FEEDING TESTS AND THEIR METHODS

This article is summarized at its close some forty pages further on in this volume. The results are concisely stated therein and cross references given.

I. INTRODUCTION

The general line of feeding tests with cows begun in the winter of 1896-97 has been followed in the work of the past season. Somewhat more stress than heretofore, however, has been laid upon the comparison of the relative food values of different rations. The equipment of the station is in many ways well adapted to experimentation in this line. The large size of its herd in particular permits wide range in choice of animals, much repetition of experiments and the conduct of a relatively large number of trials.

The feeding experiments of the past winter were designed to increase our information upon several points. Some of the tests were in continuation of similar trials in previous years (a, b, c, d, e) while the others pertained to matters not hitherto considered at this station. The questions which the trials were designed to aid in answering were as follows:

(a) What variations in production are to be expected from different rations-i. e., those made up to some extent of different materials-each containing essentially the same amounts of the various nutrients?

(b) What effect upon production has the addition to the ration of liquid fat, emulsified or unemulsified?

(c) What variations in production are to be expected from different rations, each containing essentially the same amounts of digestible dry matter, but different amounts of various nutrients, or, in other words, "medium" and "wide" rations?

(d) What variations in production are to be expected from different rations, each containing different amounts of digestible dry matter and of all digestible nutrients, or, in other words, "medium" rations fed in medium quantity and "wide" rations fed in relatively scant quantity?

(e) What is the feeding value of buckwheat middlings compared with half and half cottonseed and linseed meals? also compared with corn and bran?

66

(f) What are the comparative feeding values of corn silage and 'Improved french white artichokes?"

(g) What is the effect upon production if cows may drink water as they will instead of but twice daily?

(h) What effect has grooming cows upon production?

(i) What is the extent of the probable error inherent in feeding experiments due to the individuality of the animals used or to other causes? The feeding trials were known respectively as tests of (a) “equal balance," (b) oil feeding, (c) medium and wide rations, (d) medium and wide

scant rations, (e) buckwheat middlings, (f) artichokes, (g) watering, (h) grooming and (i) experimental error; and are so referred to throughout this article and in the appendix of tables.

II.

STATEMENT OF METHODS AND DETAILS OF CONDUCT OF

TESTS

The records of 60 cows were kept in the seven and one-half months during which the feeding trials lasted. Those of 4 cows fed for seventeen weeks during the winter and spring of 1897-98, and not hitherto reported are likewise available. The records of 8 of the original 60 cows for one reason and another are unsafe to use. Errors in feeding and shortage of supply owing to incorrect estimates seriously curtailed the data concerning the feeding value of artichoke tubers. A few cows scheduled and used in sundry experiments either began to dry off early or became ill or went off feed during the winter. It is to be expected that some troubles of this kind would follow the use of so large a number of animals fed under conditions necessitating frequent and radical changes in rations.

The 56 cows whose records are available were distributed among the tests as follows: (a) equal balance, 12 cows used in two series of trials with different rations; (b) oil feeding, 14 cows used in four series of trials, two in 1898 with 4 cows, three in 1899 with 10 cows; (c) medium and wide rations, 5 cows; (d) medium and wide scant rations, 5 cows: (e) buckwheat middlings. 6 cows; (f) artichokes, 1 cow; (g) watering, 4 cows; (h) grooming, 5 cows; (i) experimental error, 4 cows.

All the cows used were not equally well suited to our purposes as regards time of lactation. A long and careful study of previous records—extending over nearly five years with some animals-of ages, time of calving, flow and quality of milk, time of service, etc., prefaced choice for the various experiments. The equal balance and oil feeding tests on the whole had preference in selection.

DETAILS OF FEEDING

The feeding periods were either 4 or 5 weeks long.1 Regardless of the length of period the first third of the time was always considered preliminary and non-experimental. Hence the periods divide into preliminary and experimental portions as follows: 4 weeks into 10 and 18 days, 5 weeks into 12 and 23 days. The experimental portions only of the periods are considered in the discussion, and are hereinafter referred to as "the period." Results obtained and conclusions drawn are based upon them alone. The preliminary portions were considered as preparatory only, being necessary to get the animal fairly upon and accustomed to its new

I The first oil feeding period was four weeks and the second non-oil feeding period six weeks long.

diet. Notwithstanding this fact, full feeding and milk records were kept, samples of fodders and feeds taken, and analyses made in the same manner and with the same care during the preliminary as during the experimental portions. In some cases the milk was sampled in the preliminary as well as in the experimental portions of the periods.

The cows were fed twice daily, watered twice, and turned out into the yard twice. The yard exercise varied in duration from 20 to 45 minutes according to weather. Hay and grain were first fed morning and night, and after these were eaten as much silage was given as the cow would consume. The amount of hay offered each cow was uniform throughout the tests, but varied between cows according to their digestive capabilities. All fodders and feeds were weighed as fed, the orts were weighed back daily and their nature noted.

The following statement shows the general feeding plan :

The formulas, analyses, digestible constituents, etc., of the mixed feeds numbered 1 to 5, the grains from which they were formed, as well as of the roughages used which are referred to in the statement, are shown on page 260. It may be said in brief that Nos. 1 and 2 contained bran, cornmeal, cottonseed and linseed meals, No. 3 consisted of bran, cornmeal and buck wheat middlings, while Nos. 4 and 5 contained corn and bran. Mixed feed No. 1 did not differ materially in composition from No. 2.

