VIBRAR.Y 
 
 OF THE 
 
 UN1VLR.SITY 
 OF ILLINOIS^ 
 
 630.7 
 UGb 
 
 cop 
 
 AGRICULTURE 
 
NOTICE: Return or renew all Library Materials! The Minimum Fee for 
 each Lost Book is $50.00. 
 
 The person charging this material is responsible for 
 its return to the library from which it was withdrawn 
 on or before the Latest Date stamped below. 
 Theft, mutilation, and underlining of books are reasons for discipli- 
 nary action and may result in dismissal from the University. 
 To renew call Telephone Center, 333-8400 
 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN 
 
 
 L161 O-1096 
 

UNIVERSITY OF ILLINOIS 
 
 Agricultural Experiment Station 
 
 BULLETIN No. 289 
 
 EXPERIMENTS IN THE SELF-FEEDING 
 OF DAIRY COWS 
 
 BY W. B. NEVENS 
 
 URBANA, ILLINOIS, MAY, 1927 
 
SUMMARY 
 
 Eight dairy cows were fed by the self-feeder or free-choice method 
 of feeding for periods ranging from 14 to 130 weeks. 
 
 There was a general tendency for the cows to consume nutrients 
 greatly in excess of their requirements as estimated from the Armsby 
 feeding standard. For a short period following calving, however, sev- 
 eral cows failed to consume sufficient nutrients to meet their estimated 
 needs. With the exception of one case, protein was consumed in 
 greater excess than net energy. 
 
 The high level of feed intake resulted in a marked increase in 
 weight of most of the cows and presumably, therefore, a higher main- 
 tenance cost. It seemed at first that some of the cows would gradually 
 accustom themselves to a consumption of nutrients upon a level with 
 their requirements, but such an adjustment failed to appear in the 
 period covered by the experiments. For the reasons given, therefore, 
 self-feeding proved uneconomical. 
 
 The self-feeder method proved useful in studying the relative pal- 
 atability of different feeds. Not all cows showed preferences for the 
 same feeds, however. Some were consumed freely to the almost entire 
 exclusion of others to which the cow had access. The preferences of 
 several of the cows changed suddenly and the marked partiality for a 
 certain feed often extended over a long period. Mixed concentrates 
 proved less palatable than the same concentrates fed separately, as 
 judged by the total concentrate consumption, altho there were excep- 
 tions to this. No two cows exhibited the same preferences for all feeds. 
 
 Corn proved the most palatable of the low-protein concentrates 
 used. Even tho concentrates were consumed in very large amounts, 
 the amount of roughage eaten remained at a high level. 
 
 The few records of milk production secured failed to indicate that 
 milk secretion was stimulated by the self-feeder method as compared 
 with the usual method of hand-feeding. Limiting cows to high-protein 
 feeds only, seemed to have a depressing effect upon milk production. 
 
 No harmful effects from self- feeding upon the health or reproduc- 
 tive functions of the cows were shown within the period of the experi- 
 ment, but there is some possibility that injurious effects upon the body 
 tissues were being gradually brought about in cows limited to high- 
 protein feeds only. It may be that some physiological disturbances 
 were responsible for the lessened milk production which accompanied 
 the high-protein feeding. 
 
EXPERIMENTS IN THE SELF-FEEDING 
 OF DAIRY COWS 1 
 
 By W. B. NEVENS, Assistant Chief in 
 Dairy Cattle Feeding 
 
 The self-feeder or free-choice method of feeding both roughages 
 and concentrates has been employed successfully with a number of 
 different species of animals, such as swine, sheep, cattle, and poultry, 
 under both experimental and practical farm conditions. This plan of 
 feeding has been especially favored in the fattening of animals for 
 market, since it usually results in the consumption of much larger 
 quantities of feed and thus more rapid and economical gains than 
 hand-feeding. There is also, as a rule, a saving in labor. 
 
 Altho this method has been used for many years in the feeding of 
 beef cattle, but little use of it has been made with dairy cattle, ex- 
 cept to a limited extent in raising calves and in the feeding of rough- 
 ages. 2 
 
 The idea that greater milk production might be secured from 
 dairy cows by the free-choice method of feeding than by the usual 
 plan of hand- feeding was the chief motive for undertaking this inves- 
 tigation. It seemed reasonable to assume that if a cow were given ac- 
 cess constantly to a supply of feeds known to be suited to milk pro- 
 duction, she might eat more than when hand-fed and her milk yield 
 be thereby increased. Several other questions connected with the self- 
 feeding of dairy cows were also studied: namely, the economy of the 
 method; whether cows given freedom in selecting their own feeds 
 would consume the different ones in such amounts that the proper pro- 
 portions would be maintained among the several nutrients; and 
 whether self-feeding could be employed without injury. The relative 
 palatability of the different feeds was also a point of interest in the 
 study. 
 
 So far as the author has been able to ascertain, the first feeding 
 trial with dairy cows in which the free-choice method was used ex- 
 clusively was carried out at the Illinois Station under the direction of 
 Professor W. W. Yapp in 1922 (Holler 5 ). In view of the thoro man- 
 
 Submitted for publication April 15, 1926. 
 
 'Investigations in the feeding of dairy calves by the free-choice method have 
 been reported by Hulce 1 at the Illinois Station, McCandlish 2 at Iowa, and Nev- 
 ens 3 at Nebraska. Hunt 4 conducted a feeding trial in which a group of cows, after 
 being fed in the usual manner in the stable each morning, was turned into a lot 
 containing a self-feeder, where they remained until night. 
 
 427 
 
428 BULLETIN No. 289 [May, 
 
 ner in which that work was done, the results obtained are presented 
 here in connection with those secured by the author. a 
 
 HOW THE COWS WERE FED AND THE RECORDS 
 SECURED 
 
 Eight cows of moderate productivity were used in the experi- 
 ments. Until it could be demonstrated that the self-feeder method 
 might be used without injury, it was not deemed advisable to place 
 the highest-producing cows in the herd under an experiment of this 
 kind. 
 
 Records of feed consumption, live weight, and milk production 
 were secured by the author for four cows for. periods of different 
 lengths, one cow being under experiment 60 weeks, another 78 weeks, 
 the third 104 weeks, and the fourth 130 weeks. Similar records were 
 secured by Holler 5 for one cow for 26 weeks, one cow for 20 weeks, and 
 two cows for 14 weeks each. The cows were kept in individual box 
 stalls in each of which a self-feeder provided a continuous supply of 
 several different feeds (Fig. 1). They had access to the feeds at all 
 times except for periods during the day when they were turned out- 
 doors for exercise and water. The length of the exercise periods varied 
 from one to several hours, depending upon weather conditions. 
 
 The self-feeder for concentrates contained four or five hopper 
 compartments in which separate feeds were placed as needed in order 
 to keep a constant supply of each. Silage and hay were usually fed in 
 larger, separate compartments, some of which were provided with slat- 
 ted covers to prevent loss of feed. Water was provided in some stalls 
 by means of automatic water bowls. In the experiments by Holler the 
 bedding consisted of straw, but in the subsequent experiments shavings 
 were the only material used for this purpose. 
 
 The cows were gradually accustomed to the experimental condi- 
 tions during a preliminary period of two to three weeks. Roughage 
 and a concentrate mixture were offered in gradually increasing 
 amounts until the cows were consuming all they cared for. Each con- 
 centrate was then placed in a separate compartment. 
 
 Records of feed consumption were secured by weekly periods and 
 in some of Holler's work each day. At the end of each experimental 
 
 *The facilities of the Department of Dairy Husbandry were placed at the 
 disposal of Mr. L. S. Holler, a student in the College of Agriculture, for this study. 
 Most of the records of feed consumption and live weight for two of the cows 
 were obtained by Mr. Holler in person, while the same data for the other two, to- 
 gether with records of milk production and butterfat tests for all four cows, were 
 secured by herdsmen and laboratory helpers in the same manner as in the feeding 
 trials conducted by the author. 
 
