TEXAS AGRICULTURAL EXPERIMENT STATION R. D. LEWIS, Director, College Station, Texas ' 768 Influence of Moisture on Heating in Feeds 0.11 if 1- 16f 1 W53 The TEXAS AGRICULTURAL AND MECHANICAL COLLEGE SYSTEM cuss GILCHRIST, ChCIIECGHOT DIGEST Feed ingredients did not heat below a certain critical moist ’ level, which varied with different ingredients. Heating beca more rapid and intense as the moisture level increased from t_ critical level to that where maximum heating occurred. Ground oats, wheat and corn heated faster and to a hig temperature than the unground grains containing the same amo ’ of moisture. The number of molds in ingredients which heated w larger in every instance than it was in those which did not heat. j Mixtures of ‘corn meal and molasses did not heat when 15 a 20 percent of molasses containing 25.5 percent moisture were add to corn meal with 8.7 percent moisture. Corn meal containing 1 ? percent moisture heated; when this meal was mixed with 5, 1'0, and 20 percent molasses containing 27.4 percent moisture, all t mixtures heated. Heating occurred also when the same corn j was mixed with 5, 10, 15 and 20 percent molasses with 21.0 perc moisture. Mixtures containing molasses with 21.0 percent moist started to heat on the 12th day, while those containing mola‘ with 27.4 percent moisture began heating on the 7th day. 0 The addition of 10 and 20 percent molasses with 21.0, 25.2 - 30.7 percent moisture delayed heating in wheat bran containing 1 percent moisture. Molasses with 21.0 percent moisture delay heating longer than those with higher levels of moisture, and 1 percent of molasses delayed heating longer in every case than percent of molasses. Corn meal containing as much as 17.4 percent moisture w completely inhibited from heating for 42 days by the addition 0.3 percent of calcium propionate. The same meal without the ' hibitor heated rapidly to a high temperature. The addition of 0 0.15 and 0.20 percent calcium propionate delayed heating but , not prevent it. ‘t The problems involved in the heating processes are compl , To eliminate heating, standards for the moisture content of all " gredients used in feeds should be reevaluated. The absence g heating in molasses feeds will not be insured by establishing‘ standard for the moisture content of molasses alone. Standa for molasses for animal feeds should be based on the total inv sugar after inversion and on moisture. Moisture should be as 1' as practical. Some feed manufacturers are equipped to use n, lasses containing as little as 20 to 22 percent moisture, especi during the summer, but others are not equipped to use them .- less than 24 to 26 percent moisture at any time. Probably the o way a workable standard for molasses can be established is by y consideration of the problems involved by all the interested gro i, Influence of Moisture en Heating in Feeds J. V. Halick and L. H. Hichardsen* sPONTANEOUS HEATING in feed ingredients and mixed feeds ‘r a major problem in many areas 0f Texas and other states (Rich- s; son and Halick 1952). Spontaneous heating of various mater- j s may be caused by: biological activity or respiration of living terials; by the metabolic heat produced as the result of the iowth of microorganisms; or by chemical oxidation. The amount " heat produced by the biological activity or respiration of living aterials is relatively insignificant and does not cause serious fses under practical conditions. Investigators of heating in grains d other feed ingredients are agreed that the most destructive ating and deterioration processes are caused by the growth of '0 ds. The early literature on the influence of microorganisms in the ating of moist grains has been reviewed by several investigators, arlyle and Norman, 1946; Milner and Geddes, 1946ab; Oxley, 48; and Christensen, Olafson and Gleddes, 1949) and will not be viewed in detail in this bulletin. In a study with wheat main- 'ned under adiabatic conditions and containing 22 percent mois- l e, Milner, Christensen and Geddes (1947) concluded that the in- ase in temperature was directly correlated to the growth of molds ‘til a temperature of 1.25-131° F. (52-55° C.) was attained. At is temperature, the molds were killed and the heating ceased. ith wheat of a higher moisture content, it was possible for bac- ial growth to increase the temperature to 154-158° F. (68-7 0° C.) . 