CA1314754C - Livestock feeding - Google Patents

Abstract

LIVESTOCK FEEDING

ABSTRACT OF THE DISCLOSURE
To increase the efficiency of utilizing protein in feed by ruminants, feed containing a protein and a reducing sugar are mixed in quantities suitable for the Maillard reaction. The mixture is heated at a temperature, pH and time sufficient to cause early Maillard reactions but not advanced Maillard reactions. Preferably, the sugar is xylose obtained by mixing sulfite liquor with the feed.

Description

1 3 ~ ~ r(J ~ 4c L IVRSTOCK FEEDI NG

This invention relates to livestock, and more particularly to a ]ivestock feed, the preparation of a livestock feed and the feeding of livestock to increase utilization of protein by ruminants~
It is known to treat feed for ruminants to reduce the microbia] degradation of fed protein in the rumen. Various prior art methods of treating feed to reduce the microbial degradation of proteins have included (1) chemical treatment with tannin, (2i chemical treatment with formaldehyde, (3) heat treatment, (4) addition of spent sulfite liquor and (5) pelleting with calcium lignosulfonate.
Chemical treatment of feed with tannin is dis-closed in United States Patent 3,507,662. This patent discloses a process for protecting proteinaceous animal feed ~rom rumen degradation by treatment of the feed with water and tanning agents, forming a paste, and drying at a temperature not to exceed 80 degrees centigrade. Subsequent work by Driedger (1972~ J. Anim. Sci. 34:465 showed that tannin could be added to feed prior to pelleting, eliminating the paste forming step, and still effectively protect the protein from rumen 2 ~311 ~7~

degradation. Driedger used 10 percent tannin on soybean meal. Tannins, however, are subject to irre~ersible oxidative condensation which can render the protein unavai]able in the abomassum tFergusSon, 1975, page 453 in ~igestion and Metabolism in the Ruminant, Univ. New England Publ. Unit, Armidale, New South Wales, Aust.), and are not widely commercia]ly accepted for use in feed ~reatment to protect protein.
Chemical treatment of feed with formaldehyde is shown in United States Patent 3,619,200. This patent discloses a feed for ruminants composed of proteinaceou~ material protected from rumen degradation by chemical modification of the protein through treatment with formaldehyde. Formaldehyde reacts with amino groups at neutral pH to form methylol groups which further condenses to form methylene bridges. In the acid pH of the abomassum, this reaction reverses, ren~ering the protein available and liberating formaldehyde (Fergusson, 1975). ~lemsley, 1973, Australian J. Biol. Sci.
26:961 reported optimum treatment to be 0.8 to 1.2 percent formaldehyde. Higher leve]s would over-protect the protein and reduce nitrogen retention.
Crawford, 1384, J. Dairy Sci. 67:1945 reported that .

3 ~31~7~

the optimum treatment ]evel will vary depending on the rate of passage of the feed through the rumen.
Since this is highly variable, it may be difficult to use formaldehyde effectively, and in fact formaldehyde is not approved for use in feeds in the United States by the Federal Drug AdministratiOn.
Heat treatment of feed is shown in United States Patent 3,695,891. I-Ieating proteinaceous feeds reduces degradabi]ity by reducing protein solubility and by blocking sites of enzyme attack through chemical modification. The reaction, however, is sensitive, and too little heat will not provide protection while too much heat will render the protein undigestible in the lower digestive tract (Sherrod, 1964, J. Anim. Sci. 23:510, and Plegge, 1982, J. ~nim. Sci. 55:395).
Addition of spent sulfite liquor to feed is shown in Larsen, United States Patent 4,377,576.
Larsen discloses a method of feeding high producing dairy cows with a feed containing spent sulfite liquor in an amount of 0.25 - 3.0 percent by weight of the feed to increase milk production. The feed and spent sulfite liquor of Larsen is merely mixed together in a blender without any additional pro-cessinq prior to feeding dairy cows. Larsen specu-` ' . .

. .
.

~3~7~

lated that t~e ]ignin present in the spent sulfite liquor operated to protect the proteins in the feed from being destroyed by microorganisms present in the cow's first three stomachs. Additionally, Larsen speculated that the wood sugars in spent sulfite liquor may assist in better digestion of the materials present in the grains and roughage commonly found in feeds. However, as taught herein it has been shown that the lignin present in the spent sulfite liquor does not operate to protect proteins from degradation by microbes in the rumen and the the wood sugars in spent sulfite liquor do not necessarily provide be~ter digestion of feed materials.
Pelleting feed with ca]cium lignosulfonate is shown in Stern, Can. J. ~nim. Sci. h4 (Suppl.): 27-28 (Sept. 1984). Based on continuous rumen culture in vitro studies Stern concluded that pelleting soybean meal with calcium lignosulfonate has poten-tial for protecting protein from microbial degrada-tion in the rumen. However, it has been discovered that calcium lignosulfonate is not the active compo~
nent in spent sulfite liquor that protects the protein, and in fact pelleting with calcium ligno-sulfonate per se results in no protein protection.

~4 ~

The prior art methods described above may be economical under some circumstances but it is important to achieve the maximum cost saving and the best utilization of protein such as by increasing the efficiency with which fed protein is used by the animal. The prior art feeds and methods fall short of these goals by, in some cases, providing protein which has reduced nutritional value in an effort to increase the amount of protein actually transferred from the rumen to the small intestine of ruminants or have other disadvantages.
For example, in the prior art use of calcium lignosulfonate and/or spent sul~ite liquor with feed protein, it was not understood that: (1) the process requires reducing sugars; ~2) the temperature, pH, percent moisture and time of the reaction is criticalJ and/or (3) the reaction must not continue to a stage where the resulting product is not utilized effectively in the small intestine of a ruminant.
It is known to supplement high protein animal feeds with carbohydrates including sugars.
In accordance with the invention a feed for animals comprises a mixture of organic materials including at least one reaction product of a feed .

- " -.. ~ . . . . .

6 ~ 3~7~

protein and a reducing carbohydrate, the percentage of reducing carbohydrate on feed protein is about 0.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen micro-organisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract. Advantageously, feed protein is selected from a group consisting of soybean meal, other bean meal, cottonseed meal, feather meal~ blood mea], silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early b.loom legumes, fish pro-ducts, by-product protein feedstuffs like distillers and brewers grains, milk products, poultry products, hays, corn, wheat, alfalfa, barley~ milo, sorghum, and mixtures thereof and reducing carbohydrate is selected from a group consisting of sugar sources are selected from the reducing sugars xylose, glucose, fructose, mannose, lactose, ribose, hemicellu].ose extracts and their hydrolysates, sugars contained in spent sulfite liquor, molasses and its hydrolysate and corn products and their hydrolysates, and mixtures thereof.

The reducing carbohydrate may be xylose and the percentage of xylose to feed protein be about 1 ~3~ ~7~

percen~ to 6 percent or the reducing carbohydrate may be glucose ancl the percentage of glucose on feed protein is about 2 percent to about 20 percent.
Preferably, the reducing carbohydrate is a component of spent sulfite liquor or dried spent sulfite liquor.
The reducing carbohydrate is a component of spent sulfite liquor or dried sulfite liquor and the spent sulfite liquor or dried spent sulEite ]iquor includes about 10 percent to about 40 percent reducing carbohydrates on solids and the percentage of spent sulfite liquor solids on feed protein is about ~ percent to about 40 percent.
Advantageously, the feed includes at least one reaction product of feed protein and spent sulfite liquor or dried spent sulfite liquor, the percentage of spent sulfite liquor so]ids on feed is about 3.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen micro-organisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract. The percentage of spent sulfite liquor solids on feed protein is about 8 percent to about 25 percent by weight. ~he spent sulfite liquor or ~ 3 ~

dried spent sulfi~e liquor is obtained from the pulping of hardwoods.
A methocl of making an animal feed comprises the steps oF: providing a mixture of a feed protein and a reducing carbohydrate, the percentage o reducing carbohydrate on feed pro~ein being about 0.5 percent to about 40 percent by weight; and heating the mixture at a temperature, pH and percent moisture for a time sufficient to reduce the degradability of the feed protein by rumen microorganisms and provide no significant reduction in protein digestibility in the post rumen tract. Advantageously, the pH is from about 4 to about 1005, said percent moisture is from about 6 percent to about 40 percent, the temperature is from about 20 degrees centrigrade to about 150 degrees centrigrade and said time is from about 20 minutes to about 72 hours.
In one embodiment, the p~ is from about 6 to about 8.5, said percent moisture is from about 15 percent to about 25 percent and said temperature is from about 80 degrees Celsius to about 110 degrees Celsius. The said time may also be from about 1 hour to about 4 hours.

A~vantageously, the feed protein is selected from a group cons;sting of soybean meal, other bean ~ 3.~

meal~ cottonseed mea]., feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early bloom legumes, fish pro-ducts, by-product protein feedstuffs like distillers and brewers grains, mi~.k products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum, and mixtures thereof and the reflucing carbohydrate is selected from a group consisting of sugar sources are selected from the reducing sugars xylose, glu-cose, fru~tose, mannose, lactose, ribose, hemlce.llulose extracts and their hydrolysateS, sugars contained in spent sulfite liquor, molasses and its hydrolysate and corn products and their hydrolysates, and mixtures thereof. A mixture of a feed protein and a spen~ sulfite liquor or a dried spent sulfite liquor is provided such that the per-centage of spent sulfite liquor solids on feed pro-tein is about 2 percent to about 40 percent by weisht; and heating the mixture at a temperature, pH
and percent moisture for a time sufficient to reduce the degradability of the feed protein by rumen microorganisms and provide no significant reduction in protein digestibility in the post rumen tract.
The percentage of spent sulfite liquor solids on ~ 3 1 ~

feed protein is about 8 percent to about 25 percent by weight. The spent sulfite .liquor or dried spent sulfite liquor is obtained from the pulping of hardwoods.
A method of feeding animals comprises ~he steps of: selecting a protein-containing feed; and feeding to the aminals a reaction product of the feed protein and a reducing carbohydrate wherein the percentage of reducing carbohydrate or feed protein is about 0.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen microorganisms is reduced and there is no signifi-cant reduction of protein digestibility in the post rumen tract. Advantageously, the feed protein is selected from a group consisting of soybean meal, other bean meal, cottonseed meal, feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early bloom legumes, fish products, by-product protein feedstuffs like dis-tillers and brewers grains, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum, and mixtures thereof and said reducing carbohydrate is selected from a group consisting of sugar sources selected from the reducing sugars 1~ ~3~7~

xylose, glucose, Eructose, mannose, lactose, ribose, hemicellulose extracts and their hydrolysates~
sugars contained in spent sulEite liquor, molasses and its hydro]ysa~e and corn products and their hydrolysates, and mixtures thereof. A protein-containing feed suitable for a ruminant is selected;
and the ruminant is fed a reaction profluct of the feed protein and a spent sulfite liquor or a dried spent sulfite liquor wherein the percenta~e of spent sulfite liquor solids on feed protein is abou~ 2 percent to about 40 percent by weight such that degradability of the feed protein by rumen micro-organisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract The percentage of spent sulfite liquor solids on feed protein is about 8 percent to about 25 percent by weight and the spent sulfite liquor or dried spent sulfite liquor is obtained from the pulping of softwoods.
This improved protein feed may be substituted for a part or all of the usual protein feed being supplied to the animal, resulting in improved efficiency of milk, meat and/or wool production.
Specifically, increased production yields may be obtained with same feed protein levels; or same 12 ~3~

production yields may be obtained at reduced fee~
protein levels.
As can be understood from the above and following descriptions, the novel feed, method of making the feed and method of feeding animals has the advantage of providing a superior economical feed and method of feeding animals.
The above noted and other features of the in-vention will be better understood from the following detailed description when sonsidered with reference to the accompanying drawings in which:
FIG. 1 is a graph i~lustratinq the results oE
in vitro tests indicating the reduciton in microbial degradation of protein in accordance with an aspect of the invention FIG. 2 is a qraph illustratinq the results of in vitro tests indicating the reduction in microbial degradation by treatment with reducing sugars related to ratio of reducing suqar to protein in 2Q accordance with an aspect of the invention FIG. 3 is a graph illustrating the results of ln vitro tests indicating the effect of heatinq time of eed during preparation with ratios of reducing sugar to protein on microbial degradation l3 ~ 3 ~