.7557] EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 429 
 
 week the feed remaining in each compartment was weighed and the 
 amount recorded. Usually the feed was returned to the compartment 
 and, if necessary, more feed added. Occasionally the feeds remaining 
 in the hoppers at the end of the week were discarded and fresh supplies 
 ^substituted. 
 
 Common salt was supplied in a separate compartment, and in 
 some cases records of salt consumption were kept. Steamed bone meal 
 or ground limestone was also supplied to four cows and some records 
 
 FIG. 1. EQUIPMENT USED IN SELF-FEEDING DAIRY Cows 
 
 Each cow was kept constantly in a box stall where 
 she had free access to a continuous supply of a number 
 of feeds. No changes in the feed were made at calving 
 time. 
 
 of their consumption secured. All cows were weighed weekly. No 
 variation in the method of feeding or general routine was made on ac- 
 count of gestation or parturition. 
 
 The milk produced at each milking was weighed and sampled. 
 Seven-day composite samples of the milk were preserved by means of 
 mercuric chlorid tablets and were tested for butterfat by means of the 
 Babcock test. A digestion trial was conducted with one cow while be- 
 ing self-fed and in a subsequent period while being hand-fed on a lim- 
 ited ration. 
 
 The feed records show that the amount of concentrates in certain 
 compartments from which the cows apparently had not eaten often 
 weighed from .1 to .2 of a pound and occasionally as much as .3 of a 
 pound more at the end of the week than at the beginning. This may 
 have been due to several causes. The feeds in being exposed to the 
 
430 
 
 BULLETIN No. 289 
 
 [May, 
 
 
 
 GO 
 
 
 
 
 
 
 
 ll 
 
 g > s 
 
 ,j 
 
 
 i i 
 
 
 
 
 11 
 
 ^ ~" 10 Tf co ia o 
 
 
 
 
 
 
 u gj 
 G 
 
 
 
 
 
 B 
 
 e 
 
 } 
 
 1 
 
 C g 
 
 S C^ ^~^ OO J"" 1 1 1 ^ t^ CO OO 
 
 O5 
 CD 
 
 to 
 
 
 
 a 
 
 '! 
 
 ? GO '"^ C^ t^ O5 CO '^ '"^ 
 
 
 
 * 
 
 72 
 
 ) 
 
 "S 
 
 ^ 
 
 
 <N 
 
 
 -J 
 
 c 
 
 
 ^^ 
 
 
 
 
 
 S 
 
 
 . 
 
 
 
 
 
 S 
 
 ] 
 
 
 
 
 
 
 a 
 
 a 
 
 
 
 
 
 
 c 
 
 3 
 
 
 
 
 
 
 '3 
 
 
 
 _ 
 
 
 
 
 a* 
 
 
 
 
 
 
 
 GU 
 
 5 
 
 
 1 
 
 3 
 
 g--iOaeocoio>ooo 
 
 (N 
 
 CO 
 
 <9 
 X 
 
 h 
 
 o 
 
 5 
 
 j 
 
 
 SCO 
 . G 
 
 <J o 
 
 l^o^ic^^^^^ 
 
 O5 
 
 co 
 
 
 o 
 
 
 
 O 
 
 
 
 
 b 
 
 i 
 
 
 
 
 
 
 C 
 
 i 
 
 
 
 
 
 
 
 4 
 
 3 
 
 
 
 
 
 
 '"O 
 C 
 
 
 
 oo 
 , ^i 
 
 
 
 
 o 
 
 i? 
 
 i- 
 
 
 1|I 
 
 'go'* -*-<N-0 
 
 
 11 
 
 1 
 
 o 
 u 
 
 03 
 T3 
 
 e 
 
 * -rJ 
 
 
 o d '3 
 
 p. co 1 
 
 
 
 T3 
 
 5 
 
 
 
 
 
 
 
 03 
 
 2 3 
 
 'S 
 
 (-, 
 
 
 
 
 
 f Standard 
 
 1 
 
 a 
 
 0) 
 
 3 
 
 |1 
 3.C 
 
 2 3 
 
 S er 
 
 -o *" 
 
 CO 
 eo 
 
 -M 
 CO 
 
 GO 
 
 3 
 
 1 
 
 o 
 
 ~ ^f*^ 
 
 3 
 
 <3 S 
 
 
 
 
 u 
 
 s 52 
 2 1 
 
 c* *^ 
 
 '-3 
 
 
 QJ 
 
 
 
 
 
 M 
 
 _o 
 
 S <i 
 
 3 C^ 
 
 S 
 
 
 
 
 
 
 
 r 
 
 t 
 
 
 il 
 
 3 S 
 
 .gsssggg^g 
 
 o 
 
 CO 
 
 bC 
 *C 
 
 3 
 
 ; 
 
 
 O 3 
 Son 
 . G 
 
 16 
 
 ~ i-H CO rH 
 
 ^ 
 
 
 3 
 
 a 
 
 i 
 
 
 J O 
 O 
 
 
 
 
 "s 
 
 a 
 
 
 
 
 
 
 'C 
 
 B 
 
 
 
 
 
 
 4-i 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 3 
 
 
 t_ 
 
 
 
 
 oo 
 
 
 
 O "^ 
 
 
 
 
 ,9 
 
 
 
 (~" Q) 
 II 
 
 00 r^ 00 CO 10 T)H TJH 
 
 
 ^ 
 
 . V 
 
 1 
 
 
 
 
 
 
 
 *^ _c 
 
 a 
 
 i 
 
 < 
 
 
 o3 o 
 3 a 
 P 
 
 ^! r-< rH 
 
 
 
 x-5 
 
 '2 
 
 0) O 
 -kJ -u 
 03 - 
 
 4 
 
 
 
 
 
 
 c> 
 
 -S -u 
 
 
 
 
 
 
 
 G^ 
 
 
 
 
 
 
 
 S^H 
 
 
 
 
 
 
 
 O t. 
 
 
 
 
 
 
 
 G 
 
 
 
 
 
 
 
 o **^ 
 
 
 
 > 
 
 
 
 
 T3 
 
 
 
 O 
 
 
 
 
 03 b 
 
 
 
 o 
 
 
 
 
 T5 8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r -yi 
 
 
 
 
 
 Ul 
 
 
 ,? e? 
 
 
 
 
 lor-^coTtidSsd 
 
 "o 
 
 5 
 
 
 
 
 
 IM o c<i ca co co 
 
 H 
 
 ?i 
 
 
1927} 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 431 
 
 stable air for a week probably absorbed some moisture. The cows 
 sometimes nosed over the feeds without eating them, thus possibly 
 adding moisture to them. It was also observed that a cow while eat- 
 ing from one compartment moved her head about and sometimes 
 dropped small amounts of feed into other compartments or on the 
 floor. Hay and other feeds spilled on the floor were gathered up fre- 
 quently and returned to the proper compartments. While an effort of 
 
 NET tNLRGY -hmtRMS 
 
 (AMOUNT! CONSUMLP Wt 
 
 FIG. 2. Cow 25 WAS AN ECONOMICAL CONSUMER OF NET ENERGY BUT Nor OF 
 DIGESTIBLE TRUE PROTEIN 
 
 This chart shows the large amount of digestible protein supplied by alfalfa 
 hay, the only feed consumed. (See Fig. 10 for feed record.) Except during the 
 week following calving, the alfalfa hay eaten furnished 70 percent more protein 
 than was needed for body maintenance and liberal milk production. Greater milk 
 production would doubtless have been secured had some concentrates been con- 
 sumed, for at times the alfalfa hay eaten was not sufficient to meet the net energy 
 needs. With alfalfa hay as the sole feed an excess of protein was of course sup- 
 plied whenever enough hay was consumed to meet the net energy needs. 
 
 course was made to keep the records as accurate as possible, it was 
 impossible to avoid the relatively small errors incurred in these ways. 
 The amounts of digestible true protein and net energy in the feeds 
 consumed have been calculated in order to form a basis of comparison 
 with the estimated requirements for maintenance and milk produc- 
 tion. 4 
 
 "The apparent live weights of cattle fluctuate greatly from day to day and 
 from week to week, owing to variations in the water content of the body, rate of 
 excretion of waste, feed intake, and various other causes. Variations in feed in- 
 take were very evident in this investigation. In order to secure a suitable basis 
 for the calculation of data relative to live weight, the observed weights were 
 smoothed according to a method given by Rietz. 8 
 
 The digestible true protein and net energy contents of the feeds consumed, 
 were calculated according to the average values published by Borland. 7 Analyses 
 were made of some of the silage and hay fed, but these have not been used ex- 
 cept in calculating the results of the digestion trial. 
 
