is unaffected by their admixture in a ration, testifies to the inaccuracy of this basic assumption. Furthermore, the variable digestion coeffi- cients calculated for concentrates when added to different basal ra- tions possess the same significance. Finally, the securing of possible coefficients (between and 100), or of coefficients such as would be expected for a given concentrate, is no evidence that they have not to some extent been vitiated by the existence of associative relations in digestion between the concentrate and the basal ration with which it was fed. It thus appears that indirectly determined coefficients of digestion are unreliable and may be grossly in error. Rather than at- tempt indirect determinations of concentrates by a method known to be inaccurate, it may be far better to concern oneself with the digest- ibility of the concentrate in combinations with those roughages and other feeds with which it is most commonly fed. Metabolizable Energy of Soybean Products Determined for First Time Thruout this investigation all digestion trials were also metabo- lism trials. The daily excretion of urine was weighed, aliquoted, and analyzed for total nitrogen and gross energy. The gross energy of the feces and all samples of feeds and orts was also determined. Thus by using average factors for the calculation of the loss of energy in the methane, all data necessary for the calculation of the metabolizable energy of the various feeds and rations were available. The values for the metabolizable energy of soybean products obtained in this investi- gation are the only available determinations in this country. A sum- mary of the metabolizable energy of all feeds and rations fed during the entire investigation is presented in Table 19. The availability, as metabolizable energy, of the gross energy of the feeds and rations is, in decreasing order, as follows (page 278) : 1928} SOYBEAN PRODUCTS FOR SHEEP 277 i'-S-S ; CO CO CO CO CO * CO M CO CO Sx |2 fcs ININN 278 BULLETIN No. 303 [March, perct. Soybean oil meal, 450 grams (fed alone) 68.2 Soybeans, 225 grams, and alfalfa hay, 800 grams 59.4 Soybean oil meal, 112 grams, and alfalfa hay, 800 grams 56.0 Soybean oil meal, 340 grams, and alfalfa hay, 800 grams 54.9 Alfalfa hay, 800 grams 54.3 Soybean oil meal, 225 grams, and soybean straw, 910 grams 45.0 Soybean hay, 1,051 grams 43.8 Soybeans, 228 grams, and oat straw, 779 grams 41.0 Oat straw, 900 grams 33.8 Soybean straw, 1,135 grams 322 The metabolizable energy per kilogram of digestible organic mat- ter in all feeds and rations fed, including the calculated values for the concentrates, is, in decreasing order, as follows: therms Soybeans, 225 grams (with alfalfa hay, 800 grams) 5.06 Soybeans, 228 grams (with oat straw, 779 grams) 4.60 Soybeans, 225 grams, and alfalfa hay, 800 grams 4.33 Soybean oil meal, 450 grams (fed alone) 4.22 Soybeans, 228 grams, and oat straw, 779 grams 4.18 Soybean oil meal, 112 grams, and alfalfa hay, 800 grams 4.15 Alfalfa hay, 800 grams .' 