During the past 100 years the U.S. dairy industry has experienced a broad array of major changes including:

1. A sharp reduction in total cow numbers,
2. A near six-fold increase in average production per cow,
3. Substantially greater total annual milk production,
4. A steep decline in the number of herds,
5. But a large increase in the cows per herd,
6. A shift in cow numbers from the north and east to the west,
7. A cow population now dominated by Holsteins, but a recent resurgence of Jerseys,
8. A decline in the per capita consumption of whole milk equivalent and fluid milk, but a marked increase in the consumption of cheese.

The capacity and willingness of the U.S. dairy industry to produce milk has equaled or exceeded its capacity to market milk and dairy products during nearly all of the past century. Consequently, prices received for milk by producers have often been less than the costs of production, so that many producers have exited the dairy industry. Table 1 contains production variables and prices received for milk during most of the past 100 yr. In Table 2 is a list of important events which have shaped the U.S. dairy industry throughout the past century. My objectives are to identify some of the dominant changes in the U.S. dairy production industry for the past century and discuss some of the primary factors responsible for these changes.

Nearly 17 years ago, Walton (1983) predicted that by the year 2000, top individual records would approach 70,000 lbs per year and the best herd averages would be near 35,000 lbs per cow per year. Although at that time this prediction seemed unduly optimistic to me, these records are today a reality. Table 1 contains data for several production variables at 10 yr intervals. Dairy cow numbers began this century at about 17 million, peaked in 1944 at 25.6 million and today are at about 9.1 million. Annual production per cow began this century at just over 3,000 lbs, increased to 5,314 lbs in 1950 and today is above 17,000 lbs (Figure 1). So most of this large increase in production occurred since 1950. I believe that this striking increase in yield per cow has been the driving force for several of the other changes which have happened, so that the reasons for this huge increase will be identified and discussed.

Although the change in breed composition accounts for some of this increase because Holsteins now constitute about 94% of the dairy cow population, up from less than 50% in 1950. However, the major factors responsible for this large increase are a combination of improved genetics, nutrition and management.

Genetics. From genetic evaluation of Holsteins in all states on DHI test with sire identification, it was found that the genetic trend for milk increased from –2849 lbs in 1961 (using 1987 as a base of 0), to +2036 lbs in 1998 for a net gain of +4885 lbs (Everett, 1999). The average milk production (mature equivalent basis) increase of this population of cows (in 1961 about 8,000 cows and now nearly 400,000) was 13,555 lbs in 1961 and 26,194 lbs in 1998 for a phenotypic increase of 12,639 lbs (26,194 – 13,555). So one can estimate that increased genetic merit for milk accounted for 36% of this gain (4,885/12,639) since 1961, and the remainder or 64% of the gain (7,754 lbs) can be attributed to some combination of nutrition and management. On an energy corrected milk basis, the Jersey breed has made similar genetic progress especially for the past 20 years (Everett, 1999). For this large gain in genetic progress to have occurred required the simultaneous application of three technologies, all of which were developed and used successfully in this century First, milk production records had to be available for a large number of cows with sire identification. These records were available through the Dairy Herd Improvement (DHI) Program which began in 1905. Data in Table 1 show that by 1950, more than 40, 000 herds and 1 million cows were enrolled in official plans (Majeskie, 1996). By 1998 the number of herds had declined to 27,000, but because of the large increase in herd size, the number of cows had increased to 3.4 million, which represented 37% of the U. S. Cows (Table 1 and Figure 2). If those herds and cows enrolled in DHI management plans were also included, total herds would be 38,920 and total cows would be 4.4 million or 48.3% of the U. S. dairy cow population.

