VARIATIONS IN THE PERCENTAGE OF BUTTERFAT IN MILK.

New Zealand Journal of Agriculture, Volume XXX, Issue 5, 20 May 1925, Page 327

VARIATIONS IN THE PERCENTAGE OF BUTTERFAT IN MILK.

A STUDY BASED ON NEW ZEALAND C.O.R. DATA.

W. N. PATON,

Dairy Division, Wellington.

IV. YEARLY VARIATIONS.

Yearly variations in test are those which occur in the average tests for different lactation periods. The average lactational or annual test is obtained by dividing i do times the total of the butterfat by the total of the milk -produced during one lactation period. For the purposes of certificate-of-record testing a lactation period is limited to 365 days. ■. ’ ' ■ . It is not an easy matter for dairymen to ascertain what variations there are in the tests of cows from year to year, and to what factors they are - due. A cow is rarely tested each year, and, besides, the number of records available from the regular testing of one herd is hardly adequate on which to base any deductions. However, when one has large numbers of records to investigate, the nature of the variations and the causing factors are more readily apparent. Owing to the smallness of yearly test variations, considerable difficulty is experienced, even with fairly large numbers of records, in discovering the nature and assigning the causes of the many variations. A good clue to their discovery is afforded by the results of the study of monthly test variations, because of the. obvious connection . between the two. Hence we have here to consider the same set of factors as set out in the preceding article of this series. YEARLY VARIATIONS DUE TO BREED. This question was treated to some length in the first article of this seriespublished in the Journal for September, 1924 — which readers are referred. . ? YEARLY VARIATIONS DUE TO TIME OF YEAR OF CALVING. The average annual tests of the four chief breeds have been tabulated according to the season of the year in which the cows commenced test, and are given in Table 15. From a study of this table it is apparent that the figures for spring and autumn and summer and winter agree quite closely. For this reason they have been grouped as shown in the two right-hand columns of the table. With the exception of the Jerseys, the average tests indicate that cows commencing in the summer or winter months test higher than do those commencing in the spring or autumn. The annual tests for each month of commencement are given in' Table' 16. Owing to the blanks in the Jersey and Ayrshire figures for April and May, no average figures have been supplied 'or these months. Furthermore, the respective numbers of records for the months of December to June (both inclusive) are rather small, and for this reason no great weight is attachable to the. averages obtained from them. It is found that the averages for July, August, and September are slightly higher than the. average of the four breeds for

all months of commencement, which reads 4-33 per cent. The average for all breeds, grouping December, January, February, March, and June together, is 0-03 below the average for all months. The table is rather incomplete to be sufficiently reliable, so the question may be attacked from another angle. In order to obtain an indication of the months of the year during which cows test highest, lactational tests for the chief breeds have been summated and averaged for all months of the year. In other words,' this analysis gives the average test for any particular month according to the breed of cow, and irrespective of the month of commencement of test. The results are presented in Graph 11. A glance at this graph suffices to show that the variation in tests for different months is marked. The tests for June and July rank, highest in each case, May and August coming next. October is the lowest in each curve with the exception of the Friesians, in which case the. lowest point occurs for the month of December. September, October, and November are bound to be low, for the reason that grass feed is then at > its best . and most abundant. This keeps the quantity of milk - production up with the subsequent loss in test. This point is readily apparent in cases where cows are. completing a lactation period in the spring (see August to December curves in Graph. 7 of third article of series). The curves of the five breeds vary somewhat in conformation. However, the average figures of the five breeds given at the foot of the graph (No. 11) present the general position very clearly. .• All cows were not tested the full number of times, which may be either twelve or thirteen, and for this reason equal numbers are not represented for each month. In order to ascertain how this affected the results a number of Friesian records were selected, for each of which the annual test was approximately 3-55 per cent, (this being .the average annual test for. all Friesians), and for each there were twelve monthly tests. Six were selected for each month of commencement, and thus seventyrecords altogether are presented. r . The result is given in Graph 12. The curve for all Friesians is included .for purposes of comparison. The figures at the top and bottom of the graph refer respectively to the curves for the “ selected Friesians ” and for “ all Friesians." The selected Friesians give a curve considerably flattened out compared with that for all Friesians. The general trend, however, , remains practically the same, and the curve lends itself to the same division which is characteristic of the other. ‘ . The grouping of the results into periods is given in Table 18. For the selected Friesians the range of variation over the three groups has been considerably reduced from that for all Friesians, the reduction amounting to 100 per cent. It is very probable that equal reductions in the range of variation would be found in the results of standard data for the other breeds. At the same time it is expected that the trend of the curves as shown in Graph 11 would be confirmed. By adopting for each breed in Graph n similar groupings to those of Table 18, one is better able to understand the results given in Tables 15 and 16. . The groupings are not necessarily the same for each breed, and to facilitate this the figures for Graph 11 are given in Table 17. Flush