 In order to provide a suitable basis for comparison with the amounts of nu- 
 trients consumed, the requirements of digestible true protein and net energy for 
 
432 
 
 BULLETIN No. 289 
 
 [May. 
 
 x g r 
 
 8 3 
 
1927} 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 433 
 
 * s 
 
 CdNTTOd - 
 
 wnioad 
 o 
 
 4 a B Mt4*a3eM4044WSaMUM7EMMM9tMHOI04loaUZIMIOIMItai3t 
 WU.K3 
 
 Yio. 4. AN UNUSUALLY Low INTAKE OF DIGESTIBLE TRUE PROTEIN CHARACTERIZED THE FEED RECORD OF Cow 576 
 
 This cow continued to produce milk at quite a high level and to gain in weight (Fig. 12) altho her intake of digestible 
 true protein over a period of six months was continuously lower than the estimated requirements. During the experiment 
 as a whole, also, the total consumption of this nutrient was less than the estimated requirement. The net energy consump- 
 tion was much above the requirement, which is characteristic of self-fed cows offered low-protein concentrates. 
 
 
 
 ^ 
 
 - 
 
 
 <r 
 
 
 
 
 , 
 
 > 
 
 
 1 
 
 i 
 
 
 
 
 2 
 
 / 
 
 -^ 
 
 \ 
 
 t 
 
 
 
 
 
 
 { 
 
 L 
 
 ^ 
 
 
 
 J 
 
 y 
 
 ' 
 
 f 
 
 
 
 
 2 
 
 I 
 
 
 I 
 
 
 
 
 ^ 
 
 J 
 
 
 I 
 
 
 
 
 < 
 
 > ( 
 
 
 u 
 
 
 
 
 /- 
 
 z 
 
 
 
 
 
 3 
 
 
 / 
 
 
 7 
 
 
 
 
 \ 
 
 \ 
 
 j 
 
 \ 
 
 
 
 
 \ 
 
 / 
 
 
 
 
 
 
 \ 
 
 r 
 
 1 
 
 
 
 
 
 
 
 ^ 
 
 V 
 --J 
 
 \ 
 
 
 ll 
 
 It* 
 
 
 r 
 
 
 7 
 
 "(j 
 
 i 
 
 l| 
 
 
 r^ 
 
 
 j \ 
 
 j 
 
 .11 
 
 at! 
 
 
 
 
 
 
 
 H 
 
 i 
 
 L 
 
 > 
 
 > 
 
 I 
 
 
 
 
 
 > 
 
 
 \ 
 
 
 
 
 \ 
 
 
 
 i 
 
 
 
 
 j~ 
 
 > 
 
 1 
 
 \i 
 
 
 
 ^ 
 
 i 
 
 
 
 i 
 
 
 
 \ 
 
 
 \ 
 
 
 1 
 
 
 
 ^ 
 
 
 i 
 
 
 j( 
 
 
 
 L 
 
 > 
 
 \ 
 
 
 G 
 
 
 
 
 ^ 
 
 , 
 
 
 
 ' 
 
 
 .- 
 
 
 ' 
 
 > 
 
 rj 
 
 
 < 
 
 
 
 / 
 
 
 i / 
 
 Li 
 
 
 
 X 
 
 
 
 
 2 
 
 
 
 < 
 
 -^ 
 
 ( 
 
 
 S 
 
 
 
 -=s 
 
 
 i 
 \ 
 
 
 i\ 
 
 
 
 <c 
 
 -^-^ 
 ^ 
 
 ^_ 
 
 C 
 
 - 
 
 -. 
 
 ^ 
 
 ^T 
 
434 
 
 BULLETIN No. 289 
 
 [May, 
 
 SELF-FEEDER METHOD NOT SO ECONOMICAL 
 AS HAND-FEEDING 
 
 That the self-feeder method for dairy cows is not economical from 
 the standpoint of the amount of feed consumed may be seen by a 
 glance at Table 1 and Figs. 2 to 9, in which the amounts of digestible 
 true protein and net energy consumed are compared with the require- 
 ments of the animals for these substances as set forth in the Armsby 
 feeding standard. 
 
 k 
 
 -V 1 
 
 FIG. 5. Cow 267 RESEMBLED Cow 25 IN BEING AN ECONOMICAL CONSUMER OF 
 NET ENERGY BUT NOT OF DIGESTIBLE TRUE PROTEIN 
 
 During the first 15 weeks corn was offered, with the result that the intake of 
 net energy was considerably above the requirements, but during the remainder of 
 the experiment, when the concentrates consisted of high-protein feeds only, the 
 consumption of net energy was variable but on the whole slightly below the re- 
 quired amount. The difference between the amounts of digestible true protein 
 consumed and required was much less than for Cow 27. 
 
 Most of the cows at times consumed from 50 to 100 percent more 
 net energy than their estimated requirements. The cows having ac- 
 cess to some low-protein feeds, particularly corn, in general consumed 
 a greater amount of net energy above their requirements than did the 
 cows offered high-protein feeds or a concentrate mixture only. 
 
 The charts do not, however, tell the whole story with respect to 
 the economy of the method. Most of the cows remained in a high 
 
 maintenance and milk production were calculated according to the Armsby stand- 
 ard as published by Borland. 7 It is recognized that feeding standards are merely 
 guides to aid in securing approximately accurate results in practical feeding; 
 while they may be far from exact under the widely varying conditions of practice 
 and also of these experiments, it is believed that they afford a useful basis for 
 studying the data secured in an investigation such as this. 
 
 Milk yields were corrected on the basis of their gross energy values to a uni- 
 form basis of milk containing 4 percent fat, according to the method of Gaines 
 and Davidson.' This method assumes that one pound fat-corrected milk (F.C.M.) 
 has an energy value of 331 large calories. Correction is accomplished thru the 
 application of the formula, F.C.M. .4 milk (in pounds) + 15 fat (in pounds). 
 
1987] 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 435 
 
 condition of flesh except for short periods at the beginning of the ex- 
 periment and following calving. It has been shown, 9 - 10 that cattle fed 
 on a high plane of nutrition fail to make as efficient use of their feed as 
 when fed upon a low plane, and that these differences in extreme cases 
 may be as much as 30 percent. Cow 576 weighed about 150 pounds 
 more at the time of second calving during the self-feeder trial than at 
 the previous calving which occurred early in the experiment. Cow 27 
 
 26 
 
 22 
 
 R 16 
 14 
 
 155 
 1EO 
 
 <n' 05 
 
 2 00 
 
 
 
 
 Cow 
 
 274 
 
 
 
 
 --A 
 
 Zl 
 
 2; 
 
 V 
 
 
 
 /-- 
 
 
 A 
 
 f 
 
 
 
 
 
 
 
 " 
 
 
 
 s 
 
 / 
 
 
 
 
 
 PIG. 
 " PIG. 
 
