THE EFFECT OF GESTATION ON MILK AND BUTTEKFAT PRODUCTION IN DAIRY CATTLE DISSERTATION

Transcription

1 THE EFFECT OF GESTATION ON MILK AND BUTTEKFAT PRODUCTION IN DAIRY CATTLE DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By WILLIAM MATHIAS ETGEN, B. S., M. S. The Ohio State University 1958 Approved by Adviser Department of Dairy Science

2 ACKNOWLEDGMENTS I wish to express my sincere appreciation to my adviser and friend, Dr. Thomas M. Ludwick, Professor in the Department of Dairy Science, Ohio State University, and Project Leader of the Ohio NC-2 Dairy Cattle Breeding Project. I am greatly indebted to Dr. Ludwick for the inspiration, time, guidance, and technical assistance in planning this project and editing the manuscript. My sincere thanks are expressed to Dr. Fordyce Ely, Chairman of the Department of Dairy Science, Ohio State University, for his interest in this project and for his help in editing the manuscript. To Dr. George R. Johnson, Chairman of the Department of Animal Science, Ohio State University, I extend my thanks for his reading of the manuscript. I wish to express my appreciation to Dr. D. Ransom Whitney, Professor in the Department of Mathematics and Director of the Statistics Laboratory, Ohio State University, for his assistance in the statistical analysis of this project. Also, I am greatly indebted to Herman Rickard, Earl Rader, Don Richardson, Ervin Akins, Dr. Edwin Hess, Harry Barr, Dr. Harry Donoho, and Dr. C. M. Clifton. These persons have given unsparingly of their time and efforts in all phases of the study. ii

3 CONTENTS INTRODUCTION... 1 PAGE REVIEW OF LITERATURE... 3 Effect of Gestation on Production... 3 Other Factors Affecting Production... 9 Production Record Evaluation Physiological Aspects of the Effect of Gestation on Lactation... l4 MATERIALS AND M E T H O D S RESULTS AND DISCUSSION SUMMARY AND C O N C L U S I O N S...4? LITERATURE CITED APPENDIX AUTOBIOGRAPHY... 94

4 TABLES TABLE PAGE 1 AGE-CONVERSION FACTORS SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN HOLSTEIN COWS (1,508 OBSERVATIONS) SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN TWO-YEAR-OLD HOLSTEIN COWS (680 OBSERVATIONS) SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN THREE- AND FOUR-YEAR-OLD HOLSTEIN COY/S (480 O BSERVATIONS) SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN FIVE-YEAR-OLD OR OVER HOLSTEIN COWS (348 OBSERVATIONS) SUMMARY OF MULTIPLE REGRESSION COEFFICIENTS AND CALCULATED DIFFERENCES IN PRODUCTION DUE TO DAYS OPEN BETWEEN THOSE COWS CONCEIVING 85 DAYS AFTER FRESHENING AND THOSE NOT CONCEIVING DURING THE 305 DAY L A C T A T I O N FACTORS TO CORRECT FOR INFLUENCE OF GESTATION ON PRODUCTION OF HOLSTEIN COWS (ALL A G E S )... 4l 8 FACTORS TO CORRECT FOR INFLUENCE OF GESTATION ON PRODUCTION OF HOLSTEIN COWS (TWO-YEAR-OLDS) FACTORS TO CORRECT FOR INFLUENCE OF GESTATION ON PRODUCTION OF HOLSTEIN COWS (THREE- AND FOUR- YEAR-OLDS) FACTORS TO CORRECT FOR INFLUENCE OF GESTATION ON PRODUCTION OF HOLSTEIN COWS (FIVE-YEARS-OLD OR O V E R ) ORIGINAL DATA FROM THE LONDON H E R D ORIGINAL DATA FROM THE TOLEDO H E R D iv

5 ILLUSTRATIONS FIGURE PAGE 1 Regressions of milk production on days o p e n Regressions of butterfat production on days o p e n v

6 INTRODUCTION Production records of dairy cows are influenced by many variable factors. Some of these are age, live weight, number of times milked daily, length of lactation period, level of nutrition, housing conditions, care, previous calving interval, previous dry period, and length of gestation. Numerous attempts have been made by many investigators to correct for these variable factors that have had an influence on the amount of milk and butterfat produced during a lactation. Some of these correction factors are in general use today. The factors developed by the United States Department of Agriculture (51,32,33) to correct production records to a standardized 2X, 305 day, mature equivalent basis are valuable in correcting for age differences, for differences in the number of times a cow is milked daily and for differences in the length of the lactation period. It has been thought by many that the number of days a cow is pregnant during her lactation may have considerable influence on the amount of milk and butterfat she produces during that lactation. Several investigators (14,23,24,26,28,29,52,53,55) have shown that a definite inhibitory effect of gestation on lactation exists. However, most of these investigations were conducted before the standardized 305 day mature equivalent factors were adopted. Because of the lack of standardization in these