of milk-production varies somewhat with' the- breed, the general rule being that the heavier the breed the later the time. Thus Jersey cows, because they are lighter, reach flush of milk-production after calving earlier than cows of the other principal dairy breeds. Furthermore, the lactational tests for Jerseys rise continually, throughout the lactation period, this, as pointed out in the third article, not being the case for the other breeds (see Graph 4, page 78, Journal, February, 1925). For the other breeds the test decreases for a while and then rises steadily to the end. Lactational-test curves vary for different periods of the year of commencement, and so does flush of milk-production. As a rule flush of milk-production is fairly .well maintained for cows of good dairy qualities until about the sixth month. 'If these points are considered in conjunction with Table 17 one now sees the reason for the apparent anomalous result for the Jerseys in Table 15. . To sum . up, therefore, May, June, July, and August are the highest-testing months. Cows commencing so ' that they ; reach the flush of their milk-production during the above-mentioned - period will have higher annual tests, and will produce more butterfat, than for any other period of commencement. This conclusion agrees with that of an investigation made by the United States Dairy Division (see U.S.A. Bulletin 1071), in which it was found that autumn-freshened cows produced most butterfat, next in order being winter calvers, and spring and summer calvers lowest with equal amounts. In all 10,870 records were studied.

YEARLY VARIATIONS DUE TO LENGTH OF PERIOD OF GESTATION DURING TEST. In order to ascertain the effect of the length of the period of gestation on the annual test a random selection of available data has been taken and tabulated according to the varying periods between commencement and effective service during test. The results of this analysis are given in Table 19. Where cows are empty for less than eighty-three days during test they will calve within 365 days following calving at commencement (the gestation period is taken as 282 days). Such records are represented by the first groups given in the table.

The. other groups taken in order are for cows calving subsequent to test during the thirteenth, fourteenth, and fifteenth months respectively. The variations in test for this table are of a peculiar nature, and require further investigation. However, it appears clear that cows test highest for a period of about thirteen and a half months between consecutive calvings. If the period is more than this the annual test is lowered somewhat. The two- and four-year-old Friesians do not agree very well in variations for similar groups, but by combining the data the average tests for the groups reading -downwards in the table are then 3-50, 3-46, 3-52, and 3-51, and thus agree closely with the Jersey results.

By selecting Jersey two-year-olds data from the previous tabulation for which shorter, periods than 116 days occurred between calving at commencement and effective service during test, and grouping these records according to the varying number of days in milk, test variations were obtained as shown in Table 20. The results show that the shorter the lactation period the higher is the annual test.* The twenty records for the limits 356-365 days as given in this table were grouped into two, according as the “fallow” period was more or less than ninety-five days. The average annual test for the shorter period was 5-52, and that for the longer period 5-45. Thus after eliminating, the factor of length of lactation period it is still found that the factor of length of pregnancy causes variations in the annual test. . '