 TRUE PR 
 
 TRUE PR 
 
 OTEIN 
 DTEIN R 
 
 3N3UMEC 
 EQUIRED 
 
 WEEKLY 
 WEEKLY 
 
 
 
 
 
 /\ 
 
 
 
 
 
 
 
 / 
 
 '\ 
 
 
 
 
 
 
 
 / 
 
 \ 
 
 A 
 
 I 
 
 y' r ^ 
 
 ^~'- 
 
 
 ~~^\ 
 
 / 
 
 \I 
 
 s / ' 
 
 \ / 
 
 ,V \ 
 
 / 
 
 
 5 
 ,75 
 
 \ 
 
 1 
 
 
 ^ 
 
 \| 
 
 
 
 
 
 
 
 
 
 
 
 
 45, 
 
 
 
 
 x v_ 
 
 1 
 
 
 
 
 3 4 fl 1Z 16 ao E4 26 
 
 WEEKS NE.T E.NERGX CONSUMEP WEEKLY 
 NET LNLROY REQUIREP WEEKLY 
 
 FIG. 6. RATIONS OF SELF-FED Cows FAIL TO MEET RE- 
 QUIREMENTS FOR A SHORT TIME AFTER CALVING 
 
 The record of the nutrient consumption of Cow 
 274 is typical of self-fed cows in showing an uneco- 
 nomical consumption of feed except for a few weeks 
 after calving, when the nutrient intake is too low. 
 
 at the time of calving, during the 34th week of the experiment, weighed 
 about 175 pounds more than at her previous calving. These are in- 
 creases of about 12.5 percent and 14 percent respectively in the weights 
 of the cows, largely due to the method of feeding, altho they are to be 
 partly accounted for by growth. If it be assumed that the efficiency of 
 the feed was reduced when the cows were in a high state of flesh, then 
 it would appear that the feed cost of maintenance when cows are self- 
 fed over long periods is greatly in excess of that incurred by the usual 
 method of hand-feeding. 
 
 From the standpoint of digestible true protein the amounts of feed 
 consumed were quite uneconomical in most cases. The amount of net 
 energy in the feed of Cows 25, 267, and 649 conformed more nearly to 
 the requirements than in the case of the other cows, yet at times the 
 feed contained 50 to 100 percent more digestible true protein than 
 
436 
 
 BULLETIN No. 289 
 
 [May, 
 
 26 
 34 
 
 ? 
 
 1 16 
 12 
 
 165- 
 130- 
 <o 135- 
 S 120 
 105 
 
 
 c< 
 
 ?w290 
 
 
 
 
 
 
 
 7 
 
 
 / 
 
 \ /\ 
 
 
 / 
 
 : 
 
 ^y 
 
 ^ \ 
 
 
 / 
 
 
 
 \ 
 
 ib J- 
 
 / - 
 
 bi 
 
 G.TRUE R 
 
 R0TE1N Ci 
 
 DNSUMtD 
 
 WEEKLY 
 
 PI 
 
 3. TRUE P 
 
 ROTEJN F 
 
 t<?01RE.P \ 
 
 VE.E.KLY 
 / 
 
 
 A 
 
 ,/\ 
 
 
 X 
 
 , 
 
 ^ A 
 
 / I 
 
 
 / 
 
 / 
 
 V 
 
 
 
 / 
 
 
 
 
 
 / 
 
 
 
 
 ^=~^ 
 
 / ~ 
 
 
 N 
 
 EXlNLK 
 
 Y CON51 
 
 )MtP WE: 
 
 _KLY 
 
 N 
 
 IT LNLRC 
 
 T RLOUI 
 
 ?EP Wtti 
 
 cr 
 
 4 6 \Z Ifc C 
 WLLKS 
 
 FIG. 7. HAND-FEEDING PROVED MORE ECONOMICAL THAN SELF-FEEDING 
 WITH THIS Cow 
 
 From the 12th to 16th weeks, while Cow 290 was hand-fed, the intake of nu- 
 trients was practically equal to requirements. Previous and subsequent to this 
 period, while this cow was self-fed, the consumption of nutrients was greatly in 
 excess of requirements. (See Fig. 16 for records of weight and milk production). 
 
 FIG. 8. SUDDEN CHANGES OCCUR IN THE AMOUNTS OF NUTRIENTS THAT SELF-FED 
 
 Cows CONSUME 
 
 Cow 649 was fed a grain mixture instead of being offered the various in- 
 gredients separately. For a time she consumed feed much in excess of her re- 
 quirements, while for a few weeks her feed consumption was too low. 
 
19871 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 437 
 
 1I6CH 
 
 24 
 
 /N 
 
 /-/ 
 
 v\. 
 
 ^ 
 
 / 
 
 
 
 .. 
 
 \ 
 
 PlGH 
 PlG-1 
 
 7UE.PR01 
 
 RUC.PROI 
 
 E.1N CONS 
 UNRtOU 
 
 IMtP 
 IREP 
 
 
 
 ^ 
 
 A 
 
 r 
 
 
 s 
 
 n 
 
 ---, 
 
 
 
 
 
 
 
 NET 
 NET 
 
 LNEKGY 
 LNLRQY 
 
 ^ONSUN 
 
 REQUIRE 
 
 EP 
 
 P 
 
 FIG. 9. THE TOTAL NUTRIENT INTAKE OF Cow 650 REMAINED CONSISTENTLY HIGH 
 
 IN CONTRAST TO THAT OF Cow 649, EVEN THO THE PREFERENCES 
 
 OF BOTH Cows SHIFTED FREQUENTLY 
 
 The nutrients consumed by Cow 650 remained much above the estimated 
 requirements, altho extreme changes occurred in the amounts of the different 
 feeds eaten. 
 
 V 
 
 FIG. 10. ALFALFA HAY PREFERRED TO PROTEIN CONCENTRATES BY Cow 25 
 
 Alfalfa hay was the only feed consumed by this cow during the entire ex- 
 perimental period of 60 weeks, even tho she had access to a supply of linseed oil 
 meal, cottonseed meal, and soybean oil meal. Note that following calving there 
 was liberal milk production, together with an increase in live weight, with alfalfa 
 as the sole feed. 
 
438 
 
 BULLETIN No. 289 
 
 {.May, 
 
 
 
 I2AO 
 
 1160 
 
 i iao 
 
 FIG. 11. IN CONTRAST TO Cow 25, Cow 27 CONSUMED UNUSU- 
 ALLY LARGE AMOUNTS OF PROTEIN CONCENTRATES. Variations in the 
 preferences for different feeds are clearly evident from this record. 
 On the whole there was a liberal consumption of feed, resulting in a 
 marked increase in live weight. This cow ate not only large amounts 
 
 60 
 
 I 
 
 I 50 
 
 1 
 
 U 90 
 
 FEEP AlflP MIL> 
 LIVE. Wtl 
 
 COW 
 
 (F.C.M.} IN POU4JP3 PER 
 r WEEKLY INT 
 
 PRIEP E.ET PUi P 
 
 v sfre 
 :.M.i IA 
 
 WttEL 
 
 FIG. 12. LOW-PROTEIN FEEDS PROVED so PALATABLE TO Cow 576 THAT BARELY ENOUGH 
 WHEAT BRAN WAS CONSUMED TO MEET THE REQUIREMENTS FOR PROTEIN. One feed me- 
 dium high in protein (wheat bran) was offered in connection with low-protein roughages 
 and concentrates, so that the cow would have an opportunity to balance her ration with 
 
1927] 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 439 
 
 of legume hay but also exceptionally large quantities of high-protein 
 mill by-products. Peanut oil meal was offered from the beginning 
 of the experiment, but none of it was consumed until the 33d week; 
 the maximum quantity consumed weekly was greater than that of 
 any other concentrate. 
 