7 2 studies, there was considerable variation in the length of lactation that was considered. In some of the studies, allowances were made for age, and in others, age differences were not considered. There is a wide difference of opinion concerning the amount of variation in production due to gestation. Some investigators feel that the effect is considerable (as great as 2,500 pounds of milk), while others feel there is very little effect ( pounds of milk). A study conducted under standardized conditions, and designed to measure the effect of gestation on production should be of value to researchers in evaluating production records. This study should also be valuable to farmers and dairymen in analyzing production records of cows in their own herds and records of cows in other herds, from which they plan to purchase breeding stock. The object of this study is to determine the effect of gestation on milk and butterfat production in Holstein-Friesian cattle.

8 REVIEW OF LITERATURE Effect of Gestation on Production The thought that the interval of time which a cow carried a calf during lactation might have some effect on the amount of milk and butterfat she produced is not a new one. As early as 1913, Gavin (26) studied the effect that time of service might have on daily milk yield. He found that the rate of decline in production seemed to increase 16 to 20 weeks following service. He also concluded from his investigation that longer periods of gestation definitely lowered the cow's annual production. Soon after this work was published, Eckles (1^) conducted a study to determine the amount of nutrients required to develop a bovine fetus. He concluded that the nutrients required to develop a Jersey calf were equivalent to approximately 110 to 170 pounds of Jersey milk and those required for a Holstein calf were equivalent to approximately 200 to 275 pounds of Holstein milk. Brody et al. ( k ), in 1923, studied the effect of gestation on production and body weight in Guernsey and Jersey cows. In this study, they divided the production records into two groups, those records from cows which conceived during the third and fourth month of lactation and those which did not conceive during the lactation. They reported that the level of production decreased rapidly in the pregnant cows after the third or fourth

9 4 month of lactation. The pregnant cows averaged approximately 450 pounds less milk per lactation than the non-pregnant ones. These workers also reported that the pregnant cows' body weight started increasing about the same time that the rapid decline in production occurred. Sanders (52) correlated service period and lactation production (using the total lactation as long as 500 days) and derived factors to correct for the effect of gestation on production. His factors varied from + 22 percent for production records from cows bred back in less than 20 days to - 35 percent for production of cows that were open from 480 to 499 days. Hammond and Sanders (29) analyzed the production records of 1,410 cows of eight breeds from four herds in England in an effort to determine the effect of gestation on production. They divided the animals into groups according to their service periods (the time between freshening and rebreeding) by 20 day intervals (0-19, 20-39, etc.) and averaged the production records of each of these groups. They found a correlation of.33 between service period and total milk yield. They also concluded that the association of service period and total milk yield was not linear. The level of production seemed to rise more rapidly when the service period was very short and then tended to level out somewhat as the service period grew longer. Correction factors were calculated using 100 days as a standard service period. These factas ranged from + 30 percent for animals with less than

10 a 20 day service period to - 28 percent for those animals with a service period from 480 to 499 days. In 1924, Ragsdale et al. (49) studied Advanced Registry Production records of Guernsey cows, using yearly milk production as a measure of productivity. They divided the records into two groups, those records from cows bred back in the third or fourth month of lactation and those which did not conceive during the year s production. They found no difference in the two groups in the first five months production, but they observed that the production of the pregnant cows declined more rapidly than the non-pregnant cows after the fifth month of lactation. From their work, they estimated that gestation caused a reduction of 480 to 800 pounds of milk per 12 month lactation in Guernsey cattle. Sanders (55)i in 1928, analyzed 3*918 lactations from four breeds of cattle in England. He divided the lactations into two groups: first lactations, and not first lactations. The records were then divided into groups with service periods of 20 day intervals. The regression coefficient of service period on total lactation yield was calculated. The coefficient was.40 and indicated a significant association between service period and total lactation production. The regression line was tested for linearity, and it was found to be not linear. He then calculated a curvilinear line and derived correction factors from this line using an 85 day service period as the standard.

11 6 A brief summary of these correction factors is given below: Correction Factors vice Period First Lactation Others of /O % A study was also conducted to determine whether there were breed differences or whether there were differences between high and low yielders. He could find no significant differences in either case. Gowen (28) used Advanced Registry Guernsey records to conduct a study of the effect of gestation on production. He corrected the records somewhat for age differences by grouping the records of the cows by six months age intervals (2-0 to 2-6, 2-6 to 3-0,