Artificial insemination (A.I.) was the second essential technology and it became a practical reality in 1938 when the first association was established in New Jersey. Later in the 1950s, the development of frozen semen made its use much more feasible. Early growth of A.I. can be seen in Table 3 in which the dairy cows bred by A.I. totaled 7,539 in 1939 and increased to more 7 million by 1971. Data on domestic dairy semen sales from 1971 through 1998 show just over 11 million units sold in 1971 then a peak in 1980 at 14 million units. Sales have been flat since then at about 13.5 million units (Anonymous, 1999a). However, the decline in dairy cow numbers during this 28 yr interval has been from ll.8 to 9.1 million (Table 1). Doak (1999) recently estimated that about 70% of the dairy cow population and 30% of the dairy heifers are now being bred using A.I. But A.I. would never have been adopted widely unless methods became available to identify sires of superior genetic merit for production. The third technology was the construction, testing and application of sophisticated statistical models, especially by Henderson (1973) and his students at Cornell. These models used the DHI production records of daughters of sires to rank those sires based on their transmitting ability for milk and its components. Therefore, these 3 technologies, the DHI record program, A.I., and statistical models to rank sires, are responsible for most of the large gain in genetic merit for production seen especially in the past 40 years. If one assumes that embryo transfers (ET) are made only with embryos from cows and sires of superior genetic merit, then this technology has also contributed to genetic progress. A large majority of the ET have occurred in the Holstein breed where it began with 18 ET registrations in 1975, peaked in 1994 with 21,912, and declined to 16,528 in 1997 (Majeskie, 1996).

Table 3. Early growth of the Artificial Insemination Industry in the U.S.

Sire Total Cows per Sire

Year Orgs. Dairy Cows Sires

(No.) (No. Bred) (Dairy) (Average)

1939 7 7,539 33 228

1948 91 1,713,581 1,745 1,078

1958 71 6,645,568 2,676 2,483

1968 32 7,138,636 2,028 3,334

1971 26 7,285,171 1,938 3,620

from USDA and the National Assn. Animal Breeders (NAAB)

Nutrition. An important advantage of the pervasive genetic merit for high milk yields is that dairy cows are highly responsive to good environment in the broad sense of that word. During the 1960s it was discovered that cows responded very positively with more milk when concentrate feeding was increased above about 20 lbs per day a traditional amount fed. Another way dietary energy intake has been increased is with supplemental fat (Palmquist, 1986). These strategems which increased the intake of energy showed clearly that the cows of that era had the capacity to respond to the increased dietary energy with greater milk yields. Another special feature of better nutrition during the past 40 years has been the emphasis and use of forage and feed testing. This program proved an excellent educational tool to demonstrate the value of higher quality forage, but it also permitted the demonstration that cows could produce relatively large volumes of milk on diets with a high proportion of quality forage and it allowed more precise ration formulation. Last year the DHI Forage One Lab in New York analyzed well over 100,000 samples (Sirois, 1999).

.

Through most of the past century pasture was a dominant forage for most dairy herds in the U.S., followed by hay and silage. Nearly all dairies fed concentrates in the barn at the time of milking so that eating occurred during milking. As herd sizes increased during the 1960’s there was an increased emphasis on labor efficiency and a corresponding increase in the construction of milking parlors. But as production increased there was also greater mechanization in the parlor which further reduced the time cows spent there to be milked. Even when additional time was provided and precision equipment was operated carefully, there was no way to ensure that the appetite of the cow would induce her to consume the required concentrate. Most parlor concentrate feeding then became free-choice feeding for two 10-min intervals per day. And the resulting "slug feeding" of highly soluble protein and carbohydrate must have been highly disruptive to a ruminal fermentation system which functions best as a steady state system. The resolution of the parlor concentrate feeding dilemma has taken several forms, but the totally mixed ration (TMR) system, used earlier in beef cattle feedlots, has now become the feeding system of choice. This is especially so where silages or other wet forages are used which serve to bind the ingredients together. A TMR can be defined as a quantitative blend of all dietary ingredients, mixed thoroughly enough to prevent separation and sorting, formulated to specific nutrient concentrations and offered ad libitum. The TMR system has been reviewed by Coppock et al. (1981) and by Muller (1992). The latter writer noted that both research and farmer experiences have shown that from 1000 to 2000 lbs or more increases in milk production per cow per year when herds were switched to a well formulated TMR.