FIG. 4 is a graph illustrat.ing the results of ln vitro tests indicating the effect of heating time on the preparation of feeds using several reducing sugars;
FIG. 5 is a graph illustrating the effect of preparation in accordance with the invention on commercial and untoasted soybean meal;
FIG. 6 is a graph illustrating the effect of p~
on preparation of feeds in accordance with the in-vention;
FIG. 7 is a graph illustrating the effect of dry matter on the preparation of a feed in accordance with the invention;
FIG. 8 is a graph illustrating the protein efficiency of feed treated in accordance with the invention;
FIG. 9 is another graph illustrating protein efficiency of feed treated in accordance with the invention;
FIG. 10 is a graph illustrating the dependency of carbohydrate content on the effectivenesS of sulfite liquor as an additive to feeds;
FIG. 11 is a graph illustrating the use of sulfite liquor as an additive to feed in accordance with the invention;

.
- ;

7 ~ ~

FIG. 12 is a graph illustrating the stability of feed made in accordance with the invention;
FIG. 13 is a graph illustrating one aspec~ oE
the useful range of a reclucing sugar in accordance with the invention and FIG. 14 is another graph illustrating another aspect of the useful range of a reducing sugar in accordance with the inventionO
Broadly, the animal feed inc]udes a substantial lQ amount of reaction products of proteins and reducing carbohydrates. Because the more reactive the re-ducing carbohydrate is, the easier it is to form such reaction products, sugar sources are selected from the reducing sugars xylose, glucose fructose, mannose, lactose, ribose, hemicellulose extracts and their hydrolysates, sugars contained in spent sulfite liquor, molasses and its hydro.1.ysate and corn products and their hydrolysates, and mixtures thereof.
Generally, the proteins used are those found in high quality protein feed such as soybean meal, other bean meal, cottonseed meal, meat and bone meal, sunflower seed meal, canola seed meal, peanut meal, safflower meal, I.inseed meal, sesame meal, early bloom legumes, fish products, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum and the like and mixtures thereof. Preferably, the reducing sugars used are those from economical sugar sources such as spent sulfite liquor or dried spent sulfite liquor which îs a by-product of some wood industries and a source of xyloseO However, mixtures of sugars are some-times used.
In this specification, the term "orthodox feed"
means the feeds normally fed to ruminants. Such feeds are well-known in the art and include the high quality protein feeds described above and other feeds, which because they are not considered a high quality protein feed, are less likely to be used in the treatment. Such feeds included among othere, soybean mea], other bean meal, cottonseed meal, feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanu~ meal, saflower meal, linseed meal, sesame meal, early bloom legumes, ~ish products, by-product protein feedstuffs like distillers and brewers gainrs, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum and the like and mixtures thereof.

The particular feed may be se]ected for economic reasons or reasons of supply but, since the methods described herein are applicable to protein in general regardless of the feea, the steps in performing the method are the same although the actual reaction products may differ.
For reasons of economy, this process is intended prîncipa]ly for protein supplements. In this specification protein supplements are feed-stuffs containing a minimum of 20 percent protein with at least 25 percent of the protein being microbially degradable protein. Microbia3ly degradable protein in this specification is protein which is cleaved by microbial protease.
Similarly, by the term "reaction product of a sugar and a protein" when used in this specifi-cation~ means a condensation product obtained by reacting; (1) any protein useful in feeding live-stocks and commonly found in orthodox livestock feeds; and (2~ a reducing carbohydrate selected for its efficiency in reduction reaction with proteins.
Generally, it is believed that the reactions are reactions with amino groups redundant of the proteins and the carbony] groups of the reducing sugars. These reactions are well-known in the art.

17 ~ 3~

Similarly, suitable reducing carbohydrates are well-known and, generally, to shorten the time and reduce the temperature, the most reactive reducing carbohydrates are selected as described in this specification but under certain circumstances, other reducing carbohydrates may be selected.
~his improved feed may be prepared in several different ways u~ilizing different ones of the suit-able feeds and diferent ones of the reducing carbo-hydrates as raw materials. In each case, a reaction takes place between the sugar and proteins in the feed used as a raw material which reduces the degra-dation of the protein in the rumen of an animal by microbes and thus increasing the protein available for digestion in the small intestine of the animal.

With this product, there is less degradation of the protein and less conversion to other nitrogen compounds, such as ammonia, by ruminal microbes.
Most suitably, the feed material is mixed with a reducing sugar to maximize the reaction. The pH is selected along with temperature, percent moisture and time of treatment to maximize the production of compounds which resist degradation by ruminal microbes but nonetheless permit digestibi~ity and use of the protein in the post rumen tract.

~3~ 15~
1~

It is believed that the extent of the reaction in forming this feed corresponds to what has been described in the literature as the early Maillard reactions and comprise a condensation reaction between the carbonyl group of a reducing sugar and amino groups of the protein. The early Maillard reactions are well-known and from the detailed spe-cification herein, the pH, temperature, moisture and time required to carry the reaction to the optimum 10. extent can be de~ermined with little experimen-tation~
It is believed that the reaction is generally a 1 mo].e to 1 mole reaction between free amine groups and the reducing carbohydrate and with some consid-eration being given to other reactions in the feed, the quantities of sugars which are most economically utilized with the feed can be determined even though some suitable feed materials are not specifically described herein. The pl-~s shoul~ be about 4 to 2Q about 10.5 and preferably about 6 to about 8.5. The time and temperature and moisture offer more leeway since a lower temperature for a longer time may be used in some circumstances or a higher temperature for a shorter time where economy dictates.

:

~31~
1~

In general the temperature of ~he reaction ranges from about 20 degrees centigrade to about ~50 degrees centigrade with 80 degrees centigrade to 110 degrees centigrade preferred, and the time of the reactian ranges from about 20 minutes to about 72 hours with 1 hour to 4 hours preferred. The amount of water affects the reaction, and the percent mois-ture ranges from about 6 percent to about Q0 percentwith 15 percent to 25 percent preferred.

Without wishing to be bound by any particular theory, it is be]ieved the following description illustrates the reaction mechanisms involved between the proteins and reducing carbohydrates which result in the feed of the present inventiOn.
More speci~ically, it is believed a reducing sugar and a livestock feed containing protein are mixed in quantities sufficient to cause enough of the alpha and epsilon amino groups in the protein to react with the carbonyl groups in the sugar to form a reaction product when the mixture is heated at a temperature, time and pH to cause reactions corre-sponding to those in formula 1, where R is a protein having the alpha amino group or epsilon amino shown, ~1 is the remaining portion of the carbohydrate , ~ 31 ~
2~

shown in formula l; and R2 is a portion o Rl resulting from the reaction as shown.
If a simple reducing sugar is the reducing carbohydrate, it is believed that the reaction is shown in formula 2 where R is a pro~ein having the amino group shown, R3 is a methyl hydroxy uni~ which together with a]dehyde and keto groups are typica~
of a sugar, P is a number of the indicated functional groups and M is a number one group less than P. If the reducing sugar is glucose, it is believed that the reaction is shown in formula 3 in which the glucose reacts with an addition compound to result in a Schiff base which immediately pro-ceeds to glucosylamine.
The mixing of the reducing carbohydrate and feed is in proportions such as are suitable for the Maillard reaction and the mixture is heated at a temperature, pH, moisture level and time sufficient to cause early Maillard reactions but not advanced Maillard reactions. Thus, the time and temperature are selected to be sufficient to form glycosylamine but insufficient to form l-amino-l-deoxy-2-ketose.

1 3 ~

Ii'ORMULA 1 a P?U~ ~ ~5 Rl ~ ~ t ~ ~ ¦
Rl Rl 1~2 22 :L3:~ ~7~

~ORMUL~ 2 a I ~2 n ~C~I (R3 ) P (R3 ) P (R~O
If;~COll 112CO~I
COI~

23 ~ 3 ~

~ORMULA 3 ~0 ~ll ~ ~3 ~34~ C-OI~ ~ ~ C~

4 t~ 4 t~Coll~3 0 ~

1~:0 2~

Some epsilon amino groups are not avai],able for microbial acti,on hecause of inhibiting effects of other groups. These inhibiting effects may be due to the conformal structure of the protein or groups chemically bound in the vicinity. It is believed that the temperature a~ which the early Maill,ard reactions occur may affect such inhibiting by changing the conformal structure to increase or decrease hidden amino groups. The groups not available for reac~ion with microbia] protease are under some circumstances not available for reaction with the reducing sugar and may reduce the amount of sugar needed for some reactions. For example, the use of high temperatures for a short time may decrease the amount of sugar needed for the same final result in the effectiveness of the feed.
Generally, the feed is prepared by mixing a reducing sugar with a suitable protein containin9 feed at a desired percent moisture in a controlled 2Q ratio and applying temperature at a p~ and for a time suitable to cause early Maillard reactions but not so long as to cause advanced or final Maillard reactions~ Thus, condensation produc~s are formed between the carbonyl group of the reducing carbohydrate and a free amine group of an amino acid ' ' ' ~3~7~

or protein in a 1 to 1 ratio. The condensation product loses a molecule o~ water and is converted to a Schiff's base which, in turn, undergoes cyclization to the corresponding substituted sugar amine.
For example, when glucose is the ~ugar, the amino group is converted to a N-suhstituted glyco-sylamine. The reaction is terminated before there is a transition of the aldose sugar to a ketose sugar derivative by way of the Amadori rearrange-mentO In the case of glucose, this is a conversion of glycosylamine to a 1-amino-1-deoxy-2-ketose. As a further example, in the case of ketose sugars, the reaction is terminated before a rearrangement cor-responding to the Heyns rearrangement to form a 2-amino-2-deoxyaldose from the ketosylamine.
One source of reducing sugar is su]~ite liquor.
Spent sulfite liquor is that portion of the wood solubilized in the acid su]~ite pulping of plant materia]s, preferably hardwoods and/or softwoods.
The plant material is cooked at elevated tempera-tures at a p~T of less than pH 7 in a solution of MITSO3 where M is the cation which can include NH4~' Na+, Ca~, Mg~ and K~.

26 ~3~

The well-known process is commonly used in making cellulose pulp for the manufacture of paper products and/or rayon. Most of the cellulose is not dissolvea in the pulping process. The solubilized portion of the wood, spent sulfite liquor, contains a substantial portion of ~he starting wood, 20 to 70% and usually 40 to 60~. Because of pulp washing the spent sulfite liquor solids may range from about S% to about 20%. Such a solution can be used in the invention though concen~rated solutions at about 40~
solids to about 65% solids or dried spent sulfite liquor at about 90% to about ~00~ solids are pre-ferred.
Spent sulfite liquors are comprised mainly of M-lignosulfonates, about 40% to about 70~; reducing sugars, about 5~ to about 30%; and oligosaccharides from about 2~ to about 20%.
Spent sulfite liquor reducing sugars are a mixture comprised of glucose, mannose, xylose, 2Q ga]actose and arabinose. The relative proportions among the sugars vary depending upon the exact pulping conditions and the plant material used in the process. For example, spent su]fite liquor from the pulping of softwood typically contains about 6 parts of hexoses (6 carbon sugars) to 4 parts of 13~7~

pentoses (5 carbon sugars1 due to hydrolysis of gluco-mannan as the main hemicellulose in go~twoods.
Spent sulfite liquor from hardwood pulping typically contains about 7.5 parts of pentoses to about 2.5 parts of hexoses due to hydrolysis of xylan as the main hemicellulose in hardwoods.
The source of the protein is not significant as long as it is a protein suitable for livestock and such proteins are well-known. Similarly, any re-ducing carbohydrate may be used but some are more efficient than others. The most suitable reducing carbohydrates are those that are most reactive and include xylose, fructose, glucose and lactose with xylose being the most reactive. Generally, the pR
is controlled ~o be above 4 and below 10.5 and preferably at 6 to 8.5. The pH is controlled by any suitable method including the addition of sodium hydroxide.
In feeding livestock, at least 50 percent and under some circumstances a lon percent increase in the protein use efficiency may be taken into account and used either to increase the weight gain from protein limited diets or to reduce the cost of the feed, The treated feed material is intended primarily for ruminants and can be used accordingly ~ , - ' ~ 1 3 ~
2~

as a substitute for untrea~ed high-protein feed. In some cases, the corresponding untreated protein supplement that would otherwise be fed can be re-duced and the amount of treated protein feed supple-ment is less than the untreated protein supplement because of the increased protein use efficiency of the treated protein supplement.
While many of the variables can be selected by the users of this invention, the following non-limititive examples illustrate the invention.