5 this temperature, bacteria were killed and the heating ceased i ess strictly controlled adiabatic conditions were maintained. p der theseconditions, the temperature continued to rise sponta- "usly due to nonbiological or chemical oxidation. Causes of Deterioration in Moldy Feeds _, Rapidly growing molds produce substances called enzymes, and se hydrolyze carbohydrates, fats and possibly proteins, thereby nging the composition of the feed. Some of the nutritive value 3a feed is lost when it heats, but practically all the damage as p- result of heating is caused by the action of enzymes produced * molds, rather than by the heat itself. Feeds which have molded y be consumed poorly or be refused by animals, but it is gener- y assumed that molds which usually grow on grains and feed in- dients are non-toxic. spectively, technician and professor, Department of Biochemistry and utrition. _4_ Procedure Used t0 Study Heating in Feeds The apparatus used to study heating in feeds is composed 0 l multiple point Electronik precision indicator, vacuum flasks (De ar) and a storage cabinet which can be controlled at a constant tel perature and relative humidity. Two kilos of an ingredient are r in the vacuum flasks along with a thermocouple wire attached the Electronik indicator. The flask containing the ingredient ‘ the thermocouple is placed in the storage cabinet, which is mai tained at 90° F. and a relative humidity of 7 0 percent. The app, atus used in these studies is shown in Figures 1 and 2. With t apparatus, it is possible to duplicate data and to determine t, causes of and the conditions necessary for heating in feedstu with a high degree of accuracy. a Critical Moisture Level The critical moisture level of a material is the moisture c‘ tent at which it is barely safe or barely unsafe from the growth molds and the accompanying heating. The number of days quired for various ingredients containing different amounts of mo, ture along with the maximum temperatures, together with the n i ber of days required to reach the maximum temperature and A number of molds per gram of material, are summarized in Table In the preliminary studies, the temperature of the ingredient in e" flask was recorded three times daily. However, one reading §¥§%§§§'§5§§§£ Figure 1. Precision “Electronik” temperature indicator and st0_ cabinet used in studies of heating in feed ingredients and mixed feeds. ‘ Figure 2. Inside 0f storage cabinet showing vacuum flasks (Dewar) hich hold the ingredients and mixed feeds during studies on heating. ufficient information for practical purposes, and only one reading as used in the later studies. Each ingredient tested so far heated a maximum temperature which varied with its moisture content, hen it cooled off gradually until the temperature had returned to hat of the storage cabinet. The cooling-off process required sev- Weeks, and tied up equipment which could be used for studying e rate of heating in other materials. A study of the rate of eating appeared to be of greater importance than the rate of cool- 'g off; therefore, to use the equipment available at maximum ca- city for heating studies, an ingredient which had heated to a swaximum was discontinued after the temperature had declined cntinuously for 3 or 4 consecutive days. Some samples which eated were taken for mold counts and moisture content 1O days fter the test was started, while others which did not heat were ken at 42 days. p Table 1 shows that an ingredient did not heat below a certain fiitical moisture level. It heated when the moisture content was i ‘ghtly above the critical level, but the time required for the heat- g to start and to reach the maximum temperature was relatively ng and the maximum temperature was relatively low. As the yoisture level increased, an ingredient started to heat and reached a ‘gh maximum temperature in a relatively short time. For exam- ground yellow corn meal with 12.3 percent moisture did not heat i 42 days. When the moisture was increased to 13.3 percent, it _6_ started to heat in 12 days and reached a maximum temperature 108° F. in 19 days. When the moisture was increased to 15.8 ~-__ cent, the corn meal started to heat in 3 days and reached a a imum temperature of 117° F. in 5 days. 1 Table 1 shows that the rate of heating and the maximum perature reached depended on the moisture content of an ingredi Even though different ingredients started to heat at different t; ture levels, the overall picture for the heating cycle was essenti, the same for all ingredients. - Table 1. Heating and number of molds in feed ingredients containing va amounts of moisture .