Computer based least-cost ration formulation programs became widely used in the late 1960s and 1970s (Howard et al., 1968). One form of evolution of the least-cost formulations has been the Cornell Net Carbohydrate and Protein System (CNCPS) which uses both carbohydrate and protein fractions that are partitioned according to their ease and speed of degradation in the rumen (Barry et al., 1994). More recently this system has evolved to the Cornell/Penn/Miner Model . The application of this model requires an extensive array of nutrient analyses which if obtained by laboratory analyses on all ingredients used, makes its use relatively expensive but the formulation can be very precise and comprehensive.

Within the past 10 years there has been an intense interest in the nutritional management of the transition cow, defined as the time period 3 weeks before and the 3 weeks following parturition (Grummer, 1998). Initially, this concern related to the need to reduce the occurrence of milk fever, but its application has broadened upon the realization that adjusting the cation to anion relationship (Na + K) – (Cl + S) expressed as millequivalents per 100 grams of diet dry matter, in the prepartum ration would not only reduce the incidence of milk fever, but would maintain a much higher consumption of dry matter (DM) through the transition period which stimulates greater milk yields in the postpartum. This dietary adjustment usually means limiting potassium in the diet and feeding anion salts during 3 wk prepartum, so that the cation to anion relationship is slightly negative which results in urine with a negative pH. In addition to a reduction in milk fever, there is often a reduction in other metabolic disorders so that the benefits of this procedure postpartum are substantial.

Management changes which have increased the yield per cow. Though many of us are too young to know what it must have been like to live without the benefits of electricity, the arrival of this great technology must have been a momentous occasion for rural America. One writer noted that "the creation in 1935 of the Rural Electrification Administration (REA) did more to bring farmers into the 20^th century than any other single act. Thanks to the REA, nine of 10 farms were electrified by 1950, compared to one out of 10 in 1935" (Anonymous, 1999b).

Record systems in dairy cow management were pioneered by the DHI system which also first used mainframe computers in the 1950s to perform the many calculations necessary. As personal computers became available, in the early 1980s the DHI program developed the DART (Direct Access to Records by Telephone) program which allowed dairy producers to obtain record information on their cows quickly and to generate reports which were highly useful for management purposes. More recently software programs have been developed by commercial vendors which provide many useful management reports and calculations. Advanced milking systems have evolved, which with large milking parlors such as double 20s to double 30s, permit large herds to be milked 3 or even 4 times per day. Coupled with electronic identification and computer applications, not only milk yields, but body temperatures of cows and their somatic cell counts as well as other variables can be monitored daily if desired.

Programmed herd health refers to a prescribed or contractual relationship between a dairy producer and a veterinarian with clearly defined protocols for routine examinations, vaccinations and maintenance of the health of the herd. This usually means regularly scheduled visits with responsibilities and procedures clearly defined concerning the use, the storage of drugs, and the treatment of animals.

For those herds especially in the south, but to cows everywhere, another dimension to higher milk yields per cow is that the greater the feed intake, the greater the metabolic heat production. As Dennis Armstrong says, "a cow is a little furnace". And the greater the feed intake, the hotter the furnace. The primary reason for a decline in milk production during hot weather is a voluntary reduction by the cow in feed intake. So dairy producers in warmer areas of the U.S. are responding with the purchase of expensive heat stress abatement technologies including shades, fans, misters and evaporative cooling.

The commercial use of exogenous bovine somatotrophin (BST) was approved early in 1994. One survey showed that dairymen were getting an average increase in milk yield of 10.5 lbs per cow per day from its use. With a product cost of about $0.40 per cow per day, and increased feed required of about 4.5 lbs and some additional labor, the economic return has been highly favorable. It was estimated late in 1999 that 30% of the cows in the U.S. are in herds where BST is being used (Anonymous, 1999c).

Dairy farm numbers and herd sizes. The increased production per cow combined with powerful and compelling economies of scale are responsible for the large increases in average herd size and the decrease in the number of herds (Table 1). Except for 1998, the values for number of herds (or dairy farm numbers) are inflated because USDA counts any farm with one or more dairy cows as a herd. In contrast, Olson (1999) tallies the number of dairy farms by surveying regulatory agencies to find the number of farms which are actually selling milk. For 1999, he obtained 87,669, whereas the USDA number was 116,430. Unfortunately, Olson (1999) has data back only to 1992. If one uses a current estimate of total dairy cow numbers of 9.1 million and 87,669 farms, then the average herd size would be 103.8. Although the herd numbers are inflated from 1990 back, a very large increase in cows/herd has occurred since 1900. Even in traditional dairy states such as Wisconsin and New York, there are a rather large number of herds in the 500 to 1000+ cow range.