4.06 Soybean oil meal, 340 grams, and alfalfa hay, 800 grams 4.02 Soybean hay, 1,051 grams 3.73 Soybean oil meal, 225 grams, and soybean straw, 910 grams 3.69 Soybean oil meal, 225 grams (with soybean straw, 910 grams) 3.66 Soybean oil meal, 228 grams (with alfalfa hay, 800 grams, and soy- bean oil meal, 112 grams) 3.65 Oat straw, 900 grams 3.61 Soybean straw, 1,135 grams 3.57 Expressed in this manner there appears to be considerable dif- ference in the amount of metabolizable energy in different feeds and rations. For the non-nitrogenous roughages, oat straw and soybean straw, there is an average of 3.59 therms of metabolizable energy per kilogram, or 1.63 therms per pound of digestible organic matter. This value compares favorably with Armsby's 47 * value of 1.588 therms per pound of digestible organic matter in roughages in general for rumi- nants. For the oat straw alone the value of 3.61 therms per kilogram of digestible organic matter was obtained. Armsby 49 * quotes from Kellner the value of 3.740 therms per kilogram of digestible organic matter in oat straw obtained by cattle. There are no results with sheep reported. For the nitrogenous roughages, soybean hay and alfalfa hay, the average is 3.90 therms per kilogram or 1.77 therms per pound of di- gestible organic matter. For the alfalfa hay alone the value of 4.06 therms per kilogram or 1.85 therms per pound of digestible organic matter was obtained. Armsby 49 * quotes from Tangl et al the value of 4.467 therms obtained with sheep; and from his own work with cattle, the value of 3.605 therms. The value obtained in this investigation is near the average of th^se two reported ^alu^s. Christensen and Hop- 1928} SOYBEAN PRODUCTS FOR SHEEP 279 per 48 * obtained an average value with two steers of 1.700 therms per pound of digestible organic matter, a value slightly lower than ob- tained with the sheep in the present investigation. The average value for whole soybeans is 4.83 therms per kilogram or 2.19 therms per pound of digestible organic matter. This value is considerably higher than Armsby's 47 * calculated average value of 1.814 therms per pound of digestible organic matter in concentrates with more than 5 percent digestible fat. It is also considerably higher than the value of 3.977 therms per kilogram of digestible organic matter, quoted by Armsby 49 * from Voltz et al, for palm-nut meal with sheep, altho it checks almost exactly the value of 4.849, obtained by the same authors with steers. While the value of 4.22 therms of metabolizable energy per kilo- gram of digestible organic matter was obtained in the single trial in which soybean oil meal was fed alone, the average of all trials, in- cluding those in which the metabolizable energy of the soybean oil meal was calculated by difference, is 3.69 therms per kilogram or 1.67 therms per pound of digestible organic matter. This value is lower than Armsby's average values of 1.996-2.177 for oil meals, altho the directly determined value of 4.22 therms per kilogram or 1.92 therms per pound agrees very closely. The literature does not contain any experiments in which the metabolizable energy of an oil meal has been determined either directly or indirectly for ruminants. When the metabolizable energy of a feed is related to the digest- ible organic matter in the feed, different feeds give widely different values. Much of this variability is probably due to the fact that di- gestible fat contains approximately 2.25 times as much metabolizable energy as either protein or