FEEDING PLAN OF ALL TESTS

All the cows received hay in amount ranging from 12 to 24 pounds daily, about two-thirds of them ate from 12 to 30 pounds of silage daily, and a few ate 10 pounds of sugar beets each day. Artichokes were fed to several cows.

The grain feeds used in alternating periods were as follows: Equal balance: (a.) 7 pounds No. 2, 8 pounds Buffalo gluten feed; (b.) 8 pounds each Quaker oat feed and equal parts corn and bran.

Oil feeding, 1898: Emulsified cottonseed oil mixed with 8 pounds of bran, and 8 pounds bran alone; emulsified cottonseed oil mixed with 8 pounds bran and 8 pounds bran alone.

Oil feeding, 1899, (a.) Emulsified cottonseed oil mixed with 8 pounds equal parts of corn and bran, and 8 pounds equal parts corn and bran (No. 4); (b.) emulsified corn oil mixed with 8 pounds equal parts of corn and bran and 8 pounds equal parts corn and bran; emulsified linseed oil mixed with 8 pounds equal parts of corn and bran and 8 pounds equal parts of corn and bran.

Medium and wide.-8 pounds each No. 1, and Quaker oat feed.

Medium and wide-scant.—8 pounds Buffalo gluten feed and 3 pounds

corn and bran (No. 5).

Buckwheat middlings.-8 pounds each of mixed feeds 1 and 3; 8 pounds each of mixed feeds 3 and 4; 8 pounds each clear buckwheat middlings and 4.

Artichokes.—Mixed feed No. 1 throughout; silage and artichokes, alter

nating.

ting.

Watering.-Mixed feed No. 4, thoughout; methods of watering alterna

Grooming.-Mixed feed No. 1, throughout; grooming and non-grooming alternating.

Experimental error.-Mixed feeds Nos. 4, or 2 or Buffalo gluten feed throughout; no alternation.

Comparisons of the amounts of nutrients eaten by the various cows with the Wolff and the Wolff-Lehmann standards show that while there were wide variations in eating as between the different cows and rations there was generally eaten enough and to spare of each of the ingredients except protein. In several, but not in all cases, too little protein was eaten. In most cases, however, the amount of this latter nutrient eaten approximated that called for by the Wolff standard except when oat feed or corn and bran was eaten, either in equal parts or as mixed feed No. 5. More specific statements touching the comparison with standards are made in the discussion of each test.

WEIGHTS OF COWS

All the cows under test were weighed during the first three days of the opening period, and on the last three days of all periods. Average weights are shown in table I of the appendix. About half the number gained quite decidedly in weight, about one-third did not vary materially in this respect and five cows (Acme, Inez, Flora, Jeannie and Clover) lost in weight. While variations in live weight are too uncertain things to lay much stress upon, yet a study of the table appears to show certain general trends in a few cases. On the whole no clear advantage of one ration over another in increasing live weight can be seen, except as noted, when either of the following were fed: Buffalo gluten feed, either cottonseed-linseed mixture, either corn and bran mixture, Quaker oat feed. The corn and bran ration had always a tendency to increase live weight in the oil feeding experiments. This may have been due in part to the considerable wasteage of the grain when mixed with oil. Whenever buckwheat middlings were fed live weight shrunk. The cows when drinking water at will weighed more as a rule than when drinking at intervals, probably because they had more water in their digestive systems at the time of weighing. There was a tendency towards greater weight when the cows were groomed than when not thus handled.

BARN TEMPERATURES

The cows were stabled in two portions of the barn structure, the main floor and the east wing, which leads from the main portion. The temperatures of the two stables were taken daily at 5 A. M., 12 M., and 6 P. M. The exposed location of the station barn causes decided fluctuations in the temperature of the cow stables during the winter, notwithstanding the comparatively large number of animals housed therein.

The average barn temperatures, morning, noon and night, during each period, the ranges of variation and the per cent of the entire number of observations within 3° F. or the mean of each period are tabulated in the appendix. Average variations of 5° or more from means for considerable periods occurred on about 20 days. For several years a careful study has been made in connection with feeding tests of the fluctuations in yields and of qualities of the milk occurring coincident with, immediately preceding, and immediately succeeding wide thermometric changes. No definite relation has been traced between variations in barn temperatures and in the quality of the milk. Some samples have shown more or less quality fluctuations inverse to temperature changes, others in the same direction, while in many cases no change occurred. It has been held safe to assert that these few changes, mostly slight in amount, had practically no influence in modifying period averages, each of which was based on analyses of several composite samples of milk.

Careful comparisons of variations in milk yield with temperature changes were made and five cases of possible relationship were noted, as follows:

Dec. 1-5, 31° above mean, 1 per cent. increase in milk flow.

These cases, which were much the largest fluctuations occurring coincident with temperature changes extending over more than a day or two, are so small as to be quite without effect upon the average of results obtained in periods of from 18 to 23 days in length. A more uniform barn temperature would have been an advantage; but it is judged from these comparisons that temperature variations had little if any effect upon the final results. It should be noted, moreover, that since all the cows were housed together, these effects, if any, might be expected to be uniformly exerted upon all. It should be remembered also, in this connection, that all the cows were turned out daily for from twenty to forty minutes.

The following tables show the period dates and the cows in use, with the nature of the grain feed eaten during each period. The names which are italicised are those of Ayrshires registered or registerable, the black faced letters indicate Jerseys registered or registerable, while the names given in ordinary type are those of high grade Jerseys.