EXAMPLES

1. Materials and Methods -Sodium hydroxide was added to soybean mea] to adjust pH in amounts determined as follows. Ten grams of soybean meal dry matter were weighed in triplicate and hydrated with 100 ml (milliliters) distilled deionized water. Hydrated samples were homogenized for 2 minutes at moderate speed with a blender and allowed to equilibrate or 2 hours at 21 degrees Celsius. Homogenates were titrated with standardized NaOH and pH changes monitored with a saturated calomel electrode. During titration, agitation of homogenates was maintained with a mag-13~1 ~7~

netic stir bar. Quantities of NaOH required to adjust pl~ to 8.5 or 10.0 were calculated asequivalents/g soybean meal dry matter.

2. In Vitro General ConditionS
Microbial degradation of treated soybean meal samples was the response variable in all trials and was measured by the in vitro ammonia release pro-cedure described by Brit~on, R.A. and T~J.
Klopfenstein. 1986. "Zinc treated soybean meal: A

method to increase bypassn. Nebraska seef Cattle Report, MP 50. rJniversity of Nebraska, Lincoln.

pp. 45-57.
Equal volumes of ruminal fluid were collected from steers fed maintenance diets of either ground alfalfa hay or ground corn cobs containing 13 percent molasses and 17 percent soybean meal (meal dry matter basis). Following fermentation for 24 hours, ammoniacal nitrogen was determined by an automated adaptation of the indophenol method of McCullough, J. 1967. "The determination of ammonia in whole blood by a direct colorimetric method".
Clin. Chim. Acta. 17:297.

~3~7~

3. In Vitro Examples An eva~uation was made of the main effects on protein of reducing sugars, time of heating and proportions of reducing sugar and protein. In these tests: (1) reducing sugar sources were xy]ose Pructose, glucose and lactose; (2) reducing sugar levels were at 1, 3 and 5 mol/mol lysine; and ~3) heating times at were 0, 30 and 90 minutes at 150 degrees Celsius. The interactions between main effects were also evaluated. Soybean meal samples were heated with p~ and moisture altered, but without reducing sugar, to estimate the effect of sugar additions.
In these tests, the protein fraction of soybean meal was assumed to contain 6.3 percent lysical in accordance with "Nutrient Requirements of Domestic Anlmals", 1979, No. 2, 'l~utrient Requirements of Swine". National Research Council. Washington, D.C.
Dehulled, solvent extracted soybean meal which had not passed through a desolventizer-toaster and was thus untoasted during processing was the soybean meal source and contained 53.0 percent crude protein on a dry matter basis.

31 ~ 13:L~75~L

Prior to heating, appropriate quanti~ies of the reducing sugars were adfled to untoasted soybean mea]
which had previously been treated with NaOH to achieve p~ 8.5. Distilled water was ad~ed so ~hat each sample contained 83 percent dry matter. Heated samples were obtained by placing 126 g (grams) samples in 9 cm (centimeters) x 12 cm x 5 cm aluminum pans and heating to 150 degrees Celsius in a forced air oven. Following heating, samples were cooled to 23 degrees Celsius, air dried for 72 hours and ground to pass through a 2 mm (millimeter~

screen This procedure for sample preparation after heating was followed in all subsequent experiments.

Prior to ammonia release analysis, sugar content, expressed as a percent of sample dry ma~ter, was made equal in all samp]es to eliminate confounding of ammonia release by reducing sugar concentration.

Previous results with commercial soybean meal as the protein source indicated ammonia release following a 24-hour fermentation was unaffected by source of reducing sugar when sugars were added on the same weight to weight ratio with soybean meal. Samples were analyzed in duplicate for ammonia release.
Contrast coefficients for the main effect of heating 1 31 ~

time were calcu~ated. ~he results are shown in FIGS. 1, 2 and 3 respectively.
In F~G. 1, there is shown a graph of ammonia nitrogen resease against heating time Eor reducing sugars in which curve 30 represents the interaction of fructose with time of heating, curve 32 repre-sents the interaction of xylose with time of heating, and curve 34 represents the interaction of lactose with time of heating. Curve 38 indicates lQ ammonia nitrogen released in the absence of re~ucing sugar for comparison.
In FIG. 2, there is shown a graph of ammonia nitrogen released against the number of moles of reducing sugar for each mole of lysine, with the curve 40 being for fractose, the curve 42 being for glucose, the curve 44 being for lactose, and the curve 46 being for xylose.
In FIG. 3~ there is shown a graph of ammonia nitrogen released against the ratio of moles of sugar for each mole of lysine for different heating times, In this graph, curve 50 is a control without heating, curve 52 is the amount of ammonia released for a preparation with 30 minutes heating, and curve 54 is the amount of ammonia released for a prepara-tion with 90 minutes of heating.

~ 3 ~

~XAMPLE 2 Effects on ammonia release of commercial soy-bean meal containing no sugar or reducing sugars (xylose, glucose, fructose or lactose), and unheated (23 degrees Celsius) or heated for 30 or 60 minutes at 150 degrees Celsius were studied. On a dry matter basis, soybean meal without sugar contained 46.5 percent crude protein. Sugars were added to soybean meal without sugar at 3 mol/mol lysine, pH
was adjusted to 8.5 and all samples contained 80 percent dry matter.
Pans containing samples for heating were pre-pared as described for example 1 except they were sealed with aluminum foil during heating. Following heating, sugar content was equalized in all samples prior to ammonia release analysis as described for example 1.
Samples were prepared in duplicate and each analyzed for ammonia release in duplicate in two ammonia release runs. ~ata were analyzed as a randomized complete block design with a 5 x 3 factorial arrangement of treatments, and run was the blocking criterion. When no block * sugar source *
heating time interaction was observed, this term was removed from the statistical model and data were ~3~7~
3~

analy~ed for maln effects an~ sugar so~rce, by heating time interactions. The results are shown in FIG. 4, whi.ch is a graph il].ustrating the effect of heating time in preparation of the feed on microbial degradation, with curves 60, 62, 64, 66 and 68 illustrating test respectively on: (I) a contro].
feed without a reducing sugar; (2~ a feed prepared with lactose, (3) a feed prepared with fructose; (4~
a feed prepared with g].ucose; and (5) a feed pre-pared with xylose.

EXA~PLE 3 Susceptibilities of commercial soybean mea]. or untoasted soybean meal to nonenzymatic browning as measured by ln vitro ammonia rel.ease were studied.
Each soybean meal was treated with NaOH to adjust pH
to 8.5, xylose at 3 mol/mol lysine and distil]ed water to attain 80 percent dry matter in each samp].e. Samples were unheated so as to be 23 degrees Celsius or were heated at 150 degrees ~elsius for 30 or 60 minutes in a forced air oven as described for example 20 Samples were prepared in duplicate and each analyzed for ammonia re].ease in duplicate in two ammonia release runs. Data were analyzed as a ` 34 35 ~ 3 ~

randomized complete block flesign with a 2 x 3 factorial arrangement of treatments with run as the blocking criterion. When no block by soybean meal source by heating time interaction was observed, this term was removed from the statistical model, and data were analyzed for main effects and soybean meal source by heating time interactions. The resu].ts are shown in FIG. 5, in which curve 70 is for un-toasted soybean meal and curve 72 is for commercial.
soybean meal.

EXAMP~E 4 Effects of pH at each of natura]. p~l, 8.5 p~
and 10.0 pH were measured when xylose was adfled to commercial soybean meal at a rate of 3 mo~/mol lysine and were unheated or heated for 20, 40 or 60 minutes at 150 degrees Celsius. The natural pH of commercial soybean meal homogenates prior to NaOH
addition was 6.5. Samples contained 80 percent dry matter. Heating procedures were the same as described for examp].e 2.
Samples were prepared in duplicate and each analyæed for ammonia release in duplicate in two ammonia re]ease euns. The data were analyzed as a randomized complete block design with a 3 x 3 ~31~75~

factorial arrangement of treatments and run was the blocking criterion. The data were analysed fcr main effects and pH by heating time interaction. ~he resu]ts are shown in FI~. 6, in which curves 74, 76 and 78 represent preparation at natural p~l, ph 8.5 and pH 10.0, respectively.

EXAMI?LE 5 Effects on ammonia release of percent dry matter (at 65, 70, 75~ 80, 85 and 90 percent~ of commercial soybean meal containing xylose in the quantity of 3 moles of xylose for each mole of lysine was measured for samples heated at 150 degrees Celsius for 30 minutes. The p}~ of samp]es was 8.5. Additionally, the effect of retaining moisture in the pans was evaluated by sealing ha]f the pans with aluminum foil.
Samples were prepared in duplicate and each analyzed for ammonia release in duplicate in two ammonia release runs. Data were analyzed as a randomized complete block design with a 6 x 2 factorial arrangement of treatments with run as the blocking criterion. The data were analyzed for main effects and dry matter level by covering interactions. The results are shown in FIG. 79 in which curves 80 and 82 il.l.ustrate the effect on dry matter when prepared in uncovered an~ covered pans, respective].y.

4. In Vitro Results As shown in FIG. 1, interactions among fructose, lactose and g1.ucose for the linear effect of heating are not signlficant. ~owever, an inter-action was noted when fructose, lactose and glucose were compared to xylose for the linear effect of heating time.
Without heating, addition of xylose suppressed ammonia rel.ease more than fructose, lactose and glucose indicating that xy].ose reacted faster with untoasted soybean meal at room temperature, under the existing cond;tions of p~l and moisture, than the other sugars. These data further suggest that, given su-fficient heating time 190 minutes~, 1.actose and glucose can cause ammonia re]ease suppression equal to xylose.
When heated for 30 minutes, ammonia release from samp].es treated with xylose was only 20 percent of that from untoasted soybean mea]. heated without sugar as shown in FIG. 1. These data suggest sugar addition augments the effect of pH, moisture level ~ 3 ~

and heatin~ on nonenzymatic browning as measured by ammonia release.
As shown in FIG. 2, interactions were found between reducing sugar sources and levels when pooled across heating times. Iinear and quadratic contrasts of reducing sugar levels revealed no interactions between xylose, fructose and glucose.
Increasing levels of xylose, fructose and glucose from 1 to 5 mol/mol lysine resu]ted in similar rates of ammonia release suppression. However, lactose did not act similar]y and ammonia release at a]]
levels of lactose was the same.
A possible explanation for lack of response to increasing levels of lactose may be due to steric hindrance caused by the molecular size of this disaccharide. Lactose may readily react with exposed ]ysyl residues at ]ow concentrations but, because of its size, be unable to penetrate the tertiary structure of soybean meal protein and interact with lysyl residues on the interior of the molecule.
As shown in FIG. 3, interactions between samples heated 30 or 90 minutes with different levels of reducing sugar were not significant. An interaction did exist however, when samples heated ~31~7~

30 or 90 minutes were compared to unheated samples for the linear effect of sugar levels. Since temp-erature and duration of heating are considered the primary factors in~luencing rate of nonenzymatic browning, an interaction between level of reducing sugar and heating time might be expected.
Since brown;ng reactions wi]] occur at ambient temperatures on the primary reaction between casein and glucose, an interation between ]evel of re-ducing sugar and heating time might be expected.
Browning reactions will occur at temperatures slight]y above O degrees Celsius, but may require weeks to progress to a measurab]e extent. In the present studies, samples were heated within 2~ hours from the time sugar, p~ and moisture adjustments were made, and were stored at 4 degrees ~elsius in the interim. When heat was applied, however, there was a linear decrease in ammonia re]ease as sugar concentration increased from 1 to 5 mol/mol lysine.
2Q As shown in FIG. ~, an interaction was noted when commerical soybean meal treated with xylose, fructose, glucose or lactose by the linear efEect of heating time. Inclusion of reducing sugars in reaction media caused ammonia release suppression greater than could be accounted for by ef~ects of 7 ~ ~
~(~