Shift in the location of cows. As noted recently (Staff Writer, 1999) there has been a major shift in the number of cows from traditional dairy states to the west. This writer pointed out that now there is a rather clear dichotomy in the U.S. dairy industry, the West and the Non-West. Although there has been a consistent decline in total dairy cows, nationally since 1990 there has been an increase in the West. Although the total country lost 969,000 cows since 1990, a 9.6% decline, the West gained 412,000 head, a 13.9% increase. The largest increase was in the state of California, which now leads the nation in total dairy cows, as well as in total milk production. This article further notes that in the West, the average herd size was 300 cows in 1998, whereas in the Non-west it was 72 cows.

Total milk production. Despite a decline in per capita production of milk from about 829 lbs in 1940 to 570 lbs in 1970 and a slight upturn to 584 lbs in 1997, the large increase in U.S. population from 76.2 million in 1900 to 151.3 million in 1950 to an estimated 270 million in 1999, has allowed total milk production to increase from 62.5 billion lbs in 1900 to 116 billion lbs in 1950 to 157.4 billion lbs in 1998 (Anonymous 1999d and Table 1.).

Prices received. The returns for milk by dairy producers are in Table 1. Relative to current total economic costs of production which range from $12.11 to $19.18/cwt as estimated by USDA (1999d), a costs and returns comparison shows that many dairy producers are losing money and such has probably been the case for much of this century. This can be deduced from the long decline in number of dairy farms and the large capacity of the industry to produce more milk than can be profitably sold as fluid milk and dairy products. Fewer and fewer producers seem willing to subsidize their dairy enterprises with their own and family labor. Consequently, the trend toward larger dairies with their economies of scale and fewer dairy farms is expected to continue.

Per Capita Consumption of Dairy Products.

Per capita consumption (lbs/yr) of total milk equivalent declined from 819 in 1930 to 563 in 1970, but has increased slightly to 582 lbs in 1998 (Table 4). Since 1970 the decline in total fluid milk/cream and butter consumption has been completely off-set by the large increase in cheese consumption from 11.4 lbs in 1970 to 28.4 lbs in 1998 (Table 4 ) which has been of enormous benefit to dairy producers. These trends are expected to continue at least for the near future.

Table 4. Per capita consumption of selected dairy products from 1910 to 1998.

1910 1930 1950 1970 1990 1998

Total milk

Equivalent 759.0 819.0 744.1 563.8 568.4 582.3

Fluid milk

and cream 121.7 174.6 259.1 264.4 231.3 217.9

Butter 18.4 17.6 10.9 5.4 4.4 4.2

Ice cream 1.9 9.7 17.4 17.8 15.8 16.6

Total cheese 4.3 4.7 7.7 11.4 24.6 28.4

Marketing Changes.

In 1922 passage of the Capper-Volstead Act encouraged the formation of farmer co-operatives and many dairy co-operatives were begun. These co-ops have played a major role in the wholesale marketing of fluid milk. But a large consolidation has occurred since 1950 when there were 2,072 co-ops, but only 230 in 1998 (Cropp, 1999). However the percentage of milk marketed by dairy co-ops increased from 53 to 87% during this 48 yr interval. The largest 50 of these co-ops now market about 81% of the milk (Cropp, 1999).

In 1984 a mandatory check-off of $0.15 cwt of milk began which was designated for promotion. Up to $0.10 of this amount could be designated for local advertising agencies and at least $0.05 was to go to national agencies. So since 1984, all dairy farmers have contributed to the promotion of milk and dairy products.

In the 1980s there began several programs by milk co-ops and milk handlers to offer economic incentives for indicators of milk quality including low somatic cell counts, low bacterial counts, etc. These incentives, in conjunction with in-line coolers and other equipment, have resulted in significant improvements in milk quality especially during the most recent 20 years.