Referring back to Table 19,. it is now seen that for the group “ up to 85 days ” two factors are operating which tend to raise the test, whereas for the remaining groups there is only one. This perhaps explains the dip in the annual tests occurring between the first and third groups. The intermediate tests for the two-year-old Jerseys from 116 to 183 days for limits of ten days only are as follows : 6-09, 6-o6, 5-68, 5-67, 5-45, 5-12. Therefore it can be concluded that, with the exception of the high point indicated in the third group of Table 19, the shorter the fallow period during test the higher is the annual test, and the longer that period the lower the test, the latter being true up to a period of about seven and a half months, when no further lowering of tests takes place. This statement is well borne out in Graph 9, page 86, and text, page 87, Journal, February last. YEARLY VARIATIONS DUE TO NATURE OF SEASON. Variations in annual test due to the nature of the season, though very small, are nevertheless quite interesting. In Table 21 the averages' for the Friesian tests, amount of rainfall, and number of rainy days are given Season by season. Plus and minus variations are supplied in the adjoining columns for each of the three mentioned items, and for purposes of comparison the mean figures for the eleven seasons are given at the foot of the table. The average tests are estimated equivalent to maturity —that is, for each season the average milk and butterfat figures for the various classes according to age have been raised to what they would be if the records had been made at maturity, and the test obtained from this result. By this' means the factor of age is eliminated from the results. Correlation between amount of rainfall and yearly test variations seems difficult at first, as the table shows tests for six seasons varying inversely'with the rainfall, while for the five remaining the relation is direct. However, when both amount of rainfall and number of rainy days are considered with the test variations, one is able to elucidate some of the difficulties. The variations in the first and third righthand columns agree in sign for each season with but three exceptions—-1912-13, 1916-17, and 1919-20. For 1912-13 the South Island was above and the North Island below, the mean rainfall, while for season 1916-17 the contrary was the case. In 1919-20, however, the amount of rainfall for each Island was above and the number of rainy days below the mean for all seasons. In the light of this information it is not unreasonable to accept 1912-13 as conforming very closely to a season of average rainfall. The position is strengthened also in regard to season 1916-17. The number of rainy days for season 1919-20 is so much below the average that this season can be safely classed as a poor one for rainfall in general. Now that the positions in regard to these seasons have been somewhat reconciled, there are left only three other seasons (marked with an asterisk) where the test variations vary directly with the rainfall. When it is borne in mind that the percentage of cows tested in each Island varies from season to season, that the testing season somewhat overlaps twelve months, that the amount of testing is not represented equally or in all districts where rainfall-recording stations

are situated, and that the climate varies considerably throughout New Zealand, it is not at all surprising that a few of the results do not conform to the general rule. There appears, however, to be sufficient evidence to conclude that the small fluctuations which occur in the test from season to season are due to the rainfall and vary inversely with it, and that, with regard to rainfall, both amount of rainfall and number of rainy days must be given due consideration.

YEARLY VARIATIONS DUE TO CONDITION . OF COW, FEEDING, ETC. ' • As mentioned in the two preceding articles, there are no data available from Certificate-of-record testing with which to investigate for the effect of variations due to condition, feeding, &c., of the cow. It is considered that cows of approved dairy type commencing in good condition, and so maintained during test, would test higher than cows which commenced in poor condition or were poorly fed during the lactation period. YEARLY VARIATIONS DUE TO AGE. As the age advances after first calving, the annual test gradually decreases by small amounts for approximately ten lactation periods. About this stage it commences to rise slightly. Tests for the Jerseys and Friesians according to age are set out in Table 22. Figures for the Ayrshires and Milking Shorthorns are not included, for the reason that they were rather irregular, due to the numbers of records representing many of the ages being very few. The percentage total range of variation for the four breeds according to the grouping in the table is about 6 per cent, only, the respective percentages for the Jerseys and Friesians reading 5-8 and 3-7.

The trend of variation in test for the factor of age is more clearly seen in Graph 13, which represents the results given in Table 22. The most striking feature in these two curves occurs at the age of five- years for the Friesians and at six years for the Jerseys. At these points the practically uniform trend of the curves is considerably upset. These points seem quite unaccountable, but if they are omitted and the blanks bridged by the dotted lines shown in the graph it is found that the trend of the curves is quite uniform.

The abnormal point varies slightly with the breed, the ages at which it occurs for the chief New Zealand breeds being as follows : Jersey, six years; Friesian, five years; Ayrshire and Milking Shorthorn, seven years. These abnormal points are all high ones, but when United States and New Zealand Jersey tests are compared for the factor of age, as given in Table 23, quite an astonishing position is revealed. The abnormal point for the American Jerseys occurs, as in the case of the New Zealand Jerseys, at the age of six years, but for some unknown reason it is abnormally low. In the right-hand column of the table the mean of the two results is given. The abnormal point has been absorbed, showing that in the one case it was as high as it was low in the other. Unfortunately, the writer has no American data' except for the Jersey breed. It would be interesting to see if the abnormal points for the other breeds of the United States were al] low ones. One important point, however, is that in all cases the abnormal point is practically coincident with maximum butterfat - production, occurring if anything just before this stage. The ages of maximum production and maturity would be expected to coincide, and thus it appears that the abnormal-test point depends on the reaching of maturity, and is due to a physiological change in the cow at this stage. The connection between the action and changes of metabolic activity and the quality of milk and its Variations is not very well understood, but there seems good reason to believe that this factor is quite an important one.