 LIVE. WEIGHT 
 
 A 
 
 jr\ 
 
 / ^1 
 
 espect to protein. The animal showed a marked preference, however, for low-protein 
 feeds, particularly corn and oats. Corn was especially favored during the first half of the 
 experiment, but later oats took precedence over the other feeds. 
 
440 BULLETIN No. 289 [May, 
 
 needed. An excessive quantity was contained in the feed of Cow 27 
 during the 45th to 51st week. This amounted to 7 to 9 pounds daily 
 above her estimated requirements, and during the 104 weeks the 
 amount was 304 percent more than her calculated needs. 
 
 The data at hand are too limited to draw a sharp distinction be- 
 tween the high and low-protein regimes with respect to economy of 
 protein consumption. The records secured so far indicate that the 
 character of the feeds offered may have little influence upon the pro- 
 tein intake, provided the cow has access to feeds which will supply 
 sufficient protein to meet her needs amply. Cow 576 seemed to be a 
 marked exception, since she failed to consume enough protein to meet 
 her needs as estimated by the standard. 
 
 MOST SELF-FED COWS DO NOT SELECT PROPER 
 AMOUNTS OF NUTRIENTS 
 
 The data are very conclusive in showing that the cows fed by the 
 free-choice method did not select feeds in the proper amounts and pro- 
 portions to meet their nutrient requirements. It was believed early in 
 the experiment that after continued self-feeding the cows might grad- 
 ually accustom themselves to the new conditions and consume only 
 enough to meet their needs, but the extension of the experimental peri- 
 ods to 78 weeks or more failed to show that such habits were formed. 
 
 Altho feed consumption was upon a relatively high plane during 
 the greater part of the lactation period, it fell below requirements at 
 the time of parturition in the case of several of the cows (Figs. 2 to 
 6). This condition usually was reversed in four to six weeks. The 
 cows used in these experiments were of only moderate dairy ability; 
 it is possible that if cows of exceptional dairy ability were self-fed, 
 they might consume feed about on a level with their requirements 
 thruout considerable portions of their lactation periods. Plans have 
 been made to secure data upon this point. 
 
 In but few cases did the consumption of net energy and digestible 
 true protein closely approximate the calculated amounts needed (Table 
 1). Most cows not only consumed more feed than they needed, but 
 their selection of the various nutrients was in most cases inconsistent. 
 Usually the excess of net energy above the theoretical requirements 
 was either much greater or much less than the excess of digestible true 
 protein. 
 
 The discrepancy between the calculated protein requirements of 
 Cow 576 and the amounts this animal actually consumed is of interest. 
 From the 29th to the 58th week inclusive the requirement curve if 
 
1927] EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 441 
 
 above the consumption curve, and yet the cow continued to increase in 
 weight and to produce milk at quite a constant rate up to the 55th 
 week. It is thus very improbable that she was drawing upon her body 
 tissues for a supply of protein. Judged by this one case, the Armsby 
 standard seems to prescribe a somewhat too liberal amount of protein, 
 altho the apparent deficiency may be accounted for in part by the 
 method of calculation used. Feeds vary in composition, particularly 
 in dry-matter content, from season to season and during storage. This 
 is true of corn, oats, and corn silage, which formed a large proportion 
 of the feed of this cow. 
 
 INDIVIDUAL PREFERENCES FOR FEEDS 
 CHANGE SUDDENLY 
 
 It was interesting to note the decided preferences of cows for cer- 
 tain feeds and the way in which these preferences often changed sud- 
 denly. Figs. 8 to 12 and 14 to 16, which show the variations in feed 
 intake from week to week, bring out the marked preferences for cer- 
 tain feeds. Figs. 17, 18, and 19 show the relative consumption of 
 feeds upon a percentage basis. The records of each cow have been di- 
 vided into consecutive periods of 13 weeks (three months) each. Al- 
 tho the preferences of a cow often changed decidedly during three 
 months, it is believed that a period of this duration is none too long 
 to ascertain whether or not certain feeds would be consumed at all, for 
 in the case of some cows a feed might not be touched during the first 
 two or three weeks or longer and later be consumed freely. In these 
 comparisons of the consumption of various feeds, the term palatability 
 has been used to indicate the inclination of the cows to eat a certain 
 feed, whether impelled to do so by the taste of the feed or by some 
 body craving. 
 
 With the primary object of studying the relative palatability of 
 various high-protein concentrates, Cow 267 was fed 20 pounds of corn 
 silage daily as the only roughage, while the concentrates were fed ad 
 libitum. In Fig. 17 the record for the first 15 weeks has been omitted 
 because of several changes in the feeds offered. Linseed oil meal was 
 the only concentrate consumed during all five periods. About 4 pounds 
 of distillers' dried grains were consumed during Period 1 and less than 
 12 pounds of soybeans during Periods 1 and 2. These two feeds, there- 
 fore, were not offered during subsequent periods. Soybean oil meal 
 (oil-expeller process) was consumed freely during Period 1. New 
 process soybean oil meal was introduced during Period 2 and was con- 
 sumed in larger amounts than the other form. The consumption of 
 cottonseed meal did not exceed 15 pounds in any period. 
 
442 
 
 BULLETIN No. 289 
 
 [May, 
 
 
 5 i i 
 
 
 
 
 *< o * 
 
 
 
 
 
 g| || -g 
 
 
 
 S3 
 
 -rt "** .^0) "3 
 
 
 
 a 
 
 2^3 ^"S ^T 
 
 
 
 S J 
 
 ^ M 
 
 M 
 
 ^ ^ G S . w -_, g 
 
 i i ^ it ill i 
 
 
 
 3 
 
 5? fe *~^ 3 ' ^ "t^ *S ^* 
 O O aj S O >r ^2 <S _Q aj 
 t^- 1-7 r^-^ 5l''**2l r^ 
 
 ^ <; cj <;M < o 
 
 
 
 -d 
 
 
 
 
 g 
 
 
 
 
 | -^ 
 
 
 
 
 ^ 
 
 (H hfl., , 
 
 . ^5 I s * ^O ^* CO 
 
 x> O i i * 00 X ^ 
 
 
 
 M ! ' o 
 
 jO ^H f-H 
 
 
 
 .5 P* 
 
 
 
 
 O 
 d 
 
 5'* 5 
 
 fll ^n5 W 
 
 (^ QJ 
 
 t* ^ CD CO ^ 00 O5 
 (MCOTf |> ^H^ngg 
 
 ^^ ^7 *!7 
 
 o> 
 
 o 
 
 K 
 
 
 w o 
 
 
 
 
 M 
 
 ** C -u 
 
 
 
 
 5 *?* ^ 
 
 !> ^H ^* ^^ QO ^^ 
 
 ^4 
 
 
 ij 2 ^ 
 
 fcO *O I-H t^. ^-t 10 
 
 co 
 
 
 CO -** C *-* 
 
 i-H i-H CO 
 
 
 
 M 
 
 o < 
 
 1 
 
 <N b- 
 
 ojCOOCO O ^ CM < 1 
 
 00 
 
 I-H 
 
 
 bC'j" ^_ 
 
 gi H-Hi35 co , ! , | CO 
 
 11 
 
 
 ^5 1) O 
 
 .COCO^ ^ Tt<C<IC<l 
 
 TO 
 
 
 
 j 
 
 
 
 
 i 
 
 
 
 . 
 
 i i J 
 
 a 
 
 
 
 '53 
 
 '_5j 
 
 
 
 M 
 
 ^s 
 
 
 
 o 
 
 o 
 
 
 
 : : B 
 
 B 
 
 
 
 i 
 
 
 
 T3 
 B 
 
 g 
 
 r j i .s . I 
 
 .S ;S i .9 1 J 1 
 
 -2 3 3 >> "0 r^ 
 
 J2 J2 O <iJ B 
 o o _. o -d 
 B B | B | | 
 
 red Guernsey 
 
 ! unknown. 
 