Over the span of the past 100 years, the U. S. dairy industry has seen a nearly 6-fold increase in the average yield per cow, a truly remarkable biological phenomenon. This large increase can be partitioned into about 1/3 from genetic gain, and 2/3 from better nutrition and a wide array of improved management skills and technologies. This increased yield/cow in conjunction with potent economies of scale have resulted in many smaller dairies leaving the industry. The strong capacity to produce milk has usually exceeded the industry’s ability to market its primary product milk, at a price that covered all production costs.

There has been a large movement of cows from the traditional dairy states in the north and east to the west, especially to California, now the number one state in both total milk and number of dairy cows. Although large herds in the range of 500 to 1000+ cows are now present in all major dairy states, they are especially prevalent in the west. A recent veterinarian writer noted that "Veterinary services remain in demand and the future appears bright for dairy practitioners" (Herrick, 1999).

Per capita production and consumption of total milk equivalent has been rather stable since 1970 with the decline in whole milk consumption countered by an increase in low fat and skim milk, but especially by cheese which is now eaten at nearly 3 times the rate as 30 years ago. A more imaginative, vigorous marketing of milk and dairy products holds great promise for the dairy industry. Producers now recognize that selling and marketing is their job also, and they seem committed to that effort. New flavors, more convenient packaging and aggressive advertising bode well for increased sales of dairy products. Recent consolidation of dairy marketing co-operatives provides greater strength and holds much promise for dairy producers who can now play a more active role in the merchandising of milk and dairy products

Anonymous. 1999a. National Assn. of Animal Breeders, P.O. Box 1033, Columbia, MO 65205-1033.

Anonymous. 1999b. United States, history of agricultural recovery, Encyclopedia Britannica online. http://www.eb.com

Anonymous. 1999c. Status Update: POSILAC bovine somatotropin. May 11.

Web site: http://www.monsanto.com/dairy/press4.

Anonymous. 1999d. USDA data from National Agricultural Statistical Service (NASS), Economic Research Service (ERS), Agricultural Marketing Service (AMS) and The National Agricultural Library (NAL), Beltsville, MD.

Barry, M. C., D. G. Fox, T. P. Tylutki, A. N. Pell, J. D. O’Connor, C. J. Sniffen and W. Chalupa. 1994. A manual for using the Cornell Net Carbohydrate and Protein System for evaluating cattle diets. Revised for CNCPS Release 3, Sept. 1, Cornell University 14853-4801.

Bath, D. L., F. N. Dickinson, H. A. Tucker, and R. D. Appleman. 1985. Dairy Cattle: Principles, Practices, Problems, Profits. 3^rd Ed., Lea & Febiger. Philadelphia

Coppock, C. E., D. L. Bath, and B. Harris, Jr. 1981. From feeding to feeding systems. J. Dairy Sci. 64:1230-1249.

Cropp, B. 1999. Dairy co-op consolidation continues. Hoard’s Dairyman, 144: No. 17, p. 679. Oct.10..

Doak, G. A. 1999. Personal Communication. National Assn. of Animal Breeders, P.O. Box 1033, Columbia, MO 65205-1033.

Everett, R. W. 1999. Dairy genetics, Cornell University. Web site: http://www.ansci.cornell.edu/abc/dairy.html

Grummer, R. R. 1998. Transition cow energy, protein nutrition examined. Feedstuffs, 70:11. Sept. 14.

`

Henderson, C. R. 1973. Comparison of alternative sire evaluation methods. J. Anim. Sci; 40:760.

Herrick, J. B. 1999. Dairy practice looks bright. Large Animal Practice, 20: No.6. p.20.

Howard, W. T., J. L. Albright, M. P. Cunningham, R. B. Harrington, and C. H. Noller. 1968. Least-cost complete rations for dairy cows. J. Dairy Sci. 51:595.

Majeskie, J. L. 1996. Status of United States Dairy Cattle. Nat’l. Co-op. Dairy Herd Inprove. Prog. Handbook, Fact Sheet K-7. (Plus up-dated tables for 1996 and 1997.