YEARLY VARIATIONS DUE TO QUANTITY IN MILK-PRODUCTION. This phase of yearly variations was treated in the first article of this series (September, 1924), wherein it was stated that quantity and quality of milk-production were believed to be separately inheritable, but that the two had a limited interdependence and varied inversely, this being most noticeable when one factor varied considerably in either extreme. CONCLUSIONS. Summing up the general position in regard to variations in yearly tests; it is found that in all cases, except for the factor of breed, the variations are of small magnitude. Furthermore, they depend almost entirely on monthly test variations, and if the results of this article are studied in conjunction with those of the previous one a better grip of the subject should be obtained. NOTES. Period of Data. —The Jersey data used in the compilation of Tables 15, 16, and 17, and Graph 11, are for season 1918-19 only. In all other cases, unless otherwise stated, the data used for the. various breeds comprise all first-class records from the commencement of C.O.R. testing (1913) up to 31st December, 1923. . ' Corrections. —In the third article of this series— Journal, February, 1925— (1) the footnote marked f on page 77 should read, “ The equation for this line is : Number of days, in the dip of the lactational-test curve of a particular breed = 722 —(163 times the average test of that breed)”; (2) under the heading of ” Lactational Variations due to Nature of Season,” 1915-16 should be deleted from the list of good seasons.

( Series to be concluded.)

table that those tests this at

Note. —For individual monthly tests see Graph 12.

* This is not necessarily true in cases where the lactation period is terminated before a cow is actually dry, as is the case for some C.O.R.' records. Also in many C.O.R. records, for which cows have milked the full allowed ■ lactation period (365 days) the cows. are far from being dry. ' .. .... • .?

* Exceptions.

* Abnormal points.

* Abnormal points.