< 0 days % days p Corn, whole 14.4 .... ._ 230,000‘ 15.0 16 95 34 1,633,333 17.8 6 107 19 18,333,333 Corn, ground p‘ 12.3 .... .. 33,67 13.3 12 108 19 3,566,002 14.7 6 115 9 3,000,000‘ 15.8 3 117 5 19,670,000, Milo, whole ~ 12.3 .... _. 350,000 13.2 20 94 24 11,066,666 14.2 10 96 24 17,833,3 _: 14.8 9 99 14 3,430, l c‘; 16.6 4 114 14 12,230, H. Milo, ground if 12.7 .... -. 36,7, 13.7 18 113 24 1,223, 8;, 14.3 11 116 14 3,000, I 14.9 7 120 9 6,800,0 Oats, whole - , 14.5 ---- -- 157, I 14.8 5 101 38 8,830, U’ 17.0 4 108 14 5,670,0_ 19.7 3 110 14 12,330,0 I» _ Oats, ground 11.5 .... .. 66,67 12.6 1s 109 2s 5,367, y; 14.9 4 117 7 8,600, i; 17.5 3 120 4 14,230,0,, Steamed bone meal 8.2 . . . _ . . . . _ . . . . . . . . . . _ . _ . . . . .- 9.3 7 103 8 10.2 4 113 5 Wheat, Whole 14.1 .... .- 14.8 4 100 7 15.7 4 101 5 16.7 3 105 5 17.5 3 106 5 _7_ Heating and number of molds in feed ingredients containing various amounts of moisture, continued Started Max. Reached heating, temp., max. temp., Nmlfnolds days % days pe gm‘ j, Wheat, ground 1'11. .... .. 22,000 y 12. 20 95 25 .................. -. ’ 13. 12 107 15 23,000 .514. 5 112 8 93,300 14. 3 114 5 123,000 15. 4 118 6 5,167,000 j _ Wheat bran 11.6 .... .. 16,000 _' 13.4 26 106 41 220,000 _ 14.0 21 109 29 3,670,000 i, 15.3 15 111 21 22,000,000 i 16.2 8 114 12 15,670,000 17.8 6 120 7 109,300,000 : Wheat shorts '- 12.3 .... .. 33,000 ; 13.0 31 109 38 1,347,000 i 13.5 28 106 36 2,500,000 ~ 14.0 18 109 24 7,966,000 . 17.6 5 118 6 13,067,000 a 19.3 4 122 5 44,670,000 _ The critical moisture level of a few ingredients are summar- w in Table 2. These were determined at a temperature of 90° F. Y approximate the conditions that occur in many areas when ting in feeds is most frequent. A relative humidity of 70 rcent was used because it was found that the moisture content the ingredients in the vacuum flasks did not change appreciably - this relative humidity during a storage period of 42 days. These ta show that ground and underground grains and grain by- ucts heat at levels of moisture lower than those which have ‘n accepted traditionally as safe for grains of higher grades. Mold and Bacterial Counts Mold counts were made on samples which were taken when the servations were discontinued. They were made by a modification A the method described by Bottomley, Christensen and Geddes, Table 2. Critical moisture level of feed ingredients : Ingredient Critical moisture % _ Bone meal 8.7 f Corn, whole 14.7 _i Corn, ground 13.0 Milo, whole 12.7 i Milo, ground 13.0 , Oats, whole 14.5 . Oats, ground 12.3 , Wheat, whole 14.3 Wheat, ground 12.0 ‘ Wheat bran 13.0 "g Wheat shorts 12.7 _3_ 1952. The modifications Were: potato-dextrose agar, adjusted » a pH of 3.5 with citric acid, was used in the culture medium inst , of malt-salt or malt-salt-boric acid media, which was recommen ' by these authors. In our experience, the potato-dextrose agar gag a count which was higher and more easily read. The plates w‘ incubated for 3 days at room temperature, and the number of m' colonies on each dish was determined by the use of a Quebec c010‘ counter. l For the serial dilutions, 48 ml. of a hot 0.15 percent agar so tion were autoclaved in a 4-oz. screw capped medicine bottle. Af - cooling, each bottle contained approximately 45 ml. of the solution. The original dilution of a sample was made by addin gms. of material to the agar solution (approximately 45 ml.) i ;__ medicine bottle. Each dilution was thoroughly shaken to obt a uniform suspension. Serial dilutions were made for each sam’ by adding 5 ml. of the original suspension and each subsequ dilution to the sterile agar solution in a medicine bottle. number of molds was determined in 3 dilutions of e-ach sam For culturing, 1 ml. of the suspension was pipetted into a ste- Petri dish and potato-dextrose agar was added. Each dilution run in triplicate and plates which contained less than 30 colon were rejected. In general, plates which contained more than 50? 