Muller, L. D. 1992. Feeding management strategies. In Large Dairy Herd Management – 1992, p.326-335. Edited by H. H. Van Horn and C. J. Wilcox. Management Services, ADSA, Champaign, IL.

Olson, K. E. 1999. Dairy farm numbers down 4.2 percent to 87,669. Hoard’s Dairyman 144: 753. Oct. 25.

Palmquist, D. L., 1986. Supplemental fats in ruminant diets—lactating dairy cows. In: Proc. Southwest Nutr. Conf., p.52, Tempe, AZ.

Sirois, P. 1999. Personal Communication. Dairy One DHI Forage Laboratory, 730 Warren Rd., Ithaca, NY 14850.

Staff Writer. 1999. By the numbers. Hoard’s West, 144: W-12. Sept. 25.

Walton, R. E. 1983. A glimpse at dairying in the year 2000. Dairy Science Handbook, 15:511.

Table 2. Important events in the U.S. Dairy Industry from 1900 to 2000

• 1905 – Production testing (DHIA) was introduced in the United States*.
• 1908 – Chicago was first to require pasteurization to all milk unless from tuberculin tested herds*.
• 1914 – Tank trucks were first used to transport milk*.
• 1914 – The Smith-Lever Act established the Federal Extension Service.
• 1919 – Homogenized milk was sold in Torrington, Connecticut*.
• 1920s – Classified pricing of milk in all major markets was begun*.
• 1922 – The Capper-Volstead Act encouraged the formation of farm cooperatives.
• 1935 – The Rural Electrification Administration (REA) began bringing electricity to rural America.
• 1937 – The Agricultural Marketing Agreement Act established the Federal Order System of milk marketing.
• 1938 – Replacement of milk cans by bulk tanks began on dairy farms*.
• 1938 – The era of artificial insemination (A.I.) began.
• 1948 – UHT (ultra-high temperature) pasteurization of milk started*.
• 1950s – A consumer trend away from whole milk and high fat products began.
• 1950s – Sire proofs identified genetically superior sires with a resulting sharp increase in the use of A. I.
• 1968 – Electronic testing of milk fat became official and available commercially*.
• 1974 – Labeling of the nutritional content of fluid milk products began*.
• 1970s – The TMR (totally mixed ration) system of feeding became widespread.
• 1970s – Programmed herd health became an important management tool.
• 1980s – Powerful and compelling economies of scale resulted in rapid increases in herd sizes even in traditional northern and eastern dairy states.
• 1980s -- Electronic testing of milk for components other than fat was adopted*.

.

• 1980s – Embryo transfer was improved and adopted by breeders.
• 1980s – Somatic cell counts in milk were used to monitor subclinical mastitis and economic incentives for several indicators of milk quality were offered.
• 1980s – Component testing of milk and payment plans which reflected protein or solids-not–fat content in addition to milk fat were begun
• 1980s – Advanced milking systems and large milking parlors were designed for high-yielding cows which also allowed for low somatic cell counts.
• 1980s – A resurging interest and use of intensive grazing systems was seen.
• 1984 – A mandatory check-off of $0.15 per cwt was begun for dairy promotion.
• 1980s – Dairy records management programs became available through Dairy Records Processing Centers, e. g., DART (Direct Access to Records by Telephone) which was followed by computers on dairies used with commercial software.
• 1980s – The need for waste nutrient management programs and concern for water quality became more widely appreciated and plans were adopted.
• 1986 – The Dairy Termination Program (Dairy Buyout) began.
• 1980s – Producers became more aware of the need for heat stress abatement technologies which became more prevalent especially in the south and west.
• 1990s – A much greater use of comprehensive rearing programs for dairy heifers has occurred.
• 1990s – A large emphasis on the management of the transition cow was seen.
• 1990s – Comprehensive ration formulation models were developed and used.
• 1994 – Exogenous bovine somatotrophin was approved for commercial use.
• 1990s – There was a large shift of dairy cows with their resulting milk production away from traditional eastern and northern dairy states to western states.
• 1990s – A large consolidation in the number of dairy co-ops occurred.
• 1990s –A new emphasis on dairy farm labor management, training and compensation occurred.