Sprin g. Summer. Autumn. Winter. | Spring and Autumn. Summer and Winter; Breed ! Average 1 Number Average i Number : Average ! Annual of Annual i of Annual Test. ! Records. Test. j Records. ! Test. Number Average Number | Average Number Average Number • of Annual of j Annual of Annual of Records. Test. Records. 1 Test. Records. Test. Records. 1 - Friesians . . . . 3’54 6l 5 1 3'56 -55 Milking Shorthorns . . 3-98 218 1 4’°9 20 Ayrshires . . • • 4-12 75 4 -I 4 27 Tersevs . . • • 5’69 134 5'5° *8 3-55 57 3’50 u-5 0 OT 0 o- u/ 3-96 5 4-07 20 3-90 223 4-3-74 3 4-18 8 4-11 78 4 -I 5 35 5’79 4 5’5*5 7 5‘ 6 9 I 3 8 1 5~5 2 T5__ Table 16. —-Variation in Ann ual Test. of Cow s of the Chief B reeds cor nmencrnt during Differen t Month s of the Year. Breed. Aug. Sept. Oct. Nov. Dec. Jan. Feb. March. April. May. June. July. AU Months. s-s8 3’5 T 3’55 3’59 3-46 1 3-58 3’4 8 3’68 3’53 3'55 Breed. Spring. Summer. Autumn.. . Winter. Spring and Autumn. Summer and Winter. Average . Annual Test. Number of Records. Average Annual Test. Number of Records. Average AnnualTest. Number of Records. Average Annual Test. Number of. Records. Average Annual Test. 1 Number of Records. Average Annual Test. Number • of/ Records. Friesians I 1 Shorthorns Milking Shorthorns Milking . Ayrshires Ayrshires . Tersevs J .3 D / 45'65 3’97 4-19 5'7 2 0 J 1 3-98 4‘°3 5’68 3’96 4-06 5’50 4-12 4’49 5'43 4-3-98 5-3-65 3'4 2 5’83 4’49 3‘53 3’90 5-64 4’37 4’47 3T9 3-96 4‘99 4-06 53’99 4’12 5'66 3’54 45’69 615 218. 75 134 44'14 5’5° • 255 20 27 18 3'55 3-96. 3’74 5-7.9 ■ 57 5 3 4: 4■*4‘ i 8 “‘5-56 115. 20 8 T 3-54 3-98 .' 4- 11 5-69 672 223 78 138 45- . 37 .0 4° 3.5 25 Year. of of of breeds four tests . Mean : 2 4’34 436 4’30 4-28 4’39 4’5 2 4-12 4\37 1 4-ri 4-38 4'33 Table 17.— Variation in Tests of the Principal Breeds for Different Months of the Year. (The monthly tests are computed irrespective of the time of commencement of the lactation period.) Breed.. Breed. Aug. Aug. Sept. Sept. Oct. Oct. Nov. Nov. Dec. Dec. Jan. Jan. Feb.'. Feb. March, j April. May. June. July. All | Months.* March. April. .' May. June. July. ' Alli Months. Friesians ;.. Friesians . Shorthorns . Milking Shorthorns . Milking Ayrshires ... Ayrshires . Jerseys - • . Polls . Red breeds tests ■. Mean Jerseys 3’57 3’82 3’99 44'43 4- 5-5-86 3-63 44’43 53M8 3'73 53M6 3'79 4-00 4-17 5’37 — cA H 60-k 3’47 3’93 4’12 4'20 5-66 34’°3 4-4-24 5’71 3- 3-70 4- 4’35 4’17 4’33 4-46 \ 4-60 5'94 j 6-38 3'79 4’43 4-4 2 4’85 6’53 3’88 44’53 44'53 61 44'23 4’54 5-3-56 3'97 4’19 5-72 3-51 3-98 4‘°3 5-68 3-58' 53A5 4-86 • 3-98- - 5-67 3’59 3:65 3:42 5-83'. 4-3-90 • 5-64 3-58 3’53 44'47 • 3:68 '■ 3'79 .. 3-96 . 4-99 3:53 4-06. 4; 11 5:80 , 3’55 ■ 3:99 4-12 5’66 4’34 4-36 4\3O 4:28 . 4'39 • 4-52 4:12. 4’37 4,-H - 4-38. 4-33 Table 17. — Variation in Tests of the Principal Breeds for Different Months of the Year. ■ (The monthly tests are computed irrespective of the time of commencement of the lactation period.) Breed. Aug. ' Sept. , , Oct. Nov. Dec. Jan. Feb. March. : April.’ May. ' June. : July. All Months.* Friesians • . . Mean . 4-71 4' 2 5 4'07 4-16 4-22 4-28 4-3 2 4’47 ' 4-67 4-80 4'94 4‘93 4M8 Shorthorns Milking . Ayrshires . . . Red • . . Jerseys . . 3’82 44’43 55’86 3:63 3- ' 44'43 . 5 -I 5 3A8 3-737 33'79 4-4-17 .5’37 . Ln _00£3’47 3-93 4; 12: 4’2O> 5-66 44:i° 4-4-17 5-3-7° 4’35 4’334‘6o 6-38 3’79 4’43. 4'42 ■4-85 45- ' 6-59 44'53 5- . 6-49 1 3:63 4‘ 1 3 4-23 .4-54 5;85 4-7i 4A5 4-07" 4-16 4'22 .•4’28: 4:32 4-47 4-67 4:80' 4’94 4’93 4M-8 . Mean ... . .. : .

Table 15.- Variation in Annual Tests of Cows of the Chief Breeds commencing during different periods of the year.

Periods. All Friesians. Selected Friesians. Average Test. Percentage Total Range of Variation over Three Periods. ■ Average ' Test. Percentage Total Range of Variation over Three Periods. May, June, July, August March, April, September October, November, December, January, February 3-85 3-64 3'47 10-5 3-72 3-65 3-53 5’2 Average monthly test . . 3-63 3-63

Table 18.- Friesian Monthly Tests for Different Periods of the Year, irrespective of Time of Commencement of Lactation Period : Standard versus Complete Data.