12 7 etc.). A significant correlation was found between milk production and the days a cow carried calf during her 3&5 day lactation. He estimated a difference of 400 to 600 pounds of milk between cows carrying calf nine months and cows standing open during the entire lactation. He also concluded from this study that gestation effects were least on young and old cows and greatest on five-year-old cows. Gaines and Davidson (2*0 reported no differences in records of cows which carried calf less than five months compared with those not carrying calf during their 3&5 day lactation. These investigators also used Advanced Registry Guernsey records in this study. Eckles and Anthony (15) conducted a study of the effect of gestation on production. Lactation records from 19 cows, which had a calving interval of one year or less for one lactation and then did not conceive for six months or more during another lactation, were compared. They reported that in the lactation where conception did not take place for at least six months, the 365 day records of the animals averaged 1,1**2 pounds more of milk and 56 pounds more of butterfat. The decline in production seemed to increase more rapidly in the pregnant cows after the fifth or sixth month of gestation. Hevens (47) also reported that the production decreasing effect of gestation seemed to begin after the first five months of pregnancy. He concluded that a ten-month lactation was depressed no more than 200 to

13 8 500 pounds of fat-corrected milk, due to the effects of gestation. Erb et al. (17) studied the effect of gestation on production by using 82 intra-herd records, which were made under conditions where the cows were fed and managed for near maximum production. Daily production records from 40 Holsteins, eight Jerseys and five Guernseys were included in the study. These authors reported that production of the Jerseys and Guernseys dropped rapidly for the first ten days after conception. They also observed the first noticeable effect of gestation on production about 180 days after conception, regardless of the number of days after calving. The records were grouped by production levels and a variation in the intensity of decline in production was noticed; however, these differences in rate of decline were insignificant. They also observed less effect of gestation on production in two- and three-year-old cows than in older cows. No difference in effect of gestation was observed when records from cows milked two- and three-times per day were considered. The rate of decline in milk production 180 days after conception was twice as fast in cows conceiving in April and May as compared with those conceiving in June. In a later study which involved 2,491 Dairy Herd Improvement Association records, Erb (16) studied the effect of gestation on production records that were adjusted for dry period and length of lactation. He observed an average difference in

14 9 production of 1,111 pounds of four percent milk between cows carrying calf 230 to 239 days compared with thobe not carrying calf during the 305 day lactation. He concluded that a 305 day record was reduced about three quarts of four percent milk (approximately 6.^5 pounds) for each day a cow was in calf over 100 days during the lactation. Thompson et al. (58)* while studying the effects of several non-genetic factors on production, did not find a significant difference in production due to gestation. Espe and Smith (18) estimated that cows in the tenth month of lactation, which have been pregnant for eight months, will give possibly 20 percent less milk than open cows. Other Factors Affecting Production As early as 1913i Carrol (7) studied the effect of varying length of dry period on production. He compared production records made after a one-month dry period, a two-month dry period, and a three-month dry period. He observed that a two-month dry period appeared more beneficial than either one-month or three- months. Numerous other investigators (1,2,3,11,16,29,3*0 agree that a dry period of something over 30 days but under 90 days is optimum. Goodwin and Erb (27) concluded that cows with a 0-9 day dry period averaged l,*t-59 pounds less of milk and 62 pounds less of butterfat in the following lactation than did cows that had a dry period of days.

15 10 Tyler and Hyatt (60) analyzed the standardized mature equivalent records of 2,203 cows to determine if length of calving interval had any effect on production. They reported significantly lower production for those animals having a ten- or eleven-month calving interval, when compared with those having a calving interval of 12 months or more. They did not observe any significant difference between those cows having a 12 or 13 months' calving interval and those having a 1^ or 15 months' calving interval. Erb (l6) observed a considerable 305 day production loss in cows having a calving interval of less than 360 days, but very little difference between cows having a calving interval of 360 days or more. These investigators agree very closely with the findings of other researchers (7,11,25) who have studied the effect of calving interval on production. Carrol (7), in 1913, reported that cows freshening in the fall produced more milk per lactation than did cows freshening in the spring. A few years later McDowell (39) also observed that cows freshening in the fall had a higher average production than those freshening at any other time of the year. Cows freshening in the winter had the second highest production, those freshening in the spring ranked third, and those which calved in the summer had the poorest production. Other investigators (2,6,11,30,^8,53,58,59) have reported finding seasonal differences in production as great as 1,000 pounds of fat-

16 11 corrected milk, and generally agree that the fall or early winter months are the best times to freshen cows for optimum production. Some investigators (63,6*0, however, report only minor differences due to season of freshening. The influence of age at freshening on production has been demonstrated by numerous workers (2,9,11,15»22,31,32,33i58) Kendrick (31>32,33) calculated correction factors which are now commonly accepted and are popularly known as the standardized 2X, 305 day, mature equivalent factors. It was shown that twoyear-old dairy cows generally produce a smaller amount of milk and butterfat than older cows and gradually increase in production until they reach their maximum performance at six or seven years of age. They then slowly decline in level of production as age increases. It is generally agreed by those who have investigated the influence of times-milked per day on production that the more frequently a cow is milked per day (to as many as four-times per day), the more milk she produces (10,31,32,33,35,36,37). There is some disagreement, however, on the magnitude of the difference in production between cows milked two-times daily compared with those milked three- or four-times daily. Some workers (10) believe that.8 is the best factor to use to convert a threetime record to a two-time record, while others are of the opinion that.833 (32) is a better factor. Still others (36) feel that because of the greater persistency of two-year-old cows, their