p~, moisture adjustment and heating time. ~owever,interactions were also found among reducing sugars and the linear effect of heating time, which suggests rate of reactivity was different for various reducing sugar sources.
Ammonia re]ease from commerical soybean meal treated with xylose was lower at all heating times than when commercial soybean meal was treated with fructose, lactose or g]ucose. These data are in agreement with those of example 1 where xy~ose was the most reactive reducing sugar. An interaction was noted when fructose was compared to glucose by the linear effect of heating time. Fructose appeared to react similarly to glucose after heating for 30 minutes, while at 60 minutes glucose produced greater ammonia release suppression than fructose.
Data from examples 1 and 2 indicated that reducing sugars reacted with soybean meal when heated and caused ammonia release suppression 2~ greater than could be accounted for by the effect of heating soybean meal without sugars. ~hese data also demonstrated xylose to be the most reactive reducing sugar.
As shown by FIG. 5, an interaction was found between soybean mea] sources and the linear effect ~3~ ~7~
~1 of heating. Without heat applicat;on, ammonia release from untoasted soybean meal was higher than from commercial soybean mea3. The interaction between commercial soybean meal and untoasted soy-bean meal across heating times might be expected since heating proteins reduced susceptibility to degradation by ruminal microbes.
The different ammonia release values for com-mercial soybean meal and untoaste~ soybean mea] when samples were not heated (0 minutes) may be the result oE heating which occurred during commercial processing of soybean meal without sugar. However, similar ammonia release values were observed for both soybean mea3 sources for 60 minutes. These data indicate nonenzymatic browning produces similar ammonia release suppression from either untoasted soybean meal or commercial soybean meal, though at different rates.
No interactions were notefl between p~ and heating times as shown in FIG. 6. Addi~ion o NaOM
to change pM to 8.5 or 10.0 resulted in lower ammonia release than for samples heated at natural pE~ (6.5). Samples treated to p~ 10.0 showed lower ammonia release than those of p~ 8.5. ~he effect o ~1 ~3~7~
~2 heating time averaged across pH treatments reduced ammonia release in a negative quaclratic manner.
Amounts of NaO}J required to change pH to 8.~
and 10.0 were 2.01 x ]0-4 and 3.58 x 10-~ moles/g soybean meaJ, respective]y. Random testing of the supernatant from tubes containing samples treated to pH 8.5 or 10.0 fo]lowing the 24-ho~r incubations revealed values not different from tubes where soy-bean meal was not treated with NaOII.
The epsilon amino group of ]ysine is primarily afEected between pH 8 and 9 because a proton is removed, making it a stronger nucleophile than a protonated primary amine. ~pp]ication of NaOH in-duces reactions other than nonenzymatic browning if pH is allowed to rise above 10. Under these conditions, amino acids racemize and crosslinks, primarily in the form of lysinoalanine, form.
As shown by FIG. 7, an interaction was found between percent dry matter of samp]es and whether or 2Q not pans were sealed during heating when tested across the comp]ete range of dry matter levels.
When evaluated between 60 and 80 percent dry matter, however, in~eractions were not detected. The interaction appeared to manifest itself when samples contained greater than 80 percent dry matter.

13~l~7~

Samples heated in covered pans reacted more comp].etely at low moisture level.s than those in uncovered pans. Evaporative losses from uncovered pans during heating ].ikely caused mois~ure to be more limiting than in covered pans, especial.].y at high dry matter content.
Moisture is necessary for nonenzymatic browniny reactions to occur since water serves as the medium through which reactants interact. However, exces-sive moisture content in reaction mixtures can s].ow the rate of nonenzymatic browning through simp]e dilution of reactants and, because a molecule of water is produced for each amino sugar formed, through end product inhibition. Water activity is the preferred method of expressing availability of water to participate in reactions. Water content is less descriptive than water activity since proteins, as well as other mol.ecules, are able to tight].y bina water, thereby making it unavail.able to serve other 2Q purposes.
In conclusion, nonenzymatic browning reduced in_itro ammonia release from soybean meal treated under a variety of conditions. Results suggest this chemical reaction may be useful for increasing the ~Iq ~ 7~

amount of soybean meal which escapes ruminal degra-dation.

5. In Vivo General Conditions Com~,ercial soybean meal was adjusked to p~ 8.5 with sodium hydroxide, and xylose added to supply 3 mol/mol of lysine. On a dry matter basis, the mixture contained 91 percent soybean meal, 8.5 per-cent xylose and .5 percent NaOE~ Water was added to this mixture to adjust the dry matter content to 83 percent. Heat app]ication was achieved by weighing 820 g soybean meal dry matter into 28 cm by 40 cm by 6 cm aluminum pans, sealing the pans with aluminum foil and heating in a forced air oven at 150 degrees Celsius. After 30 minutes, pans were removed from the oven and the soybean meal spread in a thin layer on a plastic sheet and allowed to air dry for 24 hours. The final product was compared to commercial soybean meal and urea as a source of supplemental 2Q protein in two examples.

6. In Vivo Examples The effect of nonenzymatic browning on amount of dietary soybean meal protein escaping ruminal ~ 3 ~
~s fermentation was determined using six growing, duodenal].y cannulated Angus x ~lereford steers (2~7 kg) in a simultaneous.l,y replicated 3 x 3 Latin square design. Cannul.ae were placed approximately 10 cm from the py].orus. ~he three treatments in~
vestigated were urea~ commercial soybean mea] and the prepared feed. Diets (table 1) were formu].ated to contain 12.5 percent crude protein equivalence and 54 percent TDN (total. diges~.ab],e nutr;ents), with supplements providing 67 percent of the dietary N.
To ensure all diets supplied adequate ruminal ammonia, urea was included as 58 percent of the supplemental N (nitrogen) to diets containing commercial soybean meal and prepared feed. A].fa],fa hay ~15.9 percent crude protein equiva.lence, dry matter basis~ was inc].uded to provide rumina].
degradable protein. Dextrose was added to diets containing urea or commercial soybean meal at 0.64 2Q percent of diet dry matter to equa], the ].evel of xylose supplied by the prepared feed.
The diets are shown in table 1. In this table and in tables 2-12, S.~. is the standard error of the mean, free amino groups is al.pha amino nitrogen;

46 ~3~7~

TAlBL~ 1.
CO~POSITICiN OF DIRTS FE:D TO DI~ODE:NALLY CAN~T~:D ST~ER5 .
Tr at~nt Ingredient U CS ~-30 -% of dry ~atter~
Ensiled ground corncobs 70 . 40 70 . 40 70 . 40 Ground alfalfa hay 17 . 60 17 . 60 17 . 60 CS ---- ~,9 ----XTS-30 -- -- 7 . 52 Urea ~.47 1.50 1.50 Ground Corn 7 . 65 1. 83 1. 91 Dicalcium phosphate . 91. 72 . 74 Dextrose . 64. 64 --Salt .30 ~30 .30 ~race ~in~ral premix . 02. 02 . 02 Vitamin premix . 01. 01 . 01 _ _ X

13~l~7~
~7 V-A is venus minus arteria],; SsM is soybean mean;
GTS is glucose-trea~ed soybean meal; cGM/s~ is corn gluten meal-blood mea]; U is urea; CS is control soybean mea]; XTS-30 is xy],ose-treated soybean meal heated 30 minutes (prepared feed); XTS-55 is xylose-treated soybean meal heated 55 minutes.
The trace mineral premix contains 20 percent Mg, 12 percent Zn, 7 percent Fe, 4 percent Mn, ]
percent Cu, .3 percent I and .1 percent Co and the vitamin premix contains 30,000 IU vitamin ~, 6000 IU
vitamin D and 7.5 IU vitamin E/g.
Animals were individually penned in an environ-mentally contro],led room supplying constant light and temperature (23 degrees Celsius~. Dry matter intake was restricted to 2 percent of body weight and anima]s were fed every 2 hours to approximate steady-state ruminal conditions. Experimental periods were 14 days in length and consisted of 10 days prefeeding and 4 days collection. Duodenal and fecal samples were collectea every 8 hours, with a 10-hour interval between days to allow a shift in sampling times. This sequence of sampling allowed a sample to be obtained at every even hour of the 24-hour day. uodena] (130 m]) samples were obtained by removal of the cannulae plug and waiting for ~7 ~8 ~ 7-~

surges of digesta that were col]ected in whirl-pack bags. Feca] grab samples were obtained at the time of duodenal sampling. Ensiled corncob, alfalfa hay and supplement samples were collected once daily during collection periods. Duodenal, fecal and feed samples were stored frozen.
Duodenal samples were composited on an equal volume basis within animals and period and sub-sampled. Fecal samples were similarly composited on an equal as-is weight basis. Composites were lyophilized and ground to pass through a 1 mm screen. Ensiled corncob samples were prepared for grinding by air drying and all feed samples were ground to pass through a 1 mm screen before being composited by period.
Laboratory analyses included indigestible acid detergent fiber, which served as the solids flow marker, N, ash and diaminopimelic acid. Because of difficulties determining bacterial N:diaminopimelic ~0 acid ratios, bacterial protein synthesis was cal-culated assuming 18 g bacterial N/g diaminopimelic acid. Each animal served as its own contro] to estimate the fraction o~ commercial soybean meal or prepared eed protein escaping ruminal degradation by equation 1 where percent ~EP is the rumina]

11 3~7~
~9 escape estimate of soybean meal protein, TNFS is total duodenal nonammonia N flow when consuming soybean meal or prepared feed g/d (grams per day~, BNFS is duodenal bacterial flow ~hen consuming soy-bean meal or prepared feed (g/d~, TNFV is total NAN

(nonamonia nitrogen~ flow when consuming urea (g/d~, ~NFU is bacterial N flow when consuming urea, and SNI is soybean meal N (nitrogen~ intake (g/d~.

EXAMPL~ 7 Three six-month old Finnsheep x Suffo]k ram lambs (24.7 kg~ were emp]oyed in a 3 x 3 Latin square design to measure net FAN absorption from the portal drained vîscera when urea, commercial soybean meal or prepared feed were supplemental N sources.
Diets (table 2~ contained 12 percent crude protein equivalence (dry matter basis~ and 57 percent TDN
with 65 percent of the dietary N supplied by supple-ment.

For diets containing commercia] soybean meal, 100 percent oE the supplemental N was supplied ascommercial soybean meal, while for diets containing ~9 ~3~75~
so EQU~TION 1 REP = _ (TNFS - BNFS~ - (TNFU - BNFU) x 100 100 - ((ND - NDU)/((PNS/100)*(PN~/].00))) S~ 131~75A
TABLE: 2.
Ct~ OSITION OF DI~TS F~D TO CZ~q~HE~ERIZ:13D L~BS

TreatJnent Ingredient U CS XTS-30 96 o~ dry matter~
Ensiled 64.10 64.10 64.10 Ground alfalfa hay 12 . 0012 . 00 12 . 00 Cane molasses 5.00 5.00 5.00 CS -- 16.~9 --XTS-30 -- -- 9.59 Ground Corn 13.35 .39 6.51 Urea 2.25 - 1.05 Dextrose .~1 .81 --Dicalcium phosphate1,18 .73 .94 Potassium chloride .62 .01 .29 Ammonium sul~ate .27 -- .13 Salt .33 .33 33 Magnesium oxide .05 -- .02 Trace mineral premix .03 .03 .03 Vitamin premix .01 .Ql .01 y 52 ~ 31~ 7~

prepared ~eed, ~0 percent o~ the supplementa] N was supplied by prepared feed and 40 percent by urea.
Diet dry matter was fed at 2.5 percent of body weight in equal portions at 0600, 1200, 1800 and ~400 hours. Water was avai]ab]e afl ]ibitum. Prior to initiation of this trial, animals were fed pelleted a]fa]fa Eor five weeks.
Lambs were placed under general anethesia for surgical implantation of hepa~ic porta] vein, mesenteric vein and carotid arterial catheters.
Following surgery, catheters were flushed twice weekly with steriJe, physiological saline containing 100 units/ml heparin, 1 percent benzyl alcohol and .5 percent procaine penicillin G: dihydrostrepto-myocin. Experimental periods were 7 days in length during which animals were adapted to diets for 6 days. On day 7, blood samples taken before the 0600 feeding and then hourly until 1100 hours.
Blood flow rates were est;mated by primed, continuous infusion of 3 percent (w/v~ para-amino hippuric acid into the mesenteric vein. Samples of arterial and portal blood (20 ml) were simul-taneously drawn into heparinized syringes, placed into tubes containing 30 mg NaF ana mixed. Packed cell volume was determined immediately by centri-r~3 1 3 14 7 5 4 fugation of capillary tubes filled with b]ood. A 10 ml aliquot of whole b]ood was deproteinized for para-amino hippuric acid analysis. Plasma was deproteinized with sulfosalicylic acid for determination of FAN.
Samples of deproteinized venous and arterial whole blood were composited and analyze~ for para-amino hippuric acid. Deproteinized plasma sampJ,es were analysed for FAM. Blood flow rates were cal-culated by mù]tiplying the flow of blood by ~100-packed cell volume)/100 and dai]y net portal absorption of FAN was calculated.
Net portal FAN absorption due to cons~mption of commercia], soybean meal or prepared feed was cal-culated by subtracting FAN absorption when urea was the crude protein source from net portal absorption of FAN when commercial soybean meal or prepared feed were fed. Because commercia], soybean meal supplied 100 percent of the supplemental N and prepared feed supplied 60 percent of the supplemental N, estimates of net portal absorption of FAN above urea for commercia], soybean meal were multiplied by .6 to allow comparisons between commercial soybean meal and prepared feed.