Number of Days empty . after Commencement of Test, both Limits inclusive. z ■' ■ Two-year-old Jerseys ■ for Seasons 1914-15 and 1915-16. Two-year-Friesians for Seasons 1914-15 and 1915-16. 12,000 lb., 13,000 lb., and 14,000 lb. Milk-records of the Four-year-old Friesians for all Seasons. Average Annual Test. Number of ' Records. . Average Annual . Test. N umber of Records. Average Annual Test. Number of Records. Up to 85 5’54 26 . ■ 3-58 11 3'43 13 86-1 .. 5’52 33 3-46 ’ 12 3-47 IO 116-145 • • ' 6-o1 25 3A6 19 3-49 21 146-183 . . 5’47 11 3’59 8 3-46 11 Average tests for all 5-65 95 3’55 50 3-46 . 55 groups 5-65 95 3’55 50 3-46 55

Table 19.— Variations in Annual Tests due to Length of Period of Gestation during Test.

Length of Lactation. Period in Days. Average Annual Test. Number of Records. Up to 325 5'59 24 326-355.. 5-52 15 356-365 • • • • • ■ 5'47 20

Table 20.- Variation in Annual Tests due to Length of Lactation Period : Jersey Data.

Season. ■ Average Mature Equivalent Test. Variations from Mean Mature Equivalent Test. Average Seasonal Rainfall for New Zealand. Variations from Mean Seasonal Rainfall for New Zealand. Average ' Number of Rainy Days per Season for New Zealand. Variations from Mean Number of Rainy Days per Season for New ■ Zealand. 1912-13 .. 3'52 0-00 Inches. 42’35 Inches. — 1-20 155-0 + 4-9 1913-U •• 3'51 — O-OI 45'09 + i-54 165-2 + I5-I 1914-15 •• 3-56 0-04 32-92 — 10-63 133-2 — 16-9 1915-16* . . 3’49 -0-03 , 41-08 - 2-47 I42-9 - 7-2 1916-17 . . 3'51 —o-oi 48-54 + 4'99 - 149-5 — o-6 1917-18* . . 3-58 4-o-o6 49-02 + 5’47 161-4 4-11-3 1918-19* . . 3-51 — o-oi 39-52 - 4-03 144-9 - 5-2 . 1919-20 . . 3’59 +0-07 45’39 + 1-84 140-6 - 9-5 1920-21 . . 3’55 +0-03 38-75 — 4-80 139-5 — io-6 1921-22 . . 3-45 — 0-07 46-76 + 3-21 154-2 + 4-i 1922-23- . . • 3-48 —0-04 49-67 .+ 6-12 164-6 + 14-5 Mean . . 3’52 ±0-03 43’55 ± 4-21 150-1 - ± 9-i

Table 21. —Variation in Annual Tests due to Nature of Season experienced. Total Friesian Data.

JerseysFriesians. Age. Jerseys. Friesians. Average Test. Number of Records. Average Test. Number of Records Years. Two . . 5-60 I. 154 3A6 413 Three . . 5\56 407 3'53 213 Four , . 5'55 250 3'50 137 Five . . 5'53 ' 178 3'59* 79 Six .. 5-61* 152 . 3-5i 66 Seven 5-42 - 99 3'49 55 : Eight. . 5-41 54 3-46 26 Nine . . ' 5-32 5i 3-58 12 Ten and over . . 5'29 46 ' 3-48 4 1- . Average tests for all ages 5'55 2,391 . 3-54 042

Table 22. — Variations in Annual Tests due to Age.

Age. Average Tests of U.S.A. Jerseys. . Average Tests of N.Z. Jerseys. Mean of the Two Yearlings Two years . . 5’43 5’45 1 5-6o 5-52 J 5’60 5'52 Three years 5'4i 5'56 5-49 , Four years 5'41 5'55 5-48 Five years . 5-3° 5-53 5-42 Six years 5-21* 5-6i* 5-4i Seven years 5-28 5-4'2 5-35 Eight years 5-2.5 5-4i 5-33 Nine years 5’23 5-32 5-28 Ten years 5’22 5-28 5-25 Eleven years 5-17 5-39 5-28

Table 23. — Variations in Annual Test due to Age : United States and New Zealand Jersey Tests compared.