carbohydrates. When the metabolizable en- ergy is related to the so-called "total digestible nutrients/' in which digestible fat is expressed in carbohydrate equivalents, the differences in the metabolizable energy values of different feeds become much less. For example, the metabolizable energy values of all feeds summarized in Table 19, when expressed in terms of metabolizable energy per kilo- gram of digestible organic matter, vary from 5.06 for a ration of 225 grams of whole soybeans (calculated from the ration containing in addition 800 grams of alfalfa hay) to 3.57 for a ration of soybean straw. When expressed in therms per kilogram of total digestible nu- trients, the metabolizable energy varies only from 3.97 for a ration of 112 grams of soybean oil meal and 800 grams of alfalfa hay to 3.37 for 228 grams of soybean oil meal (calculated from a ration containing in addition 800 grams of alfalfa- hay and 112 grams of soybean oil meal). The indirect determination of the metabolizable energy of a con- centrate is, of course, open to the same errors as the indirect determi- nation of its digestibility, but since the metabolizable energy involves all of the organic nutrients, the effect of the errors on the different nu- 280 BULLETIN No. 303 [March, trients may be expected to be greatly reduced by compensation. Thus, the values in the last column of Table 19 for any one feed vary only slightly according to the method of their calculation. However, the estimated metabolizable energy values of soybeans and soybean oil meal are, with one exception, greater when computed from the results obtained with an alfalfa hay ration than from the results of a ration containing oat straw or soybean straw. It may be significant that the single result obtained with soybean oil meal alone checks closely with the values computed from the rations containing alfalfa hay. SUMMARY AND CONCLUSIONS The experimental work reported in this bulletin is concerned with the determination of the coefficients of digestibility of the nutrients and the metabolizable energy of soybean products for sheep. The soybean products studied were soybean hay, soybean straw, whole soybeans, and soybean oil meal. During the last twenty-three years there have been reported but two digestion experiments with soybean products; one was concerned with a mixed ration containing soybean hay for ruminants, and the other was concerned with soybeans for swine. There have been no American digestion experiments with either soy- bean straw or soybean oil meal. The metabolizable energy values for these products obtained in this investigation constitute the only pub- lished values, while the metabolizable energy value obtained for soy- bean oil meal is the only available directly determined value for an oil meal of any kind. The experiments were carried out during a part of 1923 and a part of 1925. During 1923 the digestibility and metabolizable energy of five rations were determined, using with each ration 12 sheep, a number three times as large as has ever been used in a single digestion experiment with soybean products on ruminants. During 1925 the digestibility and metabolizable energy of five rations were determined using 6 sheep. Digestion coefficients for soybean hay, soybean