60 colonies also were rejected, but when certain species of moi predominated, a larger number could be counted. ’ Thioglycolate agar (Baltimore Biological Laboratory) used as the culture medium for the bacterial counts. The p cedures used for the serial dilutions and culture of the bact were essentially the same as those used for the mold counts, the exception that the dilutent was 0.1 percent tryptone in 0; percent agar solution. Thioglycolate medium was used for bacterial counts because it gives the total number of bacteria, _ cluding anaerobes and aerobes (Pittman, 1946). i There were discrepancies in the values obtained for the num of molds in a few ingredients. For example, the ground ye ' corn reported in Table 1 was from the same lot as the Whole c The Whole corn contained 230,000 mold spores per gram, While same corn ground contained only 33,670. The values for .- counts were obtained on samples which had been in the heating _’ paratus. If an ingredient did not heat, it had been in the heat apparatus for at least 42 days before the sample was taken. I possible that mold spores actually increased on the whole corn i taining 14.4 percent moisture, even though the temperature _ not increase and there was no visible evidence that mold gro had occurred. For this reason, it is not interpreted that the a lute number of molds has any important significance. The im tant point is that, Within a series, the highest temperature :5 also the largest number of molds were obtained, in. general, in l; samples which contained the highest level of moisture. 7 _9__ p Both bacterial and mold counts were made on two samples of ne meal. One sample contained 7.0 percent moisture and did not eat. It contained 4,000 molds and 74,500 bacteria per gram. The ther sample contained 10.2 percent moisture and had heated to f13° F. in 5 days. It contained 316,000 molds and 81,500 bacteria gram. These data indicate that the growth of molds rather i an the growth of bacteria was the cause of the heating. Further 'dence which supports this conclusion is that the moisture con- ‘ent of all the ingredients studied was lower than that usually con- 'dered necessary" to support the growth of bacteria, and the tem- rature to which the ingredients heated was within the range ormally produced by the growth of molds. _ Since heating had ceased and the temperature had started to ecline before the samples were taken for mold counts, it is obvious p, at mold spores were not destroyed. The samples with the higher vels of moisture heated to a higher temperature than those with ss moisture. The temperature at which the samples stopped eating appears to be more closely related to the moisture content ‘i physical characteristic of the ingredient than to the absolute imperature produced. A Another important point is that ingredients which did not heat 2 ill contained a large number of mold spores. This fact empha- ‘i es the need of maintaining conditions in grains, feed ingredients 'd mixed feeds so that mold spores cannot germinate and grow. Heating in Ground and Whole Grains The fact that many feed ingredients heated at moisture levels hich ordinarily would be considered safe for grains, suggested that ; e heating cycle in ground materials might be different from that A the unground grains. In view of these observations, the heating cles in ground and whole oats, wheat and corn were compared. a data ‘for oats and wheat are summarized in Figure 3. Ground ts or wheat, with essentially the same amount of moisture as e whole grains heated more rapidly and to a much higher tem- rature than the whole grains. Ground oats heated to above 115° within 5 to 7 days, while the whole oats increased only a few de- sees in temperature during the test period. The course of heat- , in ground and whole wheat and in ground and whole corn was entially the same as that for oats. Under practical conditions, "‘ ole corn with 14.5 to 15.0 percent moisture might be relatively ‘Y e from heating, but the same corn probably would heat after it ¢ ground. Heating in Mixtures Containing Molasses A Surveys made by the Texas Agricultural Experiment Station ichardson and Halick, 1952) show that feeds containing molasses it more frequently than those containing any other ingredient. dies Were carried out to determine the influence of the moisture ._1()._ 120- "l. = Moisture Ground Ground l|.5 - 4. ° IIO - l 9 A lL. o 5 l é |05 - I53 .h 5 dOYS - " +-——-l . § |oo - a t" Whole I wig/hole . 