17 12 three-time records should be multiplied by.8 while the threetime records of older cows should be multiplied by.833 to convert them to a two-time basis* Some workers (8) also report an increased butterfat percentage when cows were milked threetimes daily as compared to those milked twice daily, while others (10) report a smaller increase in butterfat production than in milk production. Production Record Evaluation As early as 1920, McCandlish and Winters (38) computed factors that were used to standardize the production records of dairy cows. These investigators conducted a study comparing the production of the daughters of a bull with their dams. In order to make a more accurate comparison, they introduced age-allowance factors for records of the younger cows, and factored their records to a mature or five-year-old basis. Ragsdale et al. (50), in 1924, analyzed 46,000 yearly records and 104,000 seven day records in a study designed to correct for age differences. They classified these records by yearly intervals based on age at freshening. The classified groups of records were averaged and plotted, and mature equivalent factors were calculated from this plotted line. Other investigators (9,19) performed similar studies and also computed correction factors for age differences. Others (6,23,52,53,54,

18 13 55) computed factors to correct for differences in length of dry period, season of freshening, and days carried calf. In 1935* when the proved sire program was inaugurated, a set of all-breed age-conversion factors were calculated by the United States Department of Agriculture (31) These factors were based on a study of the Dairy Herd Improvement Association production records available at that time. By the use of these factors, an attempt was made to standardize for differences in age, length of lactation period, times milked per day and influence of gestation. These factors were revised in 19^1 (32), and individual factors were computed for each of the different breeds (Ayrshire, Brown Swiss, Guernsey, Holstein, Jersey, and Shorthorn). It was found that the age-conversion factors obtained for the Ayrshire, Jersey, and Guernsey breeds were not significantly different from each other, so they were combined. There were relatively few Shorthorn cows reported, so Brown Swiss data were used to develop factors for Brown Swiss and Shorthorn cows. In 1955 (33)i the factors for standardized 2X, 305 day, mature equivalent production records were revised again. The factors published at that time were calculated from more than 500,000 records from dairy cows representing six breeds. There has been some criticism of the standardized factors developed by the United States Department of Agriculture (36), but most dairymen and researchers have accepted these factors as

19 14 a valuable tool in the evaluation of dairy production records which are used in the sire proving program and in research in the field of dairy science. Factors have also been computed (21,37,61) which are valuable in calculating partial records to 305 day completed records. Records adjusted by the use of these factors have been shown to have a high correlation with the actual completed records. Physiological Aspects of the Effect of Gestation on Lactation Within the last four decades, considerable research has been conducted in an effort to determine the role of the various hormonal and other physiological factors that control lactation. Some of this research has been performed with dairy cows and some with laboratory animals. Most researchers in the field of milk secretion are of the opinion that production in dairy cattle is adversely affected, to some degree at least, by gestation. There is, however, considerable disagreement among these investigators as to the cause of this inhibition, and several theories have been proposed in an effort to explain this decreased production due to gestation. One of the early theories was that decreased production due to gestation was caused by competition for nutrients by the growing fetus (14,24,28,47). Gowen (28) computed the nutrients required to produce a bovine fetus on an energy basis and concluded that the requirement was therms. This would be

20 15 equivalent to pounds of milk. Brody et al. (4,5) observed that the body weight of the gestating cow increased more rapidly approximately the same time that the rate of decline in milk production increased. These researchers concluded that the demand for nutrients required for the growth of the gestating cow and those required to nourish the fetus, reduced by an equivalent amount the supply available for the mammary gland, and thereby reduced the production of the gestating cow. Another of the earlier theories (15) on the effect of gestation on production is that halfway through pregnancy, the corpus luteum secretion stimulates some of the secretory cells of the mammary gland to grow rather than to secrete. These authors were of the opinion that as pregnancy advances, there is more growth and less secretion by these secretory cells and as a result milk yield decreases. At parturition, or shortly before, the corpus luteum degenerates and the secretory cells gradually become functional again. The experimental evidence leading to the development of this theory was obtained from experiments performed on goats. The possibility that a hormone which inhibited milk production was present during gestation, was advocated by Gaines and Davidson (24). Since then, there have been numerous experiments in which the workers have tried to determine whether there were inhibitory effects of the various hormones present during pregnancy.