` ~4 ~ 31475~

7. Results And Discussion As shown in tab~e 3, organic matter intake was not different among treatments, as prescribed by the e~perimental protocol, nor was daily duodenal organic matter flow of fecal organic matter excretion different among treatments. Therefore, apparent ruminal and total ~ract organic matter digestibilities were not affected by treatment and averaged 50.3 and 57.8 percent, respectively.
Though differences were small, dietary N intake and soybean meal N intake were higher for steers supp]emented with prepared feed than commercia]
soybean meal (table 4). Duodena] NAN flows were higher for steers supplemented with soybean mea]
than for those supplemented with urea and were higher for steers supplemented with prepared feed than commercial soybean meal. Rumina] N
digestibilities were higher in steers fed urea than those fed soybean meal and were higher when commercial soybean mea] was fed than when prepared 2Q feed was fed.
Bacterial N flow to the duodenum of each animal was calculated by multiplying the quantity of diaminopime]ic acid reaching the duodenum by 18 g 5~

131~ ~5~
TABL~ 3.
IN'rAKE, FLOW RAT~ ANIl APPARE:NT DIG113STIBILITY OF ORGANIC ~TT~
FOR STEERS

Treat~ent Ingredient ~ CS 8TS-30SE:

Intake, g/d 4663 4605 461730 ~ O
Flow, g/d To duodenum 2328 2281 228621~ 9 Fecal excretion 1959 1937 194719 0 2 Apparent digestibility, %
Ruminal 50 . 0 50 . 5 50 3 4 . 5 Total tract 57 . 8 57 . 9 57 . 7 . 4 . . .

~3~7~

T~BL~ 4.
I~TARE, F~OW, APP~RENT DIG~TIBILI~Y AND ~U~INAL E5C~P~ OF
NITROGEN IN) F~ ST~ÆRS

7reat~ent Ingr0dient U CS ~TS-30 S~

N intake, g/d 97.~ 97.1 100~6 .6 Soybean N intaXe, g/d -- 25.8 27.3 .5 Duodenal flow, g/d Nonammonia N 65.2 71.4 79.3 1.4 Bacterial N 28.1 31.9 31.4 .8 Dietary N 37.1 39.5 47.9 1.4 Fecal excration, g/d29.3 30.3 32.9 .8 Apparent digestibility, %
Ruminal 33.6 26.2 21.4 1.6 ~otal tract 69.9 68.6 67.3 .8 Ruminal escape of soybean N, % -- 13.1 33.7 7.0 X

57 ~ 31~

bacterial N/g diaminopimelic acid. Daily duodenal flow of bacteria]. N was higher when soybean meal was fed than when urea was fed, b~t was not different between commercial soybean mea~ and prepared feed. Dietary N f]ows (including protozoal and endogenous N) were higher for soybean meal fed anima].s than for urea fed animals and were higher for animals supplemented with prepare feed than commercial soybean meal. Estimated ruminal escape values for commercia] soybean meal and prepared feed were 13.1 and 33.7 percent, respectively, and were different.
Fecal N excretion was higher when animals were fed soybean meal than when fed urea, and higher when prepared feed was fed than when commercia]
soybean meal was fed. These differences appear to be a function of the higher N intake for catt].e supplemented with prepared feed since apparent total tract N digestibility comparisons were not 2Q different. That total tract N digestibi].ity in steers supplemented with prepared feed was not lower than in ~teers supplemented with commercial soybean meal was encouraging since nonenzymatic browning reactions reduce N digestibility. Because M
digestibility was not affected, the data suggests ~7 5R ~ 31~ 7~1 protein protection occurred as a result of reversible nonenzymatic browning.
As shown in table 5, dry matter intake and packed cell volume were not dif~erent among treatments. Portal blood flow, however, was higher in soybean meal suppl.emented .l.ambs than urea supplemented lams, and tended to be higher in lambs supplemented with prepared feed than those receiving commercial soybean meal. Portal blood flow estimates observed in this example are generally higher than values reported in the .l.iterature where primed-continuous infusion of para-amino hippuric acid has been the method of measurement. In the present studies, blood samples were obtained between the 0600 and 1200 hour feedings with the intent that portal blood flow during this interval would be representative of mean daily portal blood flow.
Differences due to supplemental N sources were not statistically significant for either venous-arterial differences in FAN concentrations nor net portal FAN absorption, though values were numerically higher for lambs supplemented with prepared feed than those supplemented with commercial soybean mea].. Cal.cul.ated at equal 5 9 1 3 1 4 ~ ~ 4 T~BL~ 5~
~OD~ WEIGH~, FE~D INTAX2 AND BLOOD ~EASUR~ENTS FOR L~MBS

~reatment Ingredient U CS X~S-30 ~2 Body weight, kg 24.2 25.1 24.8 .8 Dry matter intake, g/d631 638 63111.4 Packed cell volume, %20.4917.97 21.041.60 Portal blood flow, ml/min 357 18642196 125 Portal blood ~low liters*h-l/kg.75 11.8 .6 AAN concentration, V-A
di~ference, mmol/l.180 .195 .233.045 AAN absorption, mmol/d281 447 578 113 AAN absorption above urea, mmol/d Observed 0 166 297 113 At equal SBM intake O 100 297 90 . . . _ X

. 60 v~i ~3~7~

soybean meal N intake, daiJy absorption of FAN from prepared feed was approximately three times that of commercial soybean meal.
Since uncontro]led nonenzymatic browning can produce proteins of low digestibility, testing was necessary to determine the effect of nonenzymatic browning on ruminal escape of soybean meal and whether protein digestibility was compromised.
Examples 6 and 7 suggest general agreement on the effect of nonenzymatic browning on metabolism of soybean meal. Example 6 showed ruminal escape of prepared feed to be approximately 2.6 times that of commercial soybean meal and total tract N
digestibi]ties were similar. Data from example 7 suggested, when calculated at equa.l. soybean meal protein intake, net portal absorption of FAN from soybean meal was approximately 3 times higher for prepared feed than commercial soybean meal.

8. In Vitro Examples Objectives of example 8 were: (1) to determine protein efficiency of prepared feed relative to untreated, commercial soybean meal, and (2) to determine if xylose-trea~ed soybean mea~ heated 61 ~3~7~

longer than 30 minutes wou]d cause improved or reduced protein efficiency re]ative to prepared feed. The second xylose treated soybean meal, XTS-55, was prepared similarly to prepared feed except heating was for 55 minutes at 150 degrees Celsius.
Forty-eight 3-month old Finnsheep x Suffolk lambs (22 kg) were utilized in a randomized complete block designO Twelve animals from each of three blocks (ewes (22 kg), light wethers (20 kg), heavy wethers ~26 kg~) were randomly allotted to each of four supplemental N sources, which included urea, commercial soybean ~eal, prepared feed and XTS-S5.
Four levels of soybean protein were fed within each soybean meal source. Levels of commercial soybean meal were 100, 80, 60 and 40 percent of supplementa]
N as commercia] soybean meal, the ba]ance as urea.
Levels of prepared feed and XTS-55 were 60, 45, 30 and 15 percent of supplemental N from the respec~ive source, the balance as urea.

Supplements, which comprised ~8.9 percent of diet dry matter, supplied 65 percent of the dietary crude protein equivalence. ~he diets (table fi~ were balanced for 12.~ percent crude protein equiva]ents and 57 percent total digestible nutrients. Glucose `

62 13~75~

TA~L13 6 Co~POSITION OP DI~TS F~D ~D LAM~S
. . .
Treat~ent Ingredient U CSor XTS-55 ... , . . _ . _ .
~ of dry matter----------Ensiled ground corncobs 64.10 64.10 64.10 Ground alfal~a hay 12.00 12.0012.00 Cane molasses 5.00 5.005.00 CS -- 16.59 XTS-30 or XTS-55 ~ 9.61 Ground corn 13.35 .396.49 Urea 2.2~ -- 1.05 Glucose .81 .81 Dicalcium phosphate1.18 .73.94 Potassium chloride .62 .01.29 Ammonium sulfate .27 -- .13 Salt .33 33 33 Magnesium oxide .05 -- .02 Trace mineral premix.03 .03.03 Vitamin premix ~01 .01.01 . . ~

~r ~,. . .

63 ~ 7 ~ ~

was included in diets fed to lambs consu~ing urea and commercial soybean meal at .81 percent of cliet dry matter, equalling the quantity of xylose provided by prepared feed and X~S 55. Thougho~lt the 80 day trial, animals were indiviaua]ly fed once daily. Diets were rationed as a percent of body weight determined by the quantity of feed consumed by lambs fed urea. Water was avai]ab]e ad libitum.
Initial and final weights of lambs were determined as means of three consecutive day weights. Animals were housed in a room supplying continuous light and constant temperature (23 degrees Celsius). Feed refusa]s were measured weekly and sampled for dry matter ana]ysis. Dry matter contents of Eeeds and feed refusals were determined by drying samples in a forced air oven at 60 degrees Celsius for 72 hours.
Protein efficiencies of soybean meal sources were determined. Dry matter and soybean meal protein intakes, and gain and feed efficiency data were analyzed for main effects of N source.

Apparent digestibilities of protein supplied by urea, commercial soybean meal and prepared feed and h4 13147~4 XTS-55 were measured. ~wenty-four Finnsheep x Suffolk wether ~ambs (27 kg) were fitted with canvas fecal collect;on bags and assigned to one of four dietary treatments in a completely randomized design. Diets (table 6~ were individually fed once dai ly a'c 2.t5 percent of body weight in a room supplying continuous light and constant temperature (23 degrees Ce] sius~.
The experiment consisted of 10-day adaption followed by a 7-day fecal col]ection. During the collection period, feces were weighed daily and a 10 percent al iquot frozen. Feeds were sampled daily during collection. Composites were subsampled for dry matter determination and dried in a forced air oven at 60 degrees Celsius for 72 hours. l`he remainder of composites were lyophi 1 ized and ground to pass through a 1 mm screen. Samples analyzed for N by the macro-K jeldahl producersO
Digestibility of N of soybean meal origin was estimated by equation 2 where ND is apparent N
digestibility by lambs consuming commercial soybean meal or prepared feed, NDU is mean apparent N
digestibility by lambs consuming urea, PNS is percent of supplemental N supplied by commercial soyhean meal (100 percent), prepared feed (60 ~ 7~

percent) or XTS-55 (hO percent~, and PND is percent of dietary N supp]ied by supplement (65 percent~.
Values obtained were estimates relative to urea, which was assumed to be 100 percent digestible.
nata were analyzed as a completely random design by analysis of variance.

Example 10 was conducted to determine if protein efficiency of soybean meal could be improved by treating with a less costly reducing sugar, glucose. Using an in vitro protease (ficin~
assay, soybean meal treated with 1, 3 or 5 mol glucose/mol lysine and heated for 30, fiO or 90 minutes at 150 flegrees Ce]sius was compared to prepared feed. Dry matter content (percent) and p~l of all samp]es prior to heating were 80 and 8.5, respectively.
Data (table 7~ showed flegradability of soybean meal treated with 2 or 3 mol glucose/mol lysine and heated 60 minutes was simîlar to that of prepared feed. Protease degradability data were taken to suggest that glucose-treated soybean mea~ would have a similar nutritive value as prepared feed.

6~

66 ~ 3 ~ 4 EFFECT OF GLUCOS}~ Ll~LS i~ 113ATING q~ A'lr 150 C.
oPa FICIN l)EGRAD~BILI'rY OF SOYBEA2t P~aOTE:IN

P~educing ~3ugar Contrs~l Xylose Gluco~e Le~T~l (mol/Dol lysine) ~ 3 1 2 3 Undigested N, % o~ original ~- -Min at 150 C
31 . 0 61 . ~ 36 . 8 40 . 2 38 . 1 6~ 44 . 1 55 . 5 ~4 . 2 66 . 9 53 . 8 76 . ~ 76 . 4 78 . 0 :

~3~7~

Glucose-treated soybean ~eal was prepared by adding 3 mol glucose/mol lysine, adjusting DM content and pH to 80 percent and 805, respective]y, and heating for 60 minutes according to procedures previously described.
Sixty mixed breed steers (~18 kg) were fed 105 days ~o measure protein efficiency of glucose-treated soybean meal relative to commercial soybean meal. The experimental design was a randomized complete block in which cattle were randomly assigned to one of two open front confinement barns.
Supplemental N sources were urea, commercial soybean meal, glucose-treated soybean meal and a 50:50 (protein basis) mixture of corn gluten meal and blood meal which served as the positive control.
Twelve animals were randomly assigned to receive urea, and sixteen animals randomly assigned to receive commercial soybean meal, glucose-treated soybean meal or corn gluten meal and blood meal.