 
 "d 'd O) T3 m <u OQ 
 
 W C3 * CQ tn tH O 
 
 C 3 C 9 9 C 
 
 O O OH O (S fi 
 
 "oa 
 
 
 i 
 
 
 
 
 R 
 
 
 
 
 I 
 
 ^ 
 
 
 
 
 "fl 
 
 
 
 
 
 c 
 
 O 
 
 
 
 a 
 
 
 
 
 S 
 
 
 >O l> f CO 'J' O * 
 
 d 
 
 ** 
 
 
 C^ CM CO t* t^*' Oi ^^ 
 CM lO CM C 1 ! *O 
 
 
 
1927] 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 443 
 
 A striking illustration of the changes in a cow's preference for a 
 certain feed is afforded in the consumption of wheat bran. During 
 Period 1 Cow 267 ate less bran than any other feed, so small an 
 amount (1.2 pounds) that it was not offered during Period 2. It was 
 again offered in Period 3, with the result that 520 pounds were eaten 
 in 9 weeks. In Period 4 it made up 98 percent of the concentrates 
 
 FIG. 13. MOST SELF-FED Cows INCREASE RAPIDLY 
 IN WEIGHT 
 
 This illustration shows the high condition of flesh of 
 Cow 27 after being self-fed for about five months. All 
 of the cows used in the experiments gained rapidly in 
 weight and with the exception of Cow 25, remained in 
 a higher condition of flesh than when hand-fed. Cow 
 25 was quite thin for several weeks after calving. 
 
 eaten and in Period 5, 99 percent. During Period 4 its consumption 
 was 677 times as great as that of linseed oil meal, while in Period 1 
 151 times as much linseed oil meal was eaten as wheat bran. These 
 variations in the preferences of Cow 267 for certain feeds are fairly 
 representative of all the cows used, altho in some cases the variations 
 were not so marked. 
 
 Cow 27 was offered high-protein feeds for 104 weeks. Some 
 changes in feeds during Period 6, however, make it desirable to use 
 only the first five 13-week periods for a study of palatability. Soy- 
 bean hay, gluten meal, peanuts with hulls removed, peanut oil meal, 
 and flaxseed meal were the feeds offered. 
 
 The feed record of this cow is characterized by a liberal and 
 steady consumption of soybean hay, by a relatively small or zero con- 
 
444 
 
 BULLETIN No. 289 
 
 [May, 
 
 sumption of peanuts and flaxseed meal, by a liberal but rapidly declin- 
 ing consumption of corn gluten meal, and by a sudden and strong lik- 
 ing for peanut oil meal. The consumption of peanut oil meal was prac- 
 tically negligible until the 34th week, when it suddenly mounted to 
 about 10 pounds daily, remaining at a level of 8 or 9 pounds daily 
 until Period 4, when it rose to a maximum of 23 pounds daily. The 
 consumption of this concentrate was actually greatest during Period 
 4, but relatively greatest during Period 5 on account of a decline in 
 the consumption of other feeds. 
 
 FIG. 14. THE FEED PREFERENCES OF Cow 267 CHANGED FREQUENTLY AND 
 WERE VERY DECIDED 
 
 Wheat bran was offered for 21 weeks during the early part of the experiment, 
 but practically none was consumed. It was offered again during the 46th week 
 with the result that it soon was consumed very freely to the exclusion of all other 
 concentrates offered. 
 
 Since Cow 25 ate alfalfa hay only and entirely ignored soybean oil 
 meal (oil-expeller process), soybean oil meal (N.P.), cottonseed meal, 
 and linseed oil meal (O.P.), it is impossible to draw conclusions regard- 
 ing the relative palatability of these feeds except to state that in this 
 case alfalfa hay was evidently far superior to the concentrates. This 
 cow had been used in an earlier feeding trial in which soybean oil meal 
 was compared with cottonseed meal. During the three-week period 
 preliminary to offering the feeds in separate compartments, she was 
 fed 70 pounds of alfalfa hay and 140 pounds of corn silage weekly. In 
 addition she received a concentrate mixture consisting of ground corn, 
 ground oats, and wheat bran (equal parts by weight). Of this she 
 consumed 132.5 pounds the first week, 131 pounds the second week, 
 and 126 pounds the third week. She was accustomed, therefore, to eat- 
 ing concentrates in quantity and had previously consumed readily 
 most, if not all, of the kinds of feeds offered. The concentrates when 
 offered separately did not seem so palatable as when mixed. It is 
 
EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 445 
 
 likely that this cow would have eaten certain other concentrates had 
 they been offered. She calved during the first week of Period 3 and 
 shortly afterward developed a severe attack of diarrhea. In order to 
 remedy this condition, 53.5 pounds of scorched whole oats were fed. 
 The oats were eaten readily and proved effective. 
 
 ,1540 
 
 FIG. 15. SUDDEN CHANGES IN FEED PREFERENCES WERE CHARACTERISTIC 
 
 OF Cow 274 
 
 On two different occasions the amount of ground corn consumed dropped 
 suddenly from quite a liberal amount, once over 70 pounds weekly, to zero. From 
 the 10th to 13th weeks no corn at all was eaten, and then the consumption rose 
 rapidly to about 75 pounds weekly. When the consumption of corn fell off, 
 ground oats and wheat bran were eaten more freely. 
 
 The very remarkable preference of Cow 25 for alfalfa hay to the 
 exclusion of the other feeds extended over a period of 60 weeks. This 
 cow, weighing from 1,000 to 1,100 pounds, after giving birth to a 100- 
 pound calf during the 27th week of the experiment, consumed as much 
 as 341 pounds of hay weekly and yielded as much as 242 pounds of 
 milk (F.C.M.) weekly. 
 
 It is clear from this investigation that variations in the preferences 
 of cows for certain feeds may occur suddenly and extend over several 
 weeks or months. Just how much bearing these results have on gen- 
 eral practices in feeding for economical milk production, it is impos- 
 sible to say, but it seems likely that in feeding cows during tests for 
 maximum milk production a greater feed consumption could be in- 
 duced by catering carefully to the cows' preferences. 
 
446 BULLETIN No. 289 [May, 
 
 SELF-FEEDING APPARENTLY DOES NOT STIM- 
 ULATE MILK PRODUCTION 
 
 The few milk production records secured in these tests do not of- 
 fer great promise for the self-feeder method as a means of increasing 
 production. It must be remembered, however, that records were ob- 
 tained only with cows of moderate productivity. 
 
 All but three weeks of the third and fourth lactations of Cow 267 
 are shown in Fig. 14. If the milk production (F.C.M.) of the second 
 
 1360 A 
 
 FIG. 16. GROUND CORN PROVED MOST PALATABLE TO Cow 290 
 
 The feed record of Cow 290 was typical of cows self-fed on 
 a variety of high- and low-protein feeds in that marked changes 
 in the preferences occurred and that corn was consumed most 
 freely of all the concentrates. This cow was hand-fed from the 
 12th to the 16th week inclusive. In spite of the marked re- 
 duction in the amount of concentrates fed, milk production was 
 not lowered. 
 
 lactation, which was completed about seven weeks prior to the begin- 
 ning of the self-feeder trial, be taken as 100, then the productions of 
 the third and fourth lactations have the relations of 69 and 31 re- 
 spectively. The decline in yield cannot be attributed to insufficient 
 feed, since the amounts of feed consumed were large enough to sup- 
 port a greater milk yield (Fig. 5.). 
 