21 16 Folley (20) found that large doses of estrogenic hormones administered to dairy cows (as much as 485 milligrams of oestrone and 6? milligrams of dihydro-follicular-benzoate injected in five doses over a period of two- and one-half days) had a temporary inhibitory effect on production. However, when administration of these estrogenic substances ceased, milk production returned to normal. He also observed an increase in butterfat percentage and in solids-not-fat following the administration of these estrogenic substances to dairy cows. Spielman et al. (57) injected ten to 100 milligrams of diethylstilbestrol at irregular intervals to four lactating cows. These workers did not observe any effect on milk yield, but did observe an increased butterfat and lactose percentage. These increased percentages returned to normal within five days after the last injection. Reece and Murphy (51) injected four lactating cows with diethylstilbestrol proprionate and observed a decline in production in all four animals for a period of ten days following injection. Meites and Turner (43) reported decreased lactation in rats due to injections of diethylstilbestrol. Donker and Peterson (12) found that ten milligrams of diethylstilbestrol given every second day to dairy cows increased production. When the same dosage was given every day, production was either greatly inhibited or stopped entirely. Other workers (44,46) found that small doses of estrogen stimulated lactation in goats, but that large doses inhibited lactation.

22 1? Nelson (^5) found that large amounts of oestrin were inhibitory to milk production in guinea pigs, while small amounts had little or no effect. He believed that oestrin may have inhibited the production of the lactogenic hormone or may have acted directly on the mammary gland. Meites et al. (^2) has indicated that estrogen and progesterone may act together to inhibit the action of prolactin on the mammary gland. In his work with rabbits, he found that when either estrogen or progesterone was administered with prolactin, milk production remained high, but when both of these ovarian hormones were given with prolactin, milk production was decreased. In later studies, Meites and Sgouris (40,4l) found that when rabbits were given amounts of estrogen and progesterone which were optimal for udder development, along with large doses of prolactin, a fair degree of lactation was obtained. When lesser amounts of the ovarian hormones were given, moderate amounts of prolactin were effective in initiating lactation. They concluded that the level of estrogen and progesterone which is optimal for mammary growth does not completely inhibit lactation if a sufficient amount of prolactin is present. They were of the opinion that as pregnancy continues, the level of prolactin remains high as a result of continued milking stimulus, but as the level of progesterone and estrogen increases during later pregnancy, the action of prolactin in maintaining lactation is impaired and the level of milk production decreases.

23 Walker and Mathews (62) observed that simultaneous injections of estrogen and progesterone did not prevent the initiation of lactation in rats, but that it did inhibit established lactation in this species. One could conclude, from the literature reviewed, that the competition for nutrients is partially responsible for decreased production due to gestation in dairy cattle, but that the fluctuation in the levels of the various hormones present in the dairy cow during lactation and gestation is also partially responsible for this effect.

24 MATERIALS AND METHODS These data were taken from the Ohio North Central Dairy Cattle Breeding Project.'*' This is a long-range project with the over-all objective of improving dairy cattle through breeding. An objective of this nature necessitates a long-range program; therefore, many important additional projects have developed. This analysis of the effect of gestation on production is one of these additional projects. The cattle used in this project are owned and managed by the Ohio Department of Mental Hygiene and Correction. They are purebred Holstein-Friesians but registrations have not been kept up to date on all cattle. These cattle were in two herds located in different sections of the state. One herd is located at a mental institution and the other at a correctional institution. The management conditions in these herds are standardized to the extent that both herds are under the supervision of the state's chief agriculturist. Management conditions do vary, however, because there are differences between the individual herdsmen. Also, management conditions vary from one herd to the other because of the differences in type and kind of feed availa VA cooperative dairy cattle breeding project with the United States Department of Agriculture, the Ohio Agricultural Experiment Station and the Ohio Department of Mental Hygiene and Correction. 19

25 20 ble. Herd I (London Prison Farm) has somewhat better management conditions than herd II (Toledo State Hospital), but both herds are under relatively good management conditions. Both of these herds are enrolled in a combination Herd Improvement Registry and Dairy Herd Improvement Association testing program. All production records used in this study were obtained from these two sources, and were standardized to a 2X, 305 day, mature equivalent basis. All cows were milked threetimes a day except those animals in herd II which freshened after January 1, These animals were milked twice a day. The records of all cows milked three-times a day were adjusted to two-times by multiplying by.8. The Holstein Conversion Factors shown in Table I (32) were used to standardize the records for age differences. The record of any cow going dry before 305 days was assumed to be a complete record. Any record that was started by or interrupted by an abortion (any calving of more than 151 days or less than 250 days gestation) was not used. Any record, other than first lactation records, that was preceeded by less than a ^fo day dry period or more than a 90 day dry period was not used. Any first lactation record from an animal that freshened before two years of age or over three years of age was not included in this study. This age restriction eliminated less than five percent of the records. If other environmental influences were known to severely influence (if a decline in production of at least 50 percent