straw, alfalfa hay, oat straw, and of rations containing soybean straw and soybean oil meal, oat straw and whole soybeans, alfalfa hay and whole soybeans, and alfalfa hay and soybean oil meal at two different levels, were ob- tained directly, as was also the metabolizable energy of each. The di- gestion coefficients and the metabolizable energy of soybean oil meal were obtained directly with one sheep. Indirect calculations were made in the usual manner of the digestibility and metabolizable energy of soybeans and soybean oil meal. The average results obtained are tabulated in Tables 13 and 19. It appears that certain roughages, because of their physical char- acter, are only partially edible by farm animals. Digestion trials with such feeds involve variable amounts of orts and give more or less 1928} SOYBEAN PRODUCTS FOR SHEEP 281 unsatisfactory digestion coefficients. The coefficients obtained evidently apply to the edible portion of the feed only, and should not be used in connection with weights of feed offered in computing the digestible nutrients of rations. Such coefficients do not have a general applica- bility, since the inedible portion of the roughage will vary with the conditions under which it was grown or prepared for the animal. Soy- bean straw and soybean hay are roughages of this character, and the use of the digestion coefficients and metabolizable energy values ob- tained for them must be tempered by the above considerations. From a critical study of the results obtained in this investigation, as well as those in numerous previous investigations, the authors feel that coefficients of digestion obtained by the usual method of indirect calculation are in general unreliable. Apparently the digestibility of concentrate feeds cannot be satisfactorily determined with ruminants, since they exert variable associative effects in digestion when com- bined in rations with other feeds. It is recommended that, in place of attempting such unsatisfactory determinations, the digestibility of concentrates for ruminants, combined with those feeds with which they are commonly fed, be investigated. Only the directly determined re- sults on the combined rations can be considered significant. LITERATURE CITED 1. WEISKE, H., DEHMEL, B., AND SCHULZE, B. Jour. Landw. 27, 511. 1879. 2. WEISKE, H., KENNEPOHL, G., AND SCHULZE, B. Jour. Landw. 31, 209. 1883. 3. KELLNER, O. Landw. Vers. Sta. 32, 72. 1886. 4. STURTEVANT, E. L. N. Y. (Cornell) Agr. Exp. Sta. Rpt. 1884, 45. 5. EMERY, F. E., AND KILGORE, B. W. N. C. Agr. Exp. Sta. Bui. 97. 1894. 6. LINDSEY, J. B., BEALS, C. L., AND SMITH, P. H. Mass. Agr. Exp. Sta. Bui. 181. 1917. 7. LINDSEY, J. B., SMITH, R. H., AND HOLLAND, E. B. Mass. (State) Agr. Exp. Sta. Rpt. 1894, 146. 8. PHELPS, C. S., AND WOODS, C. D. Conn. (Storrs) Agr. Exp. Sta. Rpt. 1895, 187. 9. WILLARD, J. T., AND CLOTHIER, R. W. Kans. Agr. Exp. Sta. Bui. 103. 1901. 10. LINDSEY, J. B. Mass. (Hatch) Agr. Exp. Sta. Rpt. 1904, 63. 11. LINDSEY, J. B. Mass. (Hatch) Agr. Exp. Sta. Rpt. 1905, 45. 12. FORBES, E. B., BEEGLE, F. M., FRITZ, C. M., AND MENSCHING, J. E. Ohio Agr. Exp. Sta. Bui. 271. 1914. 13. PHELPS, C. S., AND BRYANT, A. P. Conn. (Storrs) Agr. Exp. Sta. Rpt. 1896, 246. 14. PHELPS, C. S. Conn. (Storrs) Agr. Exp. Sta. Rpt. 1898, 204. 