 Cow 576 showed an increased yield under the self-feeding method, 
 completing a lactation period of 448 days with a yield of 11,118 pounds 
 
19271 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 447 
 
 of milk (F.C.M.). The yield during the first 321 days of this period 
 was 8,923 pounds. Her second lactation period of 321 days, completed 
 about ten weeks before the beginning of self-feeding, had shown a 
 yield of 7,580 pounds of milk. The maximum yield during any week 
 of the third lactation was, however, only 5 percent greater and for the 
 
 FIG. 17. SELF-FEEDING SHOWS THAT PREFERENCES FOR CERTAIN FEEDS MAY 
 CHANGE FREQUENTLY 
 
 This chart shows the relative 
 amounts of the concentrates con- 
 sumed during consecutive peri- 
 ods of 13 weeks each. The total 
 intake of concentrates during 
 each period is considered as 100 
 percent, and the amount of each 
 as a percentage of the whole. The 
 roughage, which consisted of 20 
 pounds of corn silage daily, has 
 not been considered, since it was 
 hand-fed. 
 
 During Period 1 Cow 267 al- 
 most ignored wheat bran, con- 
 suming only 12 pounds, or less 
 than 1 percent of the total con- 
 centrates, while during Periods 4 
 and 5 she ate practically no con- 
 centrates except wheat bran. Un- 
 doubtedly the freshness of the 
 feed, its freedom from mold and 
 musty odor, and the degree to 
 which the feed possesses a pleas- 
 ant aroma or taste, account to a 
 large extent for the choice of 
 feeds which a cow makes, but ev- 
 idently they do not explain all 
 
 the results found. In this case, cracked soybeans did not prove palatable, but the 
 soybean oil meal (oil-expeller process), which possessed a slightly roasted flavor 
 and odor, was eaten freely. During Period 2 soybean oil meal, manufactured 
 only a few days before delivery by a chemical solvent process, was offered and 
 was consumed in larger amounts than any other concentrate. So far as known, 
 the grade and quality of the wheat bran offered during Period 1 were the same 
 as those of the wheat bran consumed during Periods 3, 4, and 5. 
 
 100 
 
 90 
 
 eo 
 
 70 
 60 
 
 8 
 
 K 
 40 
 
 30 
 20 
 10 
 O 
 
 RELATIVE c 
 rttps BY 
 
 QM30MPTION OF 
 COW 2.67 
 
 
 
 A- UN5LEP OL MEAL, O.P. 
 B- WHEAT BRAN 
 C- SOYBEAN OIL MEAL,O.E-P 
 P- SOYBEAN OIL MEAL. N.P 
 
 
 
 E.-COTTON3E.E.P ME.AL 
 F - CRACKE.P SOYBEANS 
 
 
 
 
 - 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V\ 
 
 
 
 
 
 1 
 
 
 i 
 
 
 
 
 
 I 
 
 
 
 
 ^ 
 A 
 
 ' 
 
 B 
 
 c 
 
 i 
 
 I 
 
 r 
 
 - 
 
 A 
 
 C 
 
 bt < 
 
 ' A 
 
 1 
 
 CPE A 
 
 ~ 
 
 C PL AE 
 
 CPE. 
 
 
 .2 
 
 = 
 
 1 
 
 * 
 
 
 | 
 
 i 
 
 ^ 
 
 J 
 
 A & A |J 
 
 
 U3 
 
 CONSECUTIVE. PtRIOPS OF IS WLEKS EACH 
 
 first 321 days only 18 percent greater than during the previous lacta- 
 .tion. The self-feeding trial extended over the fourth lactation period 
 also. The production of milk (F.C.M.) during this lactation, which 
 was 304 days in length and was terminated by an abortion, was 9,454 
 pounds, or about 20 percent greater than that of the second lactation. 
 
448 
 
 BULLETIN No. 289 
 
 [May, 
 
 Cow 25 showed a decrease in yield under self-feeding, giving 288 
 pounds of milk (F.C.M.) during the week of highest production in her 
 first lactation and 242 pounds while self-fed, or a ratio of 100 to 
 84. The maximum actual yield during one week of self-feeding when 
 alfalfa hay was the sole feed consumed was 274.6 pounds containing 
 3.2 percent fat. 
 
 CON3LCUT1VE. PE.RIOP3 Or 15 WttKS tACM 
 
 FIG. 18. THE PREFERENCES OF SELF-FED Cows MAY REMAIN QUITE CONSTANT FOR 
 MANY WEEKS AND THEN CHANGE DECIDEDLY 
 
 Both roughages and concentrates are included here. The amount of silage 
 consumed was multiplied by the factor .3 in order to reduce it to a dry-matter 
 basis which would be comparable to that of the other feeds. 
 
 During Period 1 corn formed about three-fifths of all the feed eaten, but dur- 
 ing succeeding periods it became a much less important part of the ration. Dur- 
 ing Period 2 about 300 pounds more of corn were consumed than during Period 
 1, but on account of the increased consumption of other feeds during this period, 
 corn formed a smaller proportion of the whole. The corn fed up to the middle of 
 Period 10 graded No. 2 yellow. During Periods 7 to 9 it was more than one year 
 old. It is not thought that the age of the corn was a factor, since it seemed to 
 be in excellent condition. About the middle of Period 10 No. 3 yellow corn was 
 fed. Even tho the silage appeared to be of poorer quality during Period 10, 
 more was consumed than in any previous period. 
 
 Aside from the two cases noted, no other changes were observed in the qual- 
 ity of the feeds which might have a bearing upon the amounts consumed. 
 
 The maximum weekly yield by Cow 27 during the lactation period 
 which began in the 34th week of the self -feeding experiment, was 167 
 percent of the maximum during the preceding lactation period, which 
 was her first, and during the first 18 weeks of which she was hand-fed. 
 The decline in milk flow, however, was very rapid during both lacta- 
 tions (Fig. 11). The milk yield during the first lactation period of 43 
 weeks was 6,042 pounds (F.C.M. ), while during the first 43 weeks of 
 the second lactation it was 6,982 pounds (F.C.M.), or 116 percent of 
 
19271 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 449 
 
 the first. A part of the increased yield during the second lactation, as 
 compared with the first, probably is to be accounted for by the natural 
 increase in productivity which accompanies the development of young 
 cows up to maturity. 
 
 In this investigation the free-choice method of feeding, which in- 
 duced a consumption of nutrients greatly in excess of the requirements 
 
 FIG. 19. THE RELATIVE AMOUNTS OF CERTAIN FEEDS CONSUMED IN DIFFERENT 
 PERIODS BY Cow 27 CHANGED GREATLY 
 
 The ration of Cow 27 was 
 made up largely of soybean hay 
 and two protein concentrates. 
 During Period 1 soybean hay 
 comprised about two-thirds of 
 the total feed, declining to about 
 one-third in Period 5. The con- 
 centrates consumed during Peri- 
 ods 1 and 2 consisted almost en- 
 tirely of gluten meal, but begin- 
 ning with Period 3 this was grad- 
 ually replaced by peanut oil 
 meal. So appetizing did the pea- 
 nut oil meal prove that it formed 
 over half the total feed and 
 seven-eighths of the concentrates 
 consumed in Period 5. In the 
 peanut oil meal used, differences 
 were noted in the quality of var- 
 ious lots which might possibly account for the negligible consumption in Periods 
 1 and 2, followed by a sudden demand for it. 
 
 CONSECUTIVE. PtRIOPS OF 
 
 WE.E.K3 E.ACM 
 
 of the animals, failed to stimulate milk production to any higher 
 point than the usual methods of feeding, and further, the limitation of 
 cows to high-protein feeds seemed to have a distinctly depressing in- 
 fluence upon milk secretion. 
 