26 21 TABLE 1 AGE-CONVERSION FACTORS Age Factor Age Factor Age Factor U

27 22 occurred at the time of obvious accident, injury, sickness etc. and the decline persisted) a production record, this record was eliminated from this study. A preliminary survey showed approximately an equal distribution of calvings by months and seasons of the year. Since this equal distribution existed, no effort was made to adjust for season of freshening. This study involves 1,508 standardized 2X, 305 day, mature equivalent records. Of these, 680 were records from two-year- old cows, **80 were from three- and four-year-olds, and 3^8 were records made by cows five-years-old or older. The records were made in two herds over a period of ten years. The number of days open (that is, the number of days during the 305 day lactation that the cow was not pregnant) was calculated from the day of freshening to the day of conception. Since the term "days open" is easily and accurately computed, it will be used in all further discussions. It is understood that "days open" is inversely related to the period of gestation during lactation. If conception did not occur within 305 days of the date of freshening, the animal was credited with 305 days open regardless of the eventual date of conception. A multiple regression analysis, as explained by Snedecor (56), was considered a logical approach to this study. The regressions of production on days open squared and days open cubed, as well as production on days open, were calculated.

28 23 This was done to determine whether the relationship might be represented better by an exponential function than a linear one. The data were punched on IBM cards to facilitate the analysis. There were 15 independent variables involved in this analysis: 1. the year 19^7 2. the year 19^8 3. the year 19^9 k. the year the year the year the year the year 195^ 9. the year the year herd 12. age 13- days open l*f. days open. squared The two dependent variables were milk and butterfat production. 15. days open. cubed

29 z b The multiple regression formula for case 1 (the regression of milk production on days open) follows: where: Y1 = b lxla + b2x2a + b3x3a b15x15a + bl6 b.x.. thru b_x_. = variance due to years 19^7 thru la 9 9a bn_.xn_ 10 10a = variance due to herd b^i^lla = variance due to age b^^x^2a = variance due to days open b ^ X ^ a = variance due to days open squared b ^ X ^ a = variance due to days open cubed b ^ X ^ ^ = milk production bl6 =! cons^an'*' Case 2, where butterfat production was the dependent variable, was calculated using the same formula as that used when milk production was the dependent variable. Multiple regression coefficients were obtained for milk production and for butterfat production on days open, days open squared and days open cubed. These coefficients were calculated with the influences of year, herd, and age removed. These coefficients measured the effect of gestation on production devoid of many environmental influences which might bias the results. The data were also sorted into three groups according to age at freshening. The groups were (1) two-year-olds, (2) three- and four-year-olds, and (3) five-year-olds and over. A

30 25 multiple regression was calculated on each of these groups using the same basic formula described earlier. The "t test was used to test the significance of the multiple regression coefficients and an F-ratio was calculated to test the significance of the multiple correlation coefficients. Correction factors were calculated from the regressions by dividing the daily increase due to days open by mean lactation production, and adding this amount for each day open less than 85, and subtracting this amount for each day open over 85. This period of 85 days was used as the standard because those cows bred back about 85 days after freshening have a calving interval of approximately one year. This annual calving interval is considered optimum by most dairymen.

31 RESULTS AND DISCUSSION In a study of this nature it is difficult to control all environmental factors that might influence the data or the interpretation of the results. In this study, year, herd, and age differences were corrected by the use of the multiple regression analysis. Length of previous dry period was corrected to a large extent, by elimination of those records that were not preceeded by a day dry period. The production records were not corrected for differences due to season or month of freshening since these influences were assumed to be very minor because of equal distribution. The records were factored to a 2X, 305 day, mature equivalent basis. This procedure adjusted for differences due to age at freshening, times milked daily, and length of lactation. Elimination of records in which an a- bortion, severe accident or injury, disease, etc. occured should also help eliminate other environmental influences that might bias the data. It is possible that other unobserved environmental influences could have affected the data. Some of these might have been changes in types, amount and quality of feed available, changes in milking procedure and management personnel, changes in genetic composition of the herds, and unrecorded health conditions. In this study, however, an effort was made to 26

32 27 standardize the data so that the results would be an accurate estimate of the effect of days open on milk and butterfat production in Holstein dairy cows. Table 2 gives a summary of the results of the multiple regression analysis of milk and butterfat production with days open of all records included in the study (1,508 observations). The multiple regression coefficients measured the effect of days open on milk and butterfat production, after the records had been corrected for effect of year, herd and age at freshening. It can be observed from Table 2 that the regressions for the squared and cubed values were smaller than the regressions for the unit values for production on days open. The standard deviations for the exponential functions were also much higher than the standard deviations for the linear values. For this reason, the remainder of the analysis and evaluations was concerned only with the linear regression of milk and butterfat production on days open. A highly significant relationship (.01 level of probability) was observed both in the multiple correlations and in the multiple regressions between milk and butterfat production and days open. The multiple regression coefficient of for milk and.2633 for butterfat indicated that for each additional day open, the average cow in this study produced pounds more of milk and.2633 pounds more of butterfat during that lactation. At the level of production of the cows in this study