15. LINDSEY, J. B., HOLLAND, E. B., AND SMITH, P. H. Mass (Hatch) Agr. Exp. Sta. Rpt. 1907, 96. 16. EMERY, F. E., AND KILGORE, B. W. N. C. Agr. Exp. Sta. Tech. Bui. 4 (Bui. 87d) 1892. 17. HOPKINS, C. G. 111. Agr. Exp. Sta. Bui. 43. 1896. 18. MOOERS, C. A. Term. Agr. Exp. Sta. (private correspondence). 19. LINDSEY, J. B. Mass. (Hatch) Agr. Exp. Sta. Rpt. 1903, 82. 20. BARTLETT, J. M. Maine Agr. Exp. Sta. Bui. 110. 1904. 21. HENRY, W. A., AND MORRISON, F. B. Feeds and feeding. 18th ed. unabridged, 723. Henry-Morrison Co. 1923. 282 BULLETIN No. 303 [March, 22. MENTZEL, O., AND LENGERKE, A. V. Landwirtschaftliche Kalender. Pt. 1. 1913. 23. HONCAMP, F. Landw. Vers. Sta. 73, 241. 1910. 24. LINDSEY, J. B., HASKINS, H. D., SMITH, P. H., AND BEALS, C. L. Compilations of analyses. Mass. Agr. Exp. Sta. Spec. Bui. Nov. 1919. 25. FORBES, E. B. Ohio Agr. Exp. Sta. Circ. 152. 1915. 26. ARMSBY, H. P. The nutrition of farm animals, 639. Macmillan. 1917. 27. RUBNER, M. Ztschr. Biol. 21, 316, 329; 1885. Ibid. 42, 303; 1901. 28. KELLNER, O., AND NEUMAN, R. Landw. Vers. Sta. 73, 235. 1910. 29. JORDAN, W. H., AND HALL, F. H. U. S. Dept. Agr. Bui. 77, 90. 1900. 30. FRAPS, G. S. Texas Agr. Exp. Sta. Bui. 329. 1925. 31. NEVENS, W. B. 111. Agr. Exp. Sta. Ann. Rpt. 1924-25, 90. 32. KAMMLADE, W. G. 111. Agr. Exp. Sta. Ann. Rpt. 1923-24, 80; 1924-25, 75. 33. HIRSCHLER, A. Hoppe-Seyler's Ztschr. Physiol. Chem. 10, 306. 1886. 34. GOTTWALD, G. Jour. Landw. 36, 325. 1888. KELLNER, O. Die Ernsehrung der Landwirtschaftlichen Nutztiere,. 48, 51. 1912. 35. KUEHN, G. Landw. Vers. Sta. 44, 470. 1894. 36. ARMSBY, H. P. Principles of animal nutrition, 10. Wiley and Sons. 1910. 37. FORBES, E. B. Mineral nutrient requirements of farm animals. Natl. Research Council, Reprint and Circ. Ser. 60. 1924. 38. ECKLES, C. H. Mo. Agr. Exp. Sta. Res. Bui. 4. 1911. 39. MUMFORD, H. W., GRINDLEY, H. S., HALL, L. D., AND EMMETT, A. D. 111. Agr. Exp. Sta. Bui. 172. 1914. 40. SNYDER, H. U. S. Dept. Agr. Buls. 126, 1914; and 156, 1915. 41. EWING, P. V., AND WELLS, C. A. Georgia Agr. Exp. Sta. Bui. 115. 1915. 42. ATWATER, W. 0. Conn. (Storrs) Agr. Exp. Sta. Rpt. 1901, 244. 43. RUBNER, MAX. Arch. Anat. u. Physiol.: Physiol. Abt. 135-182, 1918. Zentbl. Biochem. u. Biophys. 21, 319, 1920. 44. ATWATER, W. O. Conn. (Storrs) Agr. Exp. Sta. Rpt. 1904, 183. 45. EWING, P. V., WELLS, C. A., AND SMITH, F. H. Georgia Agr. Exp. Sta. Bui. 125. 1917. 46. DIETRICH, W., AND GRINDLEY, H. S. 111. Agr. Exp. Sta. Bui. 170. 1914. 47. ARMSBY, H. P. The nutrition of farm animals, 650. Macmillan. 1917. 48. CHRISTENSEN, F. W., AND HOPPER, T. H. Amer. Soc. Anim. Prod. Proc. 39. Nov. 1925. 49. ARMSBY, H. P. The nutrition of farm animals, Table 188, 642-644. Macmillan. 1917. SOYBEAN PRODUCTS FOR SHEEP 283 APPENDIX TABLE 20. NITROGEN BALANCES: INVESTIGATION OF 1923 Sheep No. Nitrogen in feed consumed Nitrogen in feces Nitrogen in urine Nitrogen balance Period I Soybean hay 130. .. grams 165.3 grams 51.68 grams 92.0 grams +21 62 131 159.2 49.92 86 4-23 28 132 151.0 41.12 91.0 +18 88 133. . , 151.5 48.00 96.0 + 7 50 134 159.8 55.84 99 + 4 96 135 149.9 45.44 95.0 + 9 46 137 166.4 59.68 108.0 1 28 138 151.5 47.68 99.0 + 4 82 139 159.8 51.52 109 72 136. .. 