Levels of commercial soybean mea] were 100, 80, 60 or 40 percent of the supplemental N, the balance as urea. Levels of glucose-treated soybean meal and corn gluten meal and b]ood meal were 60, 45, 30 or 15 percent of supplemental N, the balance as urea.

6 ~ ~ 3 1 4 7 5 ~

Cattle were individually fed through Calan-Broadbent electronic gates.
Diets (tab]e 8) contained 11 5 percent crude protein equivalents and 55 percent total digestible nutrients. Supplements, which comprised 15.85 percent of diet dry matter, supplied 57 percent of dietary No Glucose was included in diets containing urea, commercial soybean meal and corn gluten meal and blood meal at .81 percent of diet dry matter, equalling the level supplied by glucose-treated soybean meal.
Feed was rationed once daily as a percent of body weight determined by the quantity of feed consumed by steers fed urea. Water was available ad libitum. Samples of feeds were obtained weekly and dry matter was determined by drying samples at 60 degrees Celsius for 7~ hours. Supplement samples were anlayzed for N by the macro-Kjeldah] technique to ensure proper N content. Initial and final 2Q weights of steers were determined as means of three consecutive day weights.
Protein efficiencies were determined as previously described. Daily dry matter and protein intakes, and gain and feed efficiency data were ~5g f' 131475~L

CONPOSITION CIF DI13q~ FE:D rTO S~ RS IN q~RIAL 3 AND
LANB~;

~eatmell~
Ingr~lienlt: 1~ CS GTS CG~J~I

--% o~ dry matter~

Ensiled groundcorncobs 66 .15 66 .1566 . î5 66 .15 Ground alfalfa hay 18 . 00 18 0 00 18 . 00 18 . 00 CS ---- 13~84 --------GTS --- --- 8.20----Corn gluten meal -- -~ -- 2.46 Blood meal -- -- -- 1. 81 ~:;round corn 10. 97 . 05 5 ~157 . 81 Urea 1.~2 ~ .84.81 Glucose . 80 . 80 -- . 80 Dicalcium pho~phate1.10 . 72 . 901. 05 Potassium chloride. 49 -- . 23. 49 Salt . 31 . 31 . 31. 31 Ammonium sulfate . 29 -- .17. 23 Magnesium oxide . 04 -- . 02. 04 Sulfur -- . 02 -- --Limestone - - . 08 - --Trace mineral preTQix. 02 . 02 . 02. 02 Vitamin premix . 01 . 01 . 01. 01 X' :

analyzed for main effects of protein source by analysis of variance of a randomized complete block design.

Apparent digestibi~ity of protein supplied by urea, commercial soybean meal and prepared feed was determined. Eighteen Finnsheep x Suffolk wether lambs (40 kg) were fitted with canvas fecal collection bags and assigned to three dietary treatments (urea, commercia] soybean meal and glucose-treated soybean meal; table 8) in a completely randomized design. Lambs were individually fed at an equal percent of body weight in ~etabolism crates under continuous light and constant temperature (23 degrees Celsius). Protocol and response variables for this experiment were as described in example 7.

9. Results and Discussions Protein efficiency is defined as daily gain observed above that of animals fed urea per unit of true protein supplemented. Protein efficiencies of commercial soybean meal, prepared feed and XTS-55 7~ 7 ~i 4 fed to sheep in example 7 are presented as s~opes in FIG. 8~
In FIG. 8, there is shown protein efficiency by lambs consuming control soybean meal (commercial soybean meal), xy~ose-treate~ soybean meal heated 30 minutes (prepared ~eed) and xylose-treated soybean heated 55 minutes (XTS-55), in example 8. Slopes and standard errors for commercial soybean meal (curve 94), prepared feed and XTS-55 were, respectively, .63, .16; 1.27, .31; .91, .28.
Comparisons were commercial soybean meal vs.
prepared feed (curve 90) and prepared feed vs. X~S-55 (curve 92). Protein efficiency of prepared feed was approximately two times higher than that of commercial soybean meal. Protein efficency of X~S-55 was intermediate to prepared feed and commercial soybean meal and not statistically different than prepared feed.
As intended, dry matter intakes by lambs in example 7 were not different among treatments (table 9). However, gains and feed conversions (gain/dry matter intake) were higher for lambs fed soybean meal than urea. ~o differences were observed among commercial soybean meal, prepared feed and XTS-55 72 ~3147~4 ~rABLE 9 Eæ AND P13RFORP~ANCE~ DATA OF LA~S IN lrRIAL 1 . . .
q~realt~ent Ite~ U CS ~S--30 ~rs--55 51 -Intake o~:
~ry matter, g/d 610 610 635 631 17 Dry matter, % of body weight 2.58 2.55 2.55 2.60 .03 Protein above urea-~ed lambs, g/d 35.4 197~ 18.9 3.0 Gain, g/d 35.8 52.7 56.6 46.3 7,4 Gain/dry matter intake .057 .083 .oso 0.73 .0~1 -73 ~3 for ~ain or feed conversion. ~owever, gain and fee~
conversion, when measured be]ow an anima]'s protein requirement, would be expected to reflect both the quantity and rumlnal degradability of protein fed.
nalf as much protein from prepared feed was required to achieve gains and feed conversions equal to lambs fed commercial soybean meal.
~ ry matter intakes by ~ambs in example % were not different among treatments (table 5). Apparent dry matter digestibilities were lower by lamhs consuming prepared feed than those fed XTS-55, but no explanation for this occurrence can be given.
Apparent digestibilities of N were lower for soybean meal-supplemented lambs than urea-supplemented lambs and were lower for lambs supplemented with prepared feed and XTS-55 than those fed commercial soybean meal. Apparent N
digestihility was not different among prepare feed and XTS-55. Since protein efficiency of X~S-55 was 2Q numerica]ly, but not statistically, lower than prepared feed in e~amp~e 7 and since digestibility of N fron; prepared feed was not different from that of XTS-55, heating xylose-treated soybean meal longer than 30 minutes may be unnecessary to achieve treatment.

~ 3~4~
7~

Presumably treating soybean meal by con~rolled nonenzymatic browning reduced ruminal pro~eolysis of prepared feed, thereby reducing urinary N excretion and increasing postruminal metabolizable protein flow per unit of protein consumed compared to commercial soybean meal.
In FIG. 9, there is shown protein efficiency by steers consuming commercial soybean meal, glucose~
treated soybean meal and corn gluten meal/blood meal. Slopes and standard errors for commercial soybean meal, glucose-treated soybean meal and corn gluten meal/blood meal were, respectively, .90, .10;
1.91, .21 1.85, .21. Comparisons were made between commercial soybean meal vs. g3ucose-treated soybean meal and glucose-treated soybean meal vs. corn gluten meal/blood meal.
Protein efficiency was more than two times higher for steers supplemented with glucose-treated soybean meal (curve 100) than for those supplemented with commercial soybean meal (curve 104), but was not differen~ than that from steers fed corn gluten meal/blood meal (curve 102~. The corn gluten meal and blood meal mixture was selected as the positive control because the individual proteins are high ruminal escape proteins. Protein efficiency of corn 7~

75 ~3~7~

gluten meal/blood meal relative to commercial soybean meal in the present study was within the range of values previously reported.
Dry matter intakes by steers in example 9 were not different among treatments as shown by tab]e 9.
Averaged across all levels of supplemental Nl intake of protein from commercial soybean meal was approximately two times higher than that from glucose-treated soybean meal and corn gluten meal/blood meal while anima] dai]y gains and feed conversions (gain/dry matter intake~ were similar.
Metabolizable protein was first limiting in the basal diet since steers consuming urea had lower gains and feed conversions than those consuming commercial soybean meal, glucose-treated soybean meal or corn gluten meal/blood meal. The weight gain improvement using treated feed is shown in table 10.
Dry matter intakes by lambs in example ~ were not different among treatments (tables 11 and 12) since anima]s were limit fed. ~owever, apparent dry matter digestiblities were higher for lambs supplemented with soybean meal than those supplemented with urea. It may be that alfalfa did i ~31~7S~
TABL~ 1 û
INTAR~ AND PE:RFOR~NC~ DATA OF STE~S

q reat~ellt Ite~ U cS GTSCG~/~ S~
-In~ake of:
Dry matter, kg/d4 . 96 5 . 0~ 5 .17 5 . 02 .10 Dry matter, % of ~ody weight 2.102.10 2.102.11 .01 Protein above urea-f ed steers, g/d - 204 11095 15 Gain, kg/d . 27. 46 ~ 47 . 43 . 04 Gain~dry matter inkake . 053 . 090 . 091 . 086 .007 77 131~7~4 T~BL~ 11 INT~Kæ AND DIG~S~ BILITY OF DRY NATTE~ ~ND NITROG~ (N3, AN~ R~COVERY IN FLC~S OF ACID DE~RGE~T INSOLUBL~ N ~ND PEPSI~
I~OL~BLR N FRO~ LA~BS

~rea~e~
Ite~ U CS XTS-30 ~TS 55 S~

Dry matter intake, qjd 703 716 704 729 44 Digestibility of (~):
Dry matter 59.161.0 58.761.3 1,0 Nitrogen 69.767.6 62.563.8 lol Soyb4an N ~ 96.5 77.481.5 2.5 Recovery of (%):
Acid detergent insoluble N 60.060.4 36.837.0 1.5 Pepsin insoluble N 151.8 118.4126.3 114.0 3.6 __ 78 131~7~

~A~I~ 1 2 Ipa~rAK~ AND DIOEsTIBILIq~r OF DRY NAq~R AND NITROGE~N l N
AND RECOV~RY Ill FECi :S OF ACID DE~ RGE~r INSOLUBL~3 N AN~ P13PSI~
~ 5t)LUBLB N FR~ I~S
.. . . . . . .
Tre~ent -Ite~ U CSGT5 5 Dry matter intake: 986 1009 940 33 g/d ~ of body weight ~.45 2.472.45 .01 Digestibility of (%):
Dry matter 60.~ 61.762.g .7 Nitrogen 70.0 69.968,0 l.o Soybean N ~ 99.993.4 2.15 Recovery of (%)~
Acid detergent insoluble N 65.0 62.B53.5 1.6 Pepsin insoluble N 114.6107.7112.3 3.4 .

X

13~7~
7~

not supply adequate quantities o~ ruminal degradabIe protein to support optimum microbial growth in lambs supplemented with urea.
Apparent dietary N digestibilities were not different among treatments. ~owever, calculated digestibility of N from glucose-treated soybean meal was 6~5 percent lower ~han that from commercia]
soybean mea~. Thus, a 100 percent improvement in protein efficiency was noted in example 9 as a result of treating soybean meal by nonenzymatic browning even though treatment depressed N
digestibility of glucose-treated soybean meal in example 10. These results are in general agreement with results of examples 7 and 8, although digestibility of protein from prepared feed was estimated as somewhat lower than glucose-treated soybean meal.

10. In Vitro Examples Commercial solvent extract, dehulled soybean meal (47.5 percent protein~ is dry blended with spray dried spent sulfite liquor containing 19.5 percent reducing sugars. The spent sulfite liquor is added at a rate of 5 or 10 percent on soybean meal (as is basis), depending on the specified trea~ment level. In some treatments, hydrated lime was added at a rate of 6 percent by weight on spent sulfite liquor.
The mixture is metered, at a rate of 1 kg/minute, into a cylindrical mixing chamber 18 inches in length and 8 inches in diameter where it is heated by direct application of low pressure steam (24 psi). Water is pumped into the chamber at a rate of 4 percent on the mix. Starting temperature of the mixture is 20 to 21 degrees Celsius. In less than 15 seconds, the temperature is increased to 90 to 95 degrees Celslus.
The hot feed exits the conditioning chamber into the top of a vertical holding bin where it slowly descends to the outlet emerging 90 or 120 minutes later. The reaction is exothermic and will increase in temperature from 5 to 10 degrees Fahrenheit, depending on the formulation, while in 2~ the bin.
Feed is removed from the bottom of the bin by a metering screw. The hot feed is held on wire screen as ambient air is forced upward through it. This cools and drys the feed~

8~

11. Results and Discussion The results (table 13~ showed only minor changes with changes in pH and temperature and a greater effect o the level of sulfite liquor used.
This example indicates that much less reducing sugar may be usable under controlled condi~ions. It is possible that the amount of reducing sugar may be as ]ow as 1/3 mole of reducing sugar to one mole of epsilon amino groups or ]ower and as litt]e as 0.5 percent xy]ose to the protein by weightO
Presumably, since this is below the theoretical amount, there is an inhibiting effect that reduces the epsilon amino groups subject to microbial action without reaction of all of them with carbonyl groups of reducing sugars.