95 ./.’.v -..,\ .\ . _. l! \.\_"_ |48 1% i’ \ 1’ r i 90 ' Oats Wheat Figure 3. Ground oats and wheat heated more rapidly and inte: than the Whole grains. 1 content of molasses on heating in mixtures of corn meal and i lasses. These data are summarized in Table 3. None of the n! tures heated when 15 or 20 percent 0f molasses containing 5 percent moisture was added to corn meal containing 8.7 pe moisture. A similar study was made with mixtures of corn h‘ that contained 13.2 percent moisture and 5, 10, 15 and 20 pe r- of molasses that contained 21.0 and 27.4 percent moisture. fortunately, the corn meal Without molasses started to heat i‘. days and reached a maximum temperature of 104° F. in 24 days}? the results show that the addition of molasses accelerated he“ under these conditions. When molasses were added, the mi i started to heat and reached higher maximum temperatures r; than the corn meal alone. Mixtures containing molasses Withj percent moisture started to heat in approximately 7 days, ~ mixtures containing molasses with 21.0 percent moisture did; begin to heat until the 12th day. Tests are in progress to dete { the maximum moisture content of corn meal that will be safe 4 it is mixed with various amounts of molasses containing diff amounts of moisture. Another test was carried out to determine the influenl molasses on heating when they were mixed with wheat bran‘ _11_ Table 3. Heating in mixtures of corn meal and molasses ‘Molasses llgglilsisgllllée Started Max. Reached . added, of mixture heating, Temp, max. temp., % % days . days Moisture content of molasses, 25.5% 0 8.7 15 11.2 20 11.5 Moisture content of molasses, 21% 0 13.2 18 104 24 5 13.8 13 110 16 10 14.4 11 118 14 15 14.6 12 117 15 20 14.8 13 118 7 l Moisture content of molasses, 27.4% 0 13.2 18 104 24 5 14.1 8 114 11 10 14.9 7 115 9 15 15.3 7 115 8 20 15.5 7 115 9 high in moisture. The moisture content of the molasses used as 21.0, 25.2, and 30.7 percent, and that of the bran was 17.2 per- ent. These results, are summarized in Figure 4. The bran without 1 olasses started to heat on the 4th day and reached a maximum mperature of 118° F. on the 6th day. The addition of molasses elayed heating, regardless of their moisture content, but the olasses with low moisture delayed heating longer than those with gigh moisture. Also 20 percent of molasses at every moisture level elayed heating longer than 10 percent of molasses. It is apparent ; at the effect of molasses on heating in mixtures containing mo- y:sses and another feed ingredient depend to a large extent on the oisture content of the ingredient. If the moisture content of the gredient (corn meal) is near the critical level the addition of olasses may hasten heating, but if the moisture content of the in- ’ edient (wheat bran) is high, the addition of molasses may delay _eating. In any case, these data show that a feed mixed with lOW oisture molasses will not heat as rapidly as one mixed with a high oisture molasses, but the use of a 10w moisture molasses alone will pot eliminate heating entirely. In the final analysis, the moisture ntent of the other ingredients are just as critical a factor as that .l molasses, and the moisture content of ingredients will have to w suchlthaic the total moisture in the mixed feed will be below the 'tical eve . I The Brix values shown in Figure 2 are for this particular sam- 'e of molasses, which was adjusted to contain 30.7, 25.2 and 21.0 rcent moisture. These Brix values would not necessarily be the me for another sample of molasses containing the same amount » moisture. Moisture was determined by the vacuum drying meth- ._12_ ¢>__-_.Wheat Bran Bran + I091, Molasses Bran + 20% Molasses I20- - i" A ./' n5 ‘I \‘ if, I \\ | ' ,' ‘- I;- n | | , I 2 I05- ' ' ' ~3- .' : : 2 l0O- i i " Q | 5 days | ! l- : l-“ii | l! 95- ; ‘I , : I 1 9o I l‘ I 30.7% 252% 2I.0% 75.8 ° 82.0’ 86.7’ ‘Z, = Moisture in ° = Briai Molasses Figure 4 Heating in wheat bran containing 17.2 percent moisture delayed by the addition of 10 or 2O percent of molasses. = Inhibitors A few studies were carried out to determine whether inhibi 1i could be used to prevent the growth of molds and heating in f a Propionates, particularly