171.0 51.52 106.0 +13 48 140 177.0 52.48 117.0 + 7 52 141 176.5 47.68 109.0 + 19 82 Average 161.6 50.21 100.6 +10.78 Period II Soybean straw 130 34.6 31.36 23.0 19.76 131 31.0 33.12 26.0 28 12 132 25.9 21.28 27.0 22 38 133 46.1 33.92 22.0 9 82 134 48.5 40.64 23.0 15.14 135 41.0 28.48 25.0 12 48 136. . . 42.9 37.92 25.0 20.02 137 43.8 35.04 28.0 19 24 138 37.4 33.28 22.0 17.88 139 39.5 35.68 23.0 19.18 140 38.9 37.12 30.0 28.22 141 . 38.6 29.28 20.0 10.68 Average 39.02 33.09 24.50 -18.58 Period III Oat straw and whole soybeans 130 grams 134 grams 45 grams 99 grams -10 131 132 40 97 5 132 136 50 104 -18 133. . . 138 43 94 + 1 134 . 126 40 96 -10 135 125 33 97 5 136 135 36 80 +19 137 133 36 84 + 13 138 132 31 72 +29 139 136 36 95 + 5 140 134 35 93 + 6 141 135 36 90 + 9 Average 133 38 92 + 3 284 BULLETIN No. 303 [A/arc/i, TABLE 20. Concluded Sheep No. Nitrogen in feed consumed Nitrogen in feces Nitrogen in urine Nitrogen balance Period IV Oat straw 130 41 39 26 -24 131 44 36 25 17 132 47 38 28 19 133 44 38 29 23 134 44 39 28 23 135 33 24 25 16 136 33 34 24 25 137 31 33 25 27 138 31 28 27 24 139 42 35 26 -19 140 39 34 27 22 141 33 31 26 24 Average 39 34 26 -22 Period V Soybean straw and soybean oil meal 130 . . . 137 46 88 + 3 131 138 44 96 - 2 132 134 40 89 + 5 133 136 42 85 + 9 134 135 46 86 + 3 135 132 42 71 +19 136 127 43 61 +23 137 126 38 80 + 8 138 126 40 83 + 3 139. .. 141 46 80 + 15 140 . 138 42 92 4- 4 141 140 42 92 + 6 Average 134 43 84 + 8 1928} SOYBEAN PRODUCTS FOR SHEEP 285 R3. TABLE 21. NITROGEN BALANCES: INVESTIGATION OF 1925 Sheep No. Nitrogen in feed consumed Nitrogen in feces Nitrogen in urine Nitrogen balance Period VI Alfalfa hay. 800 grams Bl.. grams 25.20 grams 5.58 grams 15.00 drams + 4 62 B2. .. 25.20 5.17 14 43 + 5 60 B3 25.20 4.94 15.94 + 4 32 Bl.. 24.40 5.95 13.98 + 4 47 R2 24.40 5 44 14 29 + 4 67 R3 15.25 3.57 10 63 + 1 05 Average 23.28 5.11 14.05 + 4.12 Period VII Alfalfa hay, 800 grams, and whole soybeans, 225 grams Bl.. 36.85 7.52 23 56 + 5 77 B2... 36.85 6.70 23.46 -J- 6.69 B3 36.85 6 82 23 46 + 6 57 Bl.. 39.89 7.87 23 31 + 8 71 R2. .. 39.89 7.12 24.00 + 8.77 R3 39.89 6.77 23.87 + 9.25 Average 38.37 7.13 23.61 + 7.63 Period VIII Alfalfa hay, 800 grams, and soybean oil meal, 112 grams Bl.. 37.63 6.78 20.77 +10.08 B2. .. 37.63 6.40 19.50 +11.73 B3 37.63 6.30 20.16 +11.17 Bl.. 28.77 7.89 19.32 + 1.56 R2 27.54 7.34 17.82 + 2.38 R3 27.54 7 23 20.46 .15 Average 32.79 6.99 19.67 + 6.13 Period IX Alfalfa hay, 800 grams, and soybean oil meal, 340 grams Bl. . 41.34 8.18 29.64 + 3.52 B2... 41.34 8.54 28.47 + 4.33 B3 41 34 8 26 32.76 + .32 Bl.. 41 68 10 34 27.28 + 4.06 R2... 37.62 8.51 25.55 + 3.56 R3 37.62 7.86 24.42 + 5.34 Average 40.16 8.62 28.02 + 3.52 Period X Soybean oil meal, 450 grams 28.22 2.64 23.01 + 2.57 286 BULLETIN No. 303 [Af arch, S o I \ * s XI 1 collection period 1- -, t- - 00 00 M IO 00 2 a ^43 00 >o sss CSCOIN coco-* (NCOO T( CN CO CO t^ CO ' ^ 0> IN pt) M -S a CD 3 J 2 IN 00 "5 ICO 00 1C CO CO CO CO CO CO CD CO . CO CO CO CN CM CO^ !* CO "5 rJ-H cOujcO CO a 8 a I 0 o 02 co OOU5O S 0500 O O CO O O US CO Til o> P 6 y il o a a ^ J3J3 > O.O.O. CD CD CD CO CO CO a a a CD CD CD CD CD CD CO CO CO a a a CD CD CD CD CD CD CO CO CO CD CD CD JS CD CD g 434343 > CO CO CO < Iff J343 > coco-J 3 a X CD ^ n J a O ^^ g * 1 3 2 ^J - a 9J 8 j- a | 43 3 c II Daily ration s a So CD"* 1,200 grams 16 pounds 1 1 So 910 grams 1,025 grams 1,135 grams a 6 i S 5 2 CD -' IS i 3 average (Ame ds, 20 rumina o . 