Four commercial lignosulfonates were added to solvent extract soybean meal at a rate of 5 percent by weight on the soybean meal, the mixtures were pelleted under identical conditions, and the resultant pellets tested for degradability of soybean protein by rumen microorganisms maintained in batch culture.

: .... . .

~ 3 ~
~2 q~ABLE: 1 3 Ten~p. ~mp.
Ran~e Range %CIS LI~ H[OLDOI~IG FIN~L 70 C-75 C 90 C--95 (:
ON %ON TI~PROCE:SS PR~13SS A~ONIA A~ONIA
SE~ IN 160--170 195--205 ~G/100NL 2qGJ1001~L

0 0 0 39 40 . 8 39 . 9 6 120 25 11 17 . ~ 12 . 8 6 so 21 . 8 14 . 7 0 120 18 . 9 11 . 1 6 120 35 19 29 9 8 19 . 9 ,~.

~ ~3~7~
~3 The four commercial lignosufonates were so].d under the brand names Toranil la trademark of Rhinelander Paper Company3, AmeriBond, Maraton and Maraton SNV, all three of the latter being trademarks of Reed Lignin Corporation. The first two o~ the .l.ignosulfonates contained less than 2 percent and 1 percent by weight, respectively, of reducing sugars and the last two contained 16 percent and 13 percent by weight, respectively, of reducing sugars.
The two samples with less than 5 percent reducing sugars showed no reduction in protein degradability indicated by curves 20 and 22 in FIG.
10 and the third and fourth columns of table 14.
The two samples with more than lS percent reducing sugars showed significantly depressed in protein degradability, as indicated by curves 24 and 26 in FIG. 10 and the last two columns of table 1~.
This comparison indicates that simple pelleting of a soybean meal-lignosulfonate mixture does not guarantee reduced proteing degradability; additional factors are involved and must be controlled.

8~ ~ 3~47~

TABL~ 14 Net in vitro ammonia production by rumen bacteria, mg/100 ml Hours SBM Torani]. AmeriBond Maratan Maratan SNV
_ 10.9 0.8 0.7 0.4 0.4 22.2 2.2 1.9 0.9 ~.6 44.7 4.2 3.8 1.1 O.q 67.4 6.8 5.9 1.3 0.9 812.0 8.2 7~8 1.8 1.1 1013.5 11.8 11.~ 2.4 1.8 2419.0 ~2.0 21.5 17.7 17.3 .. .. . . . . _ _ _ ' ~5 ~ ~3~7~

EXAMPL~ 14 l~ltrafiltra~ion was used to concentrate calcium lignosulfonate molecules (CaLSO3~ occuring in spent sulfite liquor. The permeate fraction retained low molecular weight calcium ]ignosulfonates, oligosacchrides and wood sugars ~primarily xylose~.
The original spent sulfite liquor and its concentrate and permeate fractions were spray dried to approximately ~5 percent solids. Analyses for the resulting powders are listed in table 15~
Solvent extract soybean meal was combined with 1, 2, 4 and 8 percent sulfite liquor, 4 percent concentrate, or 4 percent permeate. Addition rates are expressed as percent by weight of additive on soybean meal, as is basis (about 10 percent moisture~. The various mixtures were conditioned to 85 degrees Celsius with direct applicat;on of steam, pelleted, and returned to room temperature by evaporative cooling under a forced air stream.
Total process time above room temperature was less than 5 minutes.
~Degradability of protein by ruminal microbes was determined in batch culture for each sample.
Results are plotted in FIG. 11. Protection 86 ~3~7~

SSL COMC~ PERM.

Ca, % 3.94 3.434.18 Na, ~ 0O03 0.020.04 Total S, % 5.79 5.535.93 CaLS03, % 56.37 ~0.4246.35 Reducing Sugars, %17.135.32 22.80 . . _ . . .

~, .

~ ~3~47~
~7 increased directly with addition of spent sulfite liquor. Permeate was approximately 30 percent more effective than sulfite liquor, corresponding closely to the 33 percent increase in reducing sugars in the permeate fraction. The concentrated CaLS03 fraction did not provide protection against degradability, indicating that calcium ]ignosulfonate per se is not an effective agent for treatment of soybean meal~
As shown in FIG. 11, data point 30 represents spent sulfite Jiquor containing 17 percent reducing sugar and data point 32 represents permeate containing 22 percent reducing sugar.

EXAMPLE ]5 Permeate produced by ultrafiltration of spent sulfite liquor was washed with an alcohol-amine mixture to extract any remaining ca]cium lignosulfonate molecules. The aqueous phase containing sulfite liquor reducing sugars was concentrated and applied to solvent extract soybean mea] as a protein protection agent, as were the original sulfite ]iquor, its permeate and technica]
grade xylose. Each was dissolved in water and applied such that the solution provided 5 percent added moisture on soybean meal. Samp~es were mixed fl~ ~ ~3~

in a V-b]ender equipped with a high speed agitator and stored in plastic bags.
Blended samp]es were conditioned to 90 degrees Celsius by direct application of steam, pelleted, and returned to room temperature by evaporative cooling under a stream of forced air. Prior to pe]leting, one sample was observed to have caked and darkened slightly during s~orage. A portion of this unpe~leted meal was retained for testing. Protein degradation by ruminal microbes was determined by 6-hour batch culture fermentation~
Concentration of the spent sulfite liquor sugars, the major portion of which is known to be xylose, through ultrafiltrat;on and extraction increased the effectiveness of the protein protection agents. Technical grade xylose was also effective, indicating that re~ucing sugars alone are effective treatment agents.
It was also learned that, under some 2Q conditions, reaction can occur at room temperature.
In this example, a sulfite liquor-xylose-soybean meal blend reacted after storage at room temperature for 2 hours, reducing degradability to 82 percent versus untreated soybean meal. Pelleting this same ~ 3~7~
~9 mixture at 90 degree Celsius further reduced degradability to 4~ percent of untreated soybean meal. While it is recognized that some reaction can occur at room tempera~ure, the preferred method includes application of heat to the soybean meal-sugar mixture. ~hese results are shown in tab1e 16.

EXAMPI.E 1 6 Solvent extract soybean meal obtained from four commercial sources was mixed with a permeate (4 percent solids on soybean meal) resulting from ultrafiltration of spent sulfite liquor. Permeate supplied about 0.9 percent reducing sugars on the soybean meal. Mixtures were conditioned to 85 degrees Celsius by direct application of steam, pelleted, and the hot pellets returned to room temperature by evaporative cooling under a forced air stream.
Resultant pellets were tested in ~hour batch 2Q culture for degradability of protein by ruminal microbes. Results, listed in table 17, indicate that the process for protecting soybean meal protein is of general application, not specific to a single source of meal.

Effect of Pelleting Soybean Meal Containing Protein Protection Agents on Release of Ammonia by Ruminal Microbes Added Reducing NH3-N ~ as Sugars, % g/100 ml SBM
Control, SBM 0.0 23.5 100.0 SSL, 3% 0.6 20.4 86.6 Permeate, 3~ 0.7 18O6 7~.9 Permeate Sugars, 3% 2.7 9.3 39.7 Xylose, 1~ 1.0 15.7 66.8 SSL, 3% and xylose, 1~ 1.6 9.9 42.0 (6) Unpelleted ~.6 19.3 82.0 TA~LE 17 Release of NH3 -N (mq/ 100 m 1 ) f r om Soy Protein by Ruminal Microbes in 6-hour Batch Culture Permeate, % Degradability versus SBM, %

Source 0.0 4.0 Honeymead, Mankato, MN 36.6 27.0 73.6 Cargill, Savage, MN36.429.1 80.0 Cargill, Chicago, Il40.d33.1 82.3 Boone Valley Coop,39.3 29.7 75.6 Eagle Grove, I~

, 92 ~ ~3~7~

This example illustrates that it is possible to treat soy protein with sulfite liquor in such a manner that it will be protected from degradation by ruminal microbes, that the protection is not lost over long periods of storage time, nor is the protein's digestability by lower tract enxymes significantly reduced.
Solvent extract soybean meal was split and half was mixed to include 3 percent spent sulfite liquor solids, providing about 0.6 percent reducing sugars on soybean meal. The mixture was heated to 82 degrees Celsius by direct application of steam, pelleted, and returned to room temperature by evaporative cooling under a stream of forced air.
The entire heating and cooling cycle ~ook less than 5 minutes.
Pellets were ground and protein degradabi]ity by ruminal microbes determined in 6-hour batch culture. Ammonia nitrogen concentration in the 2Q treated pellets was only 47 percent of that generated with the pelleted soybean mea] control.
Because of this good response, this pair of samples was included in subsequent in vitro analysis over a 3-year period to provide a posi~ive control.

~3~7~

Results, expressed as percent degradability versus untreated soybean meal, are listed in ~able 18 and displayed in FIG~ 12. Protection against degradability is maintained through 40 months.
Variation is not due to the sample variability rather to the microbial populations used at different periods.
Samples were ar.alysed for pepsin digestible protein after storage for 37 months. The control soybean meal contained 43.1 percent digestible protein. Soy protein in the treated sample was ~1.1 percent digestible, indicating that no significant loss of protein had occurred during long term storage.

Commercial solvent extract soybean meal was split into four identical batches and mixed with reducing sugars as follows:
a. Controlp no additive b. 1 percent xylose c. 4 percent permeate from sulfite liquor d. 1 percent xylose and 4 percent permeate Concentration is expressed as weight percent on soybean meal, as is basis.

NT~3-N, mg/100 ml Months --------------------- Difference Stored Blank SBMSSL ~ _ 0 14.7 35.116.5 47.0 3 17.6 29.513.3 45.0 27.7 37.92~.7 75.7 6 ~1.0 33.320.4 61.3 lQ 12.3 17.55.5 31.1 16.4 28.616.2 56.5 24.5 37.222.6 60.7 17.6 28.214.2 50.3 27 16.2 27.412.1 44O3 28 11.9 24.89.2 37.2 28 16.9 32.812.9 39.5 29 16.8 32.313.3 41.3 29 16.3 27.914.4 51.6 39 22.6 32.g18.7 56.8 g5 :13~7~

Mixtures were conditioned to 85 degrees Celsius by direct s~eam addition, pelleted, and re~urned to room temperature by evaporative cooling under a forced air stream. Tota] heating period was less than 5 minutes. This portion of the process is described as treatment.
Hot pellets, approximately 100 grams, were collected in jars from each of the four batches (a-d) and placed in a 105 degree Celsius oven for 90 minutes, after which they were rapidly returned to room temperature through evaporative cooling under a steam of forced air. This portion of the process is described as treatment 2.
~reatment 3 was included as a positive control of known bypass value. This treatment consisted of soybean meal pelleted at 82 degrees Celsius, cooled, and stored for approximately 30 months. The pellets were comprised of soybean meal alone (treatment 3a~
or soybean meal mixed with 3 percent spent sulfite 2Q liquor prior to pelleting (treatment 3b).
Samples were tested for dye binding capacity and ammonia release by ruminal microbes in batch culture fermentation. Results are listed in table 19 ~

~5 96 ~3~7~

Treatment_ mg/~m mg/100 ml la 100.2 29.6 lb 110.7 27.4 lc 110.2 27.5 la 110.4 25.7 2a 118.4 29.7 2b 102.5 24.4 2c 96.2 23.1 2d 85.9 20.9 3a 101.2 27.9 3b 52.1 14.4 .

97 ~ 131~75~
Treatments 1 and 2 are arranged in a 2 b~ 3 factorial design. Analyses of the fermentation data shows additional heat, xylose~ and permeate each acted separately to reduce ln vitro N~3-N
concentration. Two factor interaction occurred between both heat and xy]ose and heat and permeate;
applica~ion of additional heat in the presence of either reducing sugar enhanced the degree of protection.