 COWS CAN BE FED SAFELY BY THE SELF-FEEDER 
 
 METHOD 
 
 So far as could be observed, no harmful effects upon the animals 
 resulted from these self-feeding experiments. As indicated in the plan 
 of procedure, no changes were made in the method of feeding on ac- 
 count of gestation or parturition. It is true that one calf was born dead 
 and another died a short time after birth, while a third died at the 
 age of about two weeks, but the last fatality is believed to have been 
 due to the condition of the milk from the dam, which suffered a gen- 
 eral toxemia as a result of a retained afterbirth. The calf had ap- 
 
450 BULLETIN No. 289 [May, 
 
 peared normal until a short time before its death. There is no evi- 
 dence indicating that the other reproductive failures were or were not 
 related to the experimental routine to which the dams were subjected. 
 In the case of Cow 576, it is only known that abortion was prevalent 
 in the herd of which she formed a part. 
 
 Cow 274 was "off feed" a number of times when she evidently had 
 eaten too much corn, but this did not seem to affect greatly her milk 
 production or general health. 
 
 A severe case of diarrhea was developed by Cow 25 shortly after 
 calving, but this condition was readily corrected. This animal was re- 
 moved from the experiment at the close of the 60th week on account 
 of a positive reaction to the tuberculin test. A post-mortem examina- 
 tion revealed only a very small lesion. 
 
 In general, the health of the cows employed in this investigation 
 seemed to be of the best. It is possible, however, that continuation of 
 the experimental conditions over a longer period may reveal pro- 
 nounced effects upon the health, particularly in connection with ani- 
 mals subsisting upon high-protein feeds. This phase of the study is 
 being continued. 
 
 SELF-FED AND NORMAL RATIONS DIFFICULT TO 
 COMPARE IN DIGESTIBILITY 
 
 ; ' The consumption of large amounts of feed by self-fed cows nat- 
 urally raises a question regarding the relative digestibility of such ra- 
 tions compared with normal rations. A digestion trial was carried out 
 with Cow 274 in an effort to determine this point. 
 
 The trial consisted of two ten-day periods, during the first of 
 which the cow was self-fed, while during the second she was fed closely 
 in accordance with her estimated requirements. It was not considered 
 feasible to provide all the feeds separately when the cow was confined 
 in a metabolism stall in which the trial was conducted, so a concen- 
 trate mixture was made up in the proportion in which the feeds had 
 been consumed previously. The mixture consisted of linseed oilmeal, 
 7 percent; wheat bran, 25 percent; ground oats, 31 percent, and ground 
 corn', 37 percent. Alfalfa hay and corn silage were offered separately. 
 Each ten-day period was subdivided into five-day periods to furnish a 
 means of checking the determinations. Altho the trial was carried out 
 when the cow was in her third month of lactation and producing about 
 32 pounds of milk daily, she unfortunately did not consume as much 
 of the concentrate mixture as she had previously eaten when the feeds 
 were offered separately. This greatly reduced the value of the com- 
 parison. 
 
1927} 
 
 EXPERIMENTS IN THE SELF-FEEDING OF DAIRY Cows 
 
 451 
 
 The amounts of the feeds consumed during each five-day period, 
 and the coefficients found for the rations made up in these proportions, 
 are shown in Table 3. The differences in the coefficients for the self- 
 feeder and the normal ration periods were practically negligible, prob- 
 ably owing to the fact that the difference in feed intake in the two 
 periods (less than 2 kilograms of dry substance daily) was so small. 
 
 TABLE 3. COMPARISON OF DIGESTIBILITY OF SELF-FED AND HAND-FED RATIONS 
 
 Amount of 
 orts for 
 
 feed and 
 period 
 
 Coefficients of digestibility 
 
 Dry 
 
 substance 
 
 Crude 
 protein 
 
 Crude 
 fiber 
 
 N-free 
 extract 
 
 Ether 
 extract 
 
 Period la Self -fed five days 
 
 Silage 
 
 grams 
 24,050 
 
 68 
 
 70 
 
 48 
 
 75 
 
 67 
 
 Concentrates 
 Hay 
 
 16,978 
 
 31,326 
 
 Orts 1 
 
 .... 7,131 
 
 
 Period Ib Self -fed five days 
 
 Silage 
 
 20,000 
 
 66 
 
 68 
 
 45 
 
 75 
 
 66 
 
 Concentrates 
 Hay 
 
 18,169 
 
 43,696 
 
 Orts 1 
 
 4,108 
 
 
 Period 2a Hand-fed five days 
 
 Silage .... 
 
 35.000 
 
 68 
 
 70 
 
 44 
 
 77 
 
 78 
 
 Concentrates 20,000 
 Hay 20.000 
 
 Orts 
 
 
 
 
 Period 2b Hand-fed five days 
 
 Silage 
 
 35,000 
 
 66 
 
 69 
 
 39 
 
 78 
 
 71 
 
 Concentrates 
 Hay.. 
 
 20,000 
 . . 20,000 
 
 Orts 
 
 
 
 'Chiefly silage. 
 
 CONCLUSIONS 
 
 The self-feeding of cows is not economical from the standpoint of 
 the amounts of feed consumed, for much more feed than enough to 
 meet their requirements for maintenance and milk production is eaten. 
 Cows tend to lay on additional weight, thus increasing the maintenance 
 cost. 
 
 The milk yields of cows of moderate productivity are not in- 
 creased appreciably by self-feeding. Limiting cows to high-protein 
 feeds only seems to depress milk production. 
 
 Cows which are self-fed usually consume feeds supplying an ex- 
 cess of net energy or digestible true protein or both. For a short period 
 
452 BULLETIN No. 289 
 
 following calving, however, the amounts of feed consumed may be too 
 small to meet the requirements. 
 
 Feeding dairy cows by the self-feeder or free-choice method 
 is not harmful to their health, provided they are gradually accustomed 
 to the method. 
 
 The method is useful in studying the relative palatability of dif- 
 ferent feeds. Corn is the most palatable of the low-protein concen- 
 trates used. 
 
 The preferences of self-fed cows for particular feeds may change 
 suddenly, and these preferences may continue over long periods. Cows 
 differ widely in their individual preferences. It is likely that greater 
 feed consumption can be secured when one studies these preferences 
 carefully and caters to them. While this may be of some value under 
 the usual conditions of good dairy practice, its greatest application is 
 in feeding cows for records of maximum milk production. 
 
 LITERATURE CITED 
 
 1. HULCE, R. S. The preservation and cost of raising dairy heifer calves. Nebr. 
 
 Dairymen's Assoc. Proc. 1917. 
 
 2. McCANDLiSK, A. C. Appetite as a guide in feeding dairy calves. Iowa Agr. 
 
 Exp. Sta. Res. Bui. 51. 1919. 
 
 3. NEVENS, W. B. The self feeder for dairy calves. Jour. Dairy Sci. 2, 435. 1919. 
 
 4. HUNT, R. C. Self feeders for dairy cows. Va. Agr. Exp. Sta. Rpt. 1917-1918. 
 
 5. HOLLER, L. S. Self feeding dairy cows. Thesis for the degree of B. S. in ag- 
 
 riculture, University of Illinois. 1922. 
 
 6. DAVENPORT, E. Principles of breeding. Appendix section 4 (Statistical meth- 
 
 ods) by H. L. Rietz. Ginn & Company, 1907. 
 
 7. BORLAND, A. A. Computation of dairy rations. Penn. Agr. Exp. Sta. Bui. 161. 
 
 1920. 
 
 8. GAINES, W. L., and DAVIDSON, F. A. Relation between percentage fat content 
 
 and yield of milk. 111. Agr. Exp. Sta. Bui. 245. 1923. 
 
 9. HOGAN, A. G., et al. The influence of the plane of nutrition on the mainten- 
 
 ance requirement of cattle. Mo. Agr. Exp. Sta. Res. Bui. 51. 1922. 
 10. McCANDLiSH, A. C., and GAESSLER, W. G. Influence of condition on main- 
 tenance requirements of dairy cattle. Iowa Agr. Exp. Sta. Res. Bui. 60. 
 1920. 
 
UNIVERSITY OF ILLINOIS URBANA