213 P H 11 fl 43 OJ 1 1 1 1 12 a c a a a a CH C3 2 d |l! co II 11 11 I 1 "I CO 1 CO I Is CO CO ,2 i CO CD 1 S -S a o g SI CO NUSCO CO CO CO CO us CO OO'^'H co Cost- i.'S .S2 cooto* -1 tOSOtO U5US co -H O US CM f.~.f. S S wg 5 C " oo 00 US US os us - 3 Ji g OONTjl OOU5^H rtto-^ I'tOt^CO -r . - -r us us to us us us us us us us Soca> t^cco co co co CD CO t* CO us r~t~t t- t>. o '5 o B U5 US o US <) O CO 1 "! 0^00 (OtOO (NO- OUST}<-^ * NO-"' t-o; CO O.! * * US TJ. fi l ^^^S5 ^H(NO CO^^O OX)OO ^ us US us ooo CO OOCOCN * t.0000 "3 'S O ^ C^ CO ^ iO CO I s - CO O3 O *^ rococo >Mco cococo CO^T*'^ ae.Q. aao. aaa aaa^ oa>o CJ O 0)0)0) 0)OO)X j=JSj= J3J5J3 -fl^J5 XJ=J=> o o 3* o o ^ III : co-* g, : til jjjg > & & -S Ifltfl'-fl "/. "A "/. COMCQ COCOCC 1 ^ !>;?< !>!>< GQGQ-< CO CQ "' -a S 1 2 5 2 ^ S > c^ O -i? C c 5 - S-S G i i 5 S '3 "S It I 1 - 0) S ol & 3 -? 9 D l_ W)^ 9 > F* *~* $ t-t &C r? ^* C^ 'oo" s^ O - >> S > s > -S T3 * I ^ rt *! "si S c Q i fe"o - Is* -" * to >* 2 | ro - 1 Sf Si "=g2 ""* Q d Q 5 rt c c-S i "O o Sjg 1 (f f 1 1 o^o 02 02 1 03 CO OQ CQ * ^ '5' . o CO * US us -.! g : Ssg f ff 0? g g i-l o >* & O O 288 BULLETIN No. 303 [March, 8 a M M i 5P 5P *^ S S t~ t- o c S 38 a o .0 CO ' O f fe 2- CN OSCC-* CNiCcO b-iOI>-00 CCOO CCO'-* O^COGC 00^00 Ot>-OO5 T^ i- 1 O *O IHQQ CO >-< O ^ SO CO C 1 ! 00 CR 00 O5 l> *-<-* O ca 1 C^C^COl "^1 rHi^t-H 111"""' II III' 1 III ' 1 hi "t^ O CO .1! 00 COi-ICO O5 00 Oi 00000000 OOt*- O^t^OO t*- O5 O 00 OC C5 CO Ci C; O C> 5 wg t~ o i ?3 00 CO Oi i i CD O500O3 O CD 00 coSSS cot; S5S co^S5 oof;* SSSoo CO 1 1 w <= o 10 CN OS S 13 'S O COO3-H lOOOi-H 00 I 2 O> O500O5 O500OS 0000000) SoO 0.00) OS05CSOJ 000000 00010300 00 S a 03 S ft a 6S kl &s o ^D 00 O ^ Q S o "'br'* OOi- S 1 CD O'-^'MCO TJ |g o Jg ^ ?. s s CO o ^ >"a S a M B"o * i * ^ o - z B B " B SS c- c j-r a a" 05 B C 3 03 . 4) >> o On ff J3 5? jQ fc< ^D ^ J3 tn O ^Q -^ co i-r CO ^"m >*>"{D ^ ^ >> O O O O CO CO CO CO o "of >- B IT i s o O5 S C) OJ T-< F-l *^ 1-1 1-H 8 1928} SOYBEAN PRODUCTS FOR SHEEP 289 Reference I t H.M.K. 'c X 'S i I _ collection period 00 - - 11 t^OOO CO & ft S * OS CN 00 CO CO o CNCM"-" CO1"O (NCO-^CO Ether extract o OS OS OS c oc 00 00 5' c- OS co >o os r^ CO CO ss 00 OS OS 00 coefficients N-free extract o to CO to CO CN OS eo c CM O o 0-* OS OS 00 cNCNCO CN- o> CO O2 O2 CO^'O CD t 00 O5O-< COCOCO COCOCO COTjH^fl^j 020202 020202 O2O2O2<1 js fa o O 1 -r OS r- CM OS CN o" 1 o .g 1 a" a a a" a" 2 O CQ CO ! o CJ J2 _a o _= S 1 " o A A a 4 ti o 10 11 ooo t-OOiO CNCN tN g,SS 22g |^r: S : oooci co co 05 CO i 1 1 1 ^Y * OJ o o IO O Sf 8i 158358 ss K - s ff s 6099 OcOOO COOcO s ^ 5 y 10 o 10 CO ^ fi cu j. S g t^COCO 00 00 CO Sod 'S i -g OJt^-O r- oo "5 oo o US o M S3 ot-co rococo ooo IO OS hi 00 00 Pf SCO CO t~00 98SS ffgg S^feS 3 CO 00 "3 E 'H itNCO KWffi ooo i-KNCO T-((MOO r-tiNCO CO P,P(D.c3 aO*Q< P,D.Do3 p, ill fcfe 1 5^1 " !> ** COOJCC CflCCC/2'*^ 02CCCC t/lCCCQ 1 *^ OS >>< (>!