33 TABLE 2 SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN HOLSTEIN COWS (1,508 OBSERVATIONS) X Sx B^ Sx of B]_ t r R F Days Open Milk U.7 Butterfat UU3» Days Open with Milk b b Days Open with Butterfat b i.lUi.Ol4b (Days Open) 2 with Milk 2.768U b b (Days Open) 2 with Butterfat U3a b (Days Qpen)3 with Milk 1.08US 10.05U b (Days Open) 3 with Butterfat.09IIU.36U b.I60U M b a - significant at.05. b - significant at.01. X - mean. Sx - standard deviation. B]l - multiple regression coefficient.. Sx of Bn - standard deviation of multiple regression. t - t-value for multiple regression. r - simple correlation coefficient. R - multiple correlation coefficient. F - F-ratio for multiple correlation. ro 00

34 29 (average 12,703 pounds of milk and pounds of butterfat), the difference in production due to gestation between a cow that conceived 85 days after freshening and one that did not conceive during the 305 day lactation was 1,832 pounds of milk and 58 pounds of butterfat. The results of the multiple regression analysis of milk and butterfat production with days open on two-year-old, first lactation records are given in Table 3«The multiple regression and correlation coefficients were smaller for this group than they were when all cows were considered. These values were 5*74l4 pounds of milk and.1595 pounds of butterfat per day open. However, they did indicate a highly significant (.01 level of probability) relationship between days open and milk and butterfat production in two-year-old Holstein cows. The average production, 13,202 pounds of milk and pounds of butterfat, was higher for this group than the group which included all cows. The average days open was about the same for both groups ( days for all cows and for the two-year-olds). The average difference in production, attributable to gestation, of cows which conceived 85 days after calving and those which did not conceive during the 305 day lactation was 1,263 pounds of wriik and 35 pounds of butterfat per lactation. In Table 4, a summary is given of the results of the multiple regression analysis of milk and butterfat production.with days open of three- and four-year-old Holstein cows. Regressions

35 TABLE 3 SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN TWO-YEAR-OID HOLSTEIN CCMS (680 OBSERVATIONS) X Sx Bi Sx of B^ t r R F Days Open 13U«lU 68.1U Milk Butterfat I46I Days Open with Milk 5.7U1U I.D4OI b U.2il98b Days Open with Butterfat U i;b U876b b - significant at.01. X - mean. Sx - standard deviation. B^ - multiple regression coefficient. Sx of B^ - standard deviation of multiple regression. t - t-value for multiple regression. r - simple correlation coefficient. R - multiple correlation coefficient. F - F-ratio for multiple correlation.

36 TABLE 1+ SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN THREE- AND FOURr-YEAR-OID HOLSTEIN CONS (U80 OBSERVATIONS) X Sx B^ Sx of B^ t r R F Days Open Milk Butterfat Days Open with Milk *3320b i 63b Days Open with Butterfat.30U b ll.l+280b b - significant at.01. X - mean. Sx - standard deviation. B]_ - multiple regression coefficient. Sx of Bj - standard deviation of multiple regression. t - t-value for multiple regression. r - simple correlation coefficient. R - multiple correlation coefficient. F - F-ratio for multiple correlation.

37 32 of pounds of milk and pounds of butterfat daily were observed in this group of cows. The multiple regression and correlation coefficients again indicated a highly significant (.01 level of probability) relationship between days open and production. The mean days open was days which was lower than that of the previous groups. The mean production of 12,073 pounds of milk and pounds of butterfat was also lower than that of the previous groups. In this study the mean difference in production due to days open between a cow that conceived 85 days after calving and one which did not conceive during the lactation was 1,887 pounds of milk and 67 pounds of butterfat per 305 day lactation. The magnitude of the effect was very similar to that obtained for all animals. A summary of the results of the multiple regression analysis on the five-year-old and over group is given in Table 5* The multiple regression coefficients of pounds per day for days open and milk and.4189 pounds per day for days open and butterfat, as well as the multiple correlation coefficients, were highly significant (.01 level of probability). The means for milk and butterfat production as well as that for days open, were nearly the same as those for the overall group. The regression coefficients were the highest for this group, indicating that gestation probably has more effect on older cows than on younger cows. In this group of cows, the difference in production due to days open between cows conceiving 85 days

38 TABLE 5 SUMMARY OF THE MULTIPLE REGRESSION ANALYSIS OF MILK AND BUTTERFAT PRODUCTION WITH DAYS OPEN IN FIVE-YEAR-OLD OR OVER HOISTEIN COWS (3U8 OBSERVATIONS) X Sx Bi Sx of B-l t r R F Days Open 133* Milk 12^ Butterfat k Days Open with Milk a b.370I Ub Days Open with Butterfat U189 * b.3600 I U638b b - significant at.01. X - mean. Sx - standard deviation. B^ - multiple regression coefficient. Sx of Bt - standard deviation of multiple regression, t - t-value for multiple regression, r - simple correlation coefficient. R - multiple correlation coefficient. F - F-ratio for multiple correlation.