calcium propionate, are used to delay v growth of molds in certain products. The results of studies calcium propionate as an inhibitor of the growth of molds in __ meal are summarized in Table 4. It is seen that 0.3 percent ofgf cium propionate completely prevented heating in corn meal’. taining 13.3, 16.0 and 17.4 percent moisture for at least 42 n} Without the calcium propionate, the meal started to heat in 2 a days and reached a maximum temperature in 3 to 19 days. time required "for the corn meal which did not contain propiona heat depended on its moisture content. Calcium propionate at l of 0.1, 0.15 and 0.2 percent delayed heating, but a level o». percent was required to prevent the growth of molds and hear, The mold counts show that the calcium propionate did not kill§ mold spores, but simply prevented their germination and growt ._..]_3._ Table 4. Calcium propionate as an inhibitor of the growth 0f molds poisture Calcium Started Max. Reached N 0 m o1 d incorn propionate, heating, temp. max. temp., ér m S v- eal,% % days °F. days p g ‘ f, 13.3 0 12 108 19 3,570,000 _ ' 0.30 ____ .. 28,300 j; 16.0 0 3 117 5 19,670,000 g 0.30 .... _. 10,000 ‘ 17.4 0 2 126 3 66,670,000 , 0.30 .... .. 23,000 a 15.3 0.10 13 112 17 30,670,000 s 0.15 25 110 32 31,000,000 0.20 37 100 48 28,000,000 Standard for Molasses There has been a demand during the past 2 years for a stan- , d for molasses used in animal feeds. The data described in this t letin illustrate the complex nature of the problems involved in e heating process. To eliminate heating, standards for the mois- g e content of all ingredients used in feeds should be reevaluated. _,e absence of heating in molasses feeds will not be insured by es- lishing a standard for the moisture content of molasses alone. y standard for molasses for animal feeds should be based on tritive value. Since carbohydrate is the principal nutrient sup- ed by molasses, it should be the first nutrient considered. Mois- e as well as carbohydrate should be included in the standard, and i- maximum should be as low as is practical. Some feed manufac- efs are equipped to use molasses containing as little as 20 to 22 , cent moisture, especially" during the summer, while others are not ipped to use them with less than 24 to 26 percent moisture at y time. Probably the only way a workable standard can be ar- ed at is by joint consideration of the problems involved by all different interested groups. Acknowledgments V, This investigation was supported in part by a grant-in-aid from ‘eSouthwestern Sugar and Molasses Company, New York, through courtesy of A. I. Kaplan, president. ' , We are indebted to E. I. du Pont de Nemours and Company, mington, Del. through the courtesy of F. M. J ornlin, for generous plies of calcium propionate (Mycoban) and propionic acid. _14_ Literature Cited Bottonley, R. A., C. M. Christensen and W. F. Geddes, 1952. Grain stor studies. X. The influence of areation, time, and moisture content on i acidity, nonreducing sugars, and mold flora of stored yellow corn. Ce i Chem., 29:53. i Carlyle, R. E. and A. G. Norman, 1941. Microbial thermogenesis in the dec position of plant materials. Part II. Factors involved. J. Bact. 41:_ Christensen, C. M., U. H. Olafson and W. F. Geddes, 1949. Grain stori“ studies. VII. Relation of molds in moist stored cottonseed to incr‘ production of carbon dioxide. Cereal Chem. 26:109. Milner, M., and W. F. Geddes, 1946a. Grain storage studies. III. The I tion between moisture content, mold growth and respiration of soybe Cereal Chem. 23:225. b, Milner, M., C. M. Christensen and W. F. Geddes, 1947. Grain storage stu VI. Wheat respiration in relation to moisture content, mold growth, ch; ical deterioration and heating. Cereal Chem. 24:182. a Milner, M., and W. F. Geddes, 1946b. Grain storage studies. IV. Biolo and chemical factors involved in spontaneous heating of soybeans. Ce Chem. 23:449. I Nagel, C. M., and G. Semeniuk, 1949. Some mold induced changes in sh corn. Plant. Physiol. 22:20. ; Oxley, T. A., 1948. The scientific principles of grain storage. The Nort Publishing Co. Ltd. Liverpool. y‘ Pittman, M., 1946. A study of fluid thioglycolate medium for the ste: test. J. Bact. 51:19. Richardson, L. R., and J. V. Halick, 1952. Moisture in molasses as a f, in the heating of feeds. Texas Agr. Expt. Sta. Bul. 754. 7