>, CNO . MI II II fc 53 C > ^ c3 w i as? '*" S S -" "of ft "3 5? "3~of- ^" .T3 p s| I| | | ^ g| a! - r3 03 ^"l"* ^ ^S "S o ^3 o M^" *3 ^3 *3e ** S c e J3>2 a a _of ^3 Sjgo jq C oj fi g 1^ r I'll 1 1,1 &-"* & &^ ra mm I" |^CN^ |1 || CO 173 a Q "S * d 10 lO o . s sTs S O3 O5 rH 1 I-H o S 1928} SOYBEAN PRODUCTS FOR SHEEP 291 s B * * * * * # -~ * * 41 z, & IL> o o Cx 00 o i "S o'E 2 S o US us us US 33 4> o a H-. US US 00 o oo us o US fjj 0^0 -I t^ t-!N^ T.!NNN -^0:0 O USCOO csco-i tOOh- 6* 1-1 us co CO O || : : : 3 SfeS t-OSMCOCS OPOOC<1 US 00 O"J- US UJU3US USTJ1U3 ^ s a 41 ?! US O O 1-1 O US o 41 SI to to to to co t-oot-r- r-cot- B O US co co 00 O US US O "3.2 *Tt o o > 020Q-< OO< a tt-& Ot- t. ~ t-l ^ 41 41 ^ 41 cJ aaag S 02 02 02 < OH JSJg > PQfcg, ? ~ r- J3J3 > 0302 1 g s ft 21 o CJ ^ - c !! 2 3 O |J , 2,340 grams 2 it a a a i rt . S !J . S 3333 O O O O aaaa S 41 > ~ & 1 '3 -'I "s^ 1 "3-S N" SSS* 3 .a - J$> 3 Q *" co -2^ oT o" T sT v (U 41-3 41 MS T3 .a^O li Ji II 2B 5 a Sdd If .33-0 S b o -sS TJ 00 al 1 ?* a ^2 ^ XI g J2 g> Oi >>C! >>;2 >-i >> CQ O2 O2 O2 c a a a ^^2 ,o.a 02 02 CC 02 Soybean s sume) B M - s'g UTS o >.-2 >.3 o o 02 03 Soybean fc (same ai seeds foi Soybean fc to fuU b ^.a "t? r* co us us to S-c o os os 00 i OS 00 ^g O 292 BULLETIN No. 303 [March, o G * * * * u s^ K ^ ill m ^ ^ CO m m g==a XJ 38 _ CO ^ (N O m co 1* 383 SS3 ? S5S5JS S SS -i cot^O OrocoOcN cococomo cocoot^cc Is o (N^"00 CO CO CO CN O O co in in co ^ co co co CO CD CO CO CO CO CO CO 1 oo co CO O CO CO oj .i; S tmco O5-HCOOSOO COOIX^ OOT)<-HT)< m cococo h-OOO-HO CNt~-*CNO B .2 S CO t- co co m CO CO 1g ". tfcoin cooo^t^co -*CNCNCN --000 t^. coco-* coi-ocot~ t~t~a>ox O ^ t>- TjlTjl * Q a 03 S m m &a CO CO CO CO CO JyS O o o P| cO co co cO CO CO CO CO CO CO CO Animal 1.3! cocci! 0.0.0,0, * So CO P .C.C.C.G > CO 02 02 CO < -COJCO g, o.ao,2 CO CO CO 2 CU CO CO ^ J3 J3 Jj III ! < mil IIHI .a 5 B 00 * < "^ o o 5 E i B 2 B 03 E M 1 1 c IN o ^ *2 G O *3 s co" M -S % a f "c" "c" a |f fl y ?, fc S-g & 'a Q 1 S b TJ Is If it "O fi T3 olS o ' n ^o ai "3 . I a B a gg B si "-' a c^ a B B B B 03 03 "Si fl >. II II |S ^ "I 2 "I 2 c8 C? i 1 c c CO 02 g Q Q CO ^ 5( o m o a 5 S-c 00 00 S 00 co co ** *" *"* T-H . 1-t I-l T-l o CO-* CO _ _ _ _ 11 CO COO'J'IN COIN 00 00 00 00 : - : - co oo o os fe o Ic -; i co o oc *t wg S B c-g O co' U5 00 V 3 1 2 - E*.E SS gsg ^*" 00 00 XX XCOCO t-xoo CON-*-* o o 12 gs fa O CD CD CO CD *O O CO* S o * NX CDOCO COt^cO CO OS OS COX CO COt* CO .5 Q t>& 11 "5 OS fa 00 "D id Q g CDOSINCS TjOOO OCO i I-l O tN5 corpus I^INt~.'0 S "3 D O IIM CO CO CO COCOCO j, S ao.o.3 1 0.0. n'o.* -UN tC - - -" M fa fa- - - - 0,0, ag J3J3J3 * -C-C _^_g >. 2 S > J=J=J= GQbQbQ'*^ QtlCC 1KC -'" .1? a o o B a -T* a 1-1 6 oo } a a. 93 1 affa CON 2 o ^-o a " " * . a g a * O ^3 rS O ~ H ^* ^J 2 3 2 .^, ^i os os o> os cs H B 1 'SS 1 S^ g| 1- 0)>4 >iJ3 _E B o M >>"5, . t- g=3S. X O CO "sjl XlNO OX-J o * 00 Q~o3 = coxa *** ZZSS S SS SSS o OOOO COCOCO COCOCOCO CO t X co co co c??^^ 3 S5SS g;g co^ 3-" 322 o I O.O.O. o.ao.g a -- --- --- aaag c.aa aaag a coo c co oSS SSS OOO OCUQ QJOO^ CO CO CO COCCOQ-< CO O330 XMGQ COOQCO C03[;32-o ooo "c a)** s II IN OCCOM tOOCOO OO3O5 CO oot^r^oo f^ SI o c o a i 11 O OOOOOO I*- 1*- t>> t* ooo oooooo Kf:g^ S a o u, I- 2 11 E b| QI 00 t^- co oo co i oo co 55 co Animal COGOGO CQGCGO-< IN CO PQKffl J2J3JZ GO GO GO il .5 "3 "3 c! "3 o ei -3 c 33 c o g 1 M i C3 S) CO Q ~ c v a Eg _ 60 Is Eg _ ao '5o 3 g^ 1* ft ^* 1^ O! 2'S