39 after freshening and those not conceiving was 2,685 pounds of milk and 92 pounds of butterfat for each 305 day lactation. Figures 1 and 2 are graphic presentations of the regressions of milk production on days open and butterfat production on days open for two-year-old cows, three- and four-year-old cows, fiveyear-old or over cows, and for all cows. A summary of the regression coefficients of all studies and the calculated differences in production due to days open between cows which conceived 85 days after freshening and those which did not conceive during the 305 day lactation is given in Table 6. It can be observed from this table that the regression coefficients and the differences in production due to an additional 220 days open for the all-cows group and for the three- and fouryear-old group is very similar. The regression coefficients and the difference in production for the two-year-old group are considerably less than any of the other groups. This difference is probably due, in part, to the higher persistency of first lactation cows. The regression coefficients and production differences in the five-year-old group were more than twice as high as the two-year-old group and somewhat higher than that of the three- and four-year-old group. Decreased persistency in older cows probably accounts for part of this difference. It is also shown in Table 6 that the mature equivalent production level was the highest in the two-year-old group, second highest in the five-year-old and over group, and lowest in the

40 r- All Cows Two-Year-Old Cows... Three- and Four-Year-Old Cows Five-Year-Old and Over Cows. l h O O O Pounds of Kilk _L _L Days Open 2ii0 300 Figure 1. Regressions of Kilk Production on Days Open.

41 All Cows Two-Year-Old Cows... Three- and Four-Year-Old Cows, Five-Year-Old and Over Cows.. -* *- 520 Pounds of Butterfat hqo hho U00 -L *0 300 Days Open Figure 2# Regressions of Butterfat Production on Days Open.

42 TABLE 6 SUMMARY OF MULTIPLE REGRESSION COEFFICIENTS AND CALCULATED DIFFERENCES IN PRODUCTION DUE TO DAYS OPEN BETWEEN THOSE COWS CONCEIVING 85 DAYS AFTER FRESHENING AND THOSE NOT CONCEIVING DURING THE 305 DAY LACTATION Age Group N X Prod. MILK % Pounds Difference X Prod. BUTTERFAT Bp Pounds Difference 1. Two-Year-Olds U1U 1263 : 1* Three- and Four-Year-Olds U U 1886 U L Five-Year-Olds or Over 3U * All Cows *1* N - number of cows. X - mean. Bp - multiple regression coefficient.

43 38 three- and four-year-old group. This fact indicates that possibly the age-conversion factors used in this study were slightly high for the two-year-old cows and somewhat low for the three- and four-year-old cows. A difference in production from 1,263 to 2,685 pounds of milk and from 35 to 92 pounds of butterfat per 305 day lactation due to the effect of carrying calf 220 days or not carrying calf during the lactation could be important to many people. Those in charge of selecting bulls for use in artificial insemination could more accurately evaluate the performance of the animals involved. At present, much time and effort is expended by bull selection personnel in evaluating environmental conditions under which the daughters of a bull have made their production records. At best, these evaluations are somewhat of an estimate of hov; much the daughters' records were influenced by varied environmental conditions. It is the opinion of the author that it would be time well spent to correct these records for effect of days open on production. This would discourage breeders from the practice of delayed breeding which is sometimes used in an effort to obtain an impressive proof on a bull. Farmers and breeders would also do well to consider the effects of days open on production when selecting young bulls or female replacement stock for their herds. Many cows are not rebred or fail to conceive until they are well along in their lactation, thus giving them an opportunity to make a more im

44 39 pressive record. Many times bull calves or female offspring are sold from these cows because they have made this high record. If these records had been adjusted to correct for effect of days open, certainly the prospective buyer would have had a better indication of the producing ability of these cows. Production records of dairy cows are the criteria used to evaluate many research projects in the field of dairy science. Feeding programs are often tested on the difference in production of small groups of cows receiving the different rations to be tested. Results of these feeding trials could be biased considerably, if there were a wide variation in days open between the groups of cows and if production differences due to variations in days open were not adjusted in the analysis. The difference in production due to effect of days open could be as large or larger than the difference in production due to the effect of the different rations. Results of other dairy cattle research projects such as breeding experiments, housing projects, grazing trials, and many others could be misinterpreted if there were large variations in days open between the groups being compared. Because of the magnitude of the effect of days open on miiv and butterfat production in dairy cattle, it was thought that correction factors, based on the results of this study should be calculated. It is realized that the number of animals (1,508) in this study is not exceedingly large and that only one breed