Method G: Naphthol slue Black The second goal of this experiment was to evaluate a new method of testing the degree of protein protection. Napthol blue black is know to bind to protein amino groups and to compete for these sites with other known protein protection agents, e.g., forma~dehyde. When soybean meal is added to a solution of dye, disappearance of the dye is an indictor of the protein content~ Lysine which has reacted with another reagent will not absorb dye from solution. Since the mechanism of protein protection reac~ion is believed to be binding of a reducing sugar to lysine in the protein mo]ecule, adsorption of napthol blue black by treated soybean can indicated the degree to which the protein has ` 98 ~ 3~75~

been succe~3fully protected when compared to ab~orption by untreated meal.
Dye solution wa~ prepared according to USDA Technical Bulletin No. 1369, "The Dye Binding of Milk Protein," by N. P.
Tarassuk, dated 1967, available from NTIS Deyt. of Commerce, order Dept., 5285 Port Royal Rd., Springfield, VA 22161, U.S.A. Samples were ground to pass a U.S. No. 20 seive and 0.100 gms of each placed in 50 ml centrifugc tubes. Thirty milliliters of dye ~olution waQ added to each tube, tubes were 3haken at room temperature for one hour, followed immediately by 15 minutes centrifuging at 2500 rpm. Exactly one milliliter of supernatant was withdrawn from each tube and diluted to 25 milliliters. Ab~orbance of this ~olution at 615 nanometer3 was determined using a spectrophotometer. Result~
were compared to the absorbance of a 1:25 dilution of ~tock dye of known concentration. Beer's Law allows calculation of the concentration of dye in the test aolution.
Dye binding capacity is determined by dividing the mass of dye the sample has absorbed by the ma~q of the sample.
Typically, untreated soybean meal has a dye binding capacity near 100 mg of dye per gram of sample. Dye binding capacity is compared to in vitro NH3-N in curve 110 in FIG. 13.
Correlation is good between the two te3t~.

:`

99 ~ 31 ~ 7~

The purpose of this experiment was ~o examine the usefu] range of xylose in trea~ing so]vent extract soybean meal. Soybean meal contains about 3.2 percen~ ]ysine. To react with all this ]ysine on an equimolar basis would require 3.5 xylose.
This could be considered the theoretical maximum.
Deviation from this maximum occurs if xylose reacts at other sites, i.e., the terminal amine, or, if xylose binding sites are not exposed due to the tertiary structure of the protein.
Several levels of xylose, listed in table 20, were disso]ved in distilled water anfl ~ixed into soybean meal to provide 20 percent added moisture.
From these mixtures, 0.100 gm samples were removed, placed in prewarmed centrifuge tubes, covered, and heated for 1 and 2 hours at 80 degrees Celsius~
Samples were removed from the oven, cooled, and tested for dye binding capacity.

Results ~curve 120, FI~. 14) show dye binding capacity to decrease through 20 percent addition, indicating the hinding sites were not yet saturated.
Additional heating reduced dye binding capacity at all levels of xylose, showing that the reaction had ~9 ~31~7~

D~C vs. Heating 2 hr. %
Change Xylose One ~
~ 1 hr. 2 hr. Control Control 0.0 77.2 72.1 0.0 0.5 6~.1 57.4 20.3 1.0 60.2 47.4 34.3 2.0 48.0 39.7 44.9 4.0 42.5 NA NA
10.0 36.3 33.5 53.5 20.0 33.8 29.2 59.5 .. . . . _ _ ...

13~ 154 in no cas~ gone to completion. It should be noted from an economical viewpoint that effectiveness per dose decreases rapidly; heated for 2 hours 20 percent xylose reduced dye binding capaci~y by 59.5 percent bu~ more than half of this reduction was provided by the first 1 percent xylose added.

12. In Vivo Examples Soybean mea] was metered into a So]idaire dryer at a rate of 4 kg/minute. ~he dryer was steam jacketed to allow application of indirect heat. A
spray of water, 8 percent xylose solution, or 30 percent spent sulfite liquor solution was applied to ~he meal as it fell into the dryer. This spray supplied 11 to 12 percen~ moisture to the soybean meal and acted as the carrier for the xylose, insuring it was dissolved and able to penetrate the flakes. Moistened soybean meal entered the dryer at ambient temperature (21 degrees Celsius~ and was retained for approximately three minutes, during which time it was heated to approximately 100 degrees Celsius. Hot feed exited the dryer and was transferrefl to an insulated container where it was ~3~5~

held ~or 45 minutes, fo]lowing which the feed was cooled and dried with ambient air.
Four lactating llolstein cows fitted with ruminal, duodenal and ileal cannulae were used in a 4 x 4 Latin square design to eva]uate treated soybean meal as a source of rumen protected protein.
Treatmen~s included untreated soybean meal, heated H2O-soybean mea], heated xylose-soybean meal and heated spent sulfite liquor soybean mea]. A diet consisting of 40 percent corn silage, 10 percent alfalEa cubes and 50 percent concentrate mix (dry matter basis) was fed four ~imes daily. Diets averaged 16.8 percent crude prctein with 50 percent of the total ration protein being derived from the respective soybean meal sources. Acid detergent lignin and diaminopimelic acid were used as digestibility and microbial markers, respectively.

13. Results The results are shown in table 21. They show treatment of soybean meal with spent sulfite liquor or xylose decreased ruminal N~3_N concentration, ruminal protein degradation, bacteria~ protein ~3~7~

1% 4%
Item SBM T-T20-SMB Xy~ose-sgM LS03-SBM

R~en N~3-N, ~g/100 ml 22.0 19.1 15.2 14.8 Ruminal protein degradation, ~ 70.6 fi9.6 55.8 53.7 Total tract protein digestion, % 77.4 75.4 73.6 71.4 Bacterial protein synthesis, g N/kg CMTD 41.5 34.9 31.4 33.4 .

lo~ ~ 13~7~

synthesis and total tract protein digestion compared to untreated soybean meal. Ruminal fiber digestion was not affected by treatment.
The data demonstrate controlled nonenzymatic browning is an effective method of protecting a highly degradable protein source like soybean meal from ruminal degradation and thereby increase efficiency of protein utilization for growth. These data further demonstrate similar responses in protein efficiency relative to commercial soybean meal when either xylose or glucose were used as reflucing sugars, though less heating was required when xylose was used due to its high rate of reactivity.
As can be understood from the above description, the novel feed, method of making the feed and method of feeding anima]s has the advantage of providing a superior economical feed and method of feeding animals.
2Q Although a preferred embodiment has been described with some particularity, many modifications and variations may be made in~ the preferred embodiment without deviating from the invention. Accordingly, it is to be understood that, within the scope of the appended claims, the 10~

105 ~ 7 ~ 4 in~tention may be practiced other than as speci f ica 1. ly descr ibed .

Claims (30)

1. A feed for animals comprising a mixture of organic materials including at least one reaction product of a feed protein and a reducing carbohydrate, the percentage of reducing carbohydrate on feed protein is about 0.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen microorganisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract.

2. A feed according to claim 1 wherein said feed protein is selected from a group consisting of soybean meal, other bean meal, cottonseed meal, feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early bloom legumes, fish products, by-product protein feedstuffs like distillers and brewers grains, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum, and mixtures thereof.

3. A feed according to claim 1 or 2 wherein said reducing carbohydrate is selected from a group consisting of sugar sources are selected from the reducing sugars xylose, ylucose, fructose, mannose, lactose, ribose, hemicellulose extracts and their hydrolysates, sugars contained in spent sulfite liquor, molassas and its hydrolysate and corn products and their hydrolysates, and mixtures thereof.

4. A feed according to claim 1 or 2 wherein said reducing carbohydrate is xylose and the percentage of xylose to feed protein is about 1 percent to 6 percent.

5. A feed according to claim 1 or 2 wherein said reducing carbohydrate is glucose and the percentage of glucose on feed protein is about 2 percent to about 20 percent.

6. A feed according to claims 1 or 2 wherein said reducing carbohydrate is a component of spent sulfite liquor of dried spent sulfite liquor.

7. A feed according to claim 1 or 2 wherein said reducing carbohydrate is a component of spent sulfite liquor or dried sulfite liquor and the spent sulfite liquor or dried spent sulfite liquor includes about 10 percent to about 40 percent reducing carbohydrates on solids and the percentage of spent sulfite liquor solids on feed protein is about 2 percent to about 40 percent.

8. A feed for animals according to claim 1 including at least one reaction product of feed protein and spent sulfite liquor or dried spent sulfite liquor, the percentage of spent sulfite liquor solids on feed is about 3.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen microorganisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract.

9. A feed according to claim 8, wherein said percentage of spent sulfite liquor solids on feed protein is about 8 percent to about 25 percent by weight.

10. A feed according to claim 8 wherein said spent sulfite liquor or dried spent sulfite liquor is obtained from the pulping of hardwoods.

11. A method of making an animal feed comprising the steps of: providing a mixture of a feed protein and a reducing carbohydrate, the percentage of reducing carbohydrate on feed protein being about 0.5 percent to about 40 percent by weight; and heating the mixture at a temperature, pH
and percent moisture for a time sufficient to reduce the degradability of the feed protein by rumen microorganisms and provide no significant reduction in protein digestibility in the post rumen tract.

12. A method according to claim 11 wherein the pH is from about 4 to about 10.5

13. A method according to claims 11 or 12 wherein said percent moisture is from about 6 percent to about 40 percent.

14. A method according to claims 11 or 12 wherein said temperature is from about 20 degrees centrigrade to about 150 degrees centrigrade.

15. A method according to claims 11 or 12 wherein said time is from about 20 minutes to about 72 hours.

16. A method according to claims 11 or 12 wherein the pH is from about 6 to about 8.5.

17. A method according to claims 11 or 12 wherein said percent moisture is from about 15 percent to about 25 percent.

18. A method according to claims 11 or 12 wherein said temperature is from about 80 degrees Celsius to about 110 degrees Celsius.

19. A method according to claims 11 or 12 wherein said time is from about 1 hour to about 4 hours.

20. A method according to claim 11 wherein said feed protein is selected from a group consisting of soybean meal, other bean meal, cottonseed meal, feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early bloom legumest fish products, by-product protein feedstuffs like distillers and brewers grains, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, sorghum, and mixtures thereof.

21. A method according to claim 11 wherein said reducing carbohydrate is selected from a group consisting of sugar sources are selected from the reducing sugars xylose, glucose, fructose, mannose, lactose, ribose, hemicellulose extracts and their hydrolysates, sugars contained in spent sulfite liquor, molasses and its hydrolysate and corn products and their hydrolysates, and mixtures thereof.

22. A method of making an animal feed according to claim 11 wherein a mixture of a feed protein and a spent sulfite liquor or a dried spent sulfite liquor is provided such that the percentage of spent sulfite liquor solids on feed protein is about 2 percent to about 40 percent by weight; and heating the mixture at a temperature, pH and percent moisture for a time sufficient to reduce the degradability of the feed protein by rumen microorganisms and provide no significant reduction in protein digestibility in the post rumen tract.

23. A method according to claim 22 wherein the percentage of spent sulfite liquor solids on feed protein is about 8 percent to about 25 percent by weight.

24. A method according to claim 22 wherein said spent sulfite liquor or dried spent sulfite liquor is obtained from the pulping of hardwoods.

25. A method of feeding animals comprising the steps of: selecting a protein-containing feed; and feeding to the aminals a reaction product of the feed protein and a reducing carbohydrate wherein the percentage of reducing carbohydrate or feed protein is about 0.5 percent to about 40 percent by weight, such that degradability of the feed protein by rumen microorganisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract.

26. A method of feeding animals according to claim 25 wherein said feed protein is selected from a group consisting of soybean meal, other bean meal, cottonseed meal, feather meal, blood meal, silages, meat and bone meal, sunflower seed meal, canola meal, peanut meal, safflower meal, linseed meal, sesame meal, early bloom legumes, fish products, by-product protein feedstuffs like distil]ers and brewers grains, milk products, poultry products, hays, corn, wheat, alfalfa, barley, milo, soryhum, and mixtures thereof.

27. A method of feeding animals according to claims 25 wherein said reducing carbohydrate is selected from a group consisting of sugar sources selected from the reducing sugars xylose, glucose, fructose, mannose, lactose, ribose, hemicellulose extracts and their hydrolysates, sugars contained in spent sulfite liquor, molasses and its hydrolysate and corn products and their hydrolysates, and mixtures thereof.

28. A method of feeding animals according to claim 25 wherein a protein-containing feed suitable for a ruminant is selected; and the ruminant is fed a reaction product of the feed protein and a spent sulfite liquor or a dried spent sulfite liquor wherein the percentage of spent gulfite liquor solids on feed protein is about 2 percent to about 40 percent by weight such that degradability of the feed protein by rumen microorganisms is reduced and there is no significant reduction of protein digestibility in the post rumen tract.

29. A method according to claim 28 wherein the percentage of spent sulfite liquor solids on feed protein is about 8 percent to about 25 percent by weight.

30. A method according to claim 28 wherein said spent sulfite liquor or dried spent sulfite liquor is obtained from the pulping of softwoods.