Genetics Research at Fort Keogh:
S. Newman, Geneticist
"At what point in the history of mankind did human beings first become aware of problems of heredity and of the transmission of features and characters from one generation to another? We do not know. We can assume, however, that conscious or unconscious selection of characters first occurred during the prehistoric period of transition from hunting and food gathering to tillage and animal husbandry." Stubbe, History of Genetics, 1972.
Foundations of modern livestock improvement extend back to the mid-1700's, when Robert Bakewell began his animal breeding work at Dishley, England. He had definite "ideals" for selection of beef cattle, e.g., a low-set, blocky, quick-maturing animal almost unknown at that time. He bred best to best regardless of relationship, which often meant rather close inbreeding.
Modern animal breeding is built upon the work of Bakewell in the sense that: 1) it is of primary importance to define a goal or objective in a breeding program; and 2) measures of genetic merit like expected progeny differences (EPD's) account for all performance information from relatives. Since its inception in 1924, Fort Keogh Livestock and Range Research Laboratory (LARRL) has been a leader in development of tools for genetic improvement that enhance the accuracy with which we choose superior breeding stock. Beef cattle have been emphasized to a greater extent than other livestock species at Fort Keogh. Recurring themes in research over the last 70 years include combined effects of inbreeding and selection on performance in turkeys, swine, and beef cattle; foundations of performance testing in beef cattle; and crossbreeding in swine and beef cattle.
Although LARRL is expressly involved in research with beef cattle at the present time, that was not the case at its inception. In 1927, the animal inventory for LARRL included:
"... A purebred Belgian Stallion with 49 pure-bred and draft horses, one purebred Morgan stallion, and one thoroughbred stallion with 26 purebred Morgan mares and 46 grade mares and geldings of saddle type; 250 purebred Hereford cattle; 2,000 grade and purebred Rambouillet sheep; and 44 head of breeding hogs representing Duroc Jersey, Chester White and Yorkshire Breeds."
Turkeys were brought to LARRL in 1929. A Milking Shorthorn dairy herd was maintained beginning in the early 1930's. The milk was sold to the employees, but the animals were not used extensively for research purposes. The herd was dispersed in the late 1930's.
What follows is a brief description of research pertaining to genetic improvement in the various species of livestock that maintained residence at LARRL. Readers are encouraged to inquire about printed records of these investigations, many of which are maintained in the Fort Keogh Library.
Turkey Improvement Programs
Studies with Bronze turkeys started in 1929 and involved approximately 1500 young turkeys and 350 breeding hens. Studies consisted of feeding, breeding and rearing experiments. The only breeding research alluded to in a 1935 Research Report stated that:
"Turkey breeding experiments resulted in a simple, easily operated plan for maintaining the average egg weight and improving the fertility and hatchability of eggs and egg production."
An article that appeared in the 1931 Yearbook of Agriculture reported that these improvements were due to both modern methods of sanitation and to genetic improvement. A project to study effects of inbreeding on genetic improvement of turkeys began in 1931 and was terminated in 1935. Outbred (non-inbred) matings were maintained to affect adequate comparison of the effects of inbreeding. General conclusions were: 1) mild (inbreeding = .125 to .218) and close (.25 to .411) inbreeding had little effect on fertility, egg production, and egg weight when compared to outbred matings; 2) mild and close inbreeding had an adverse effect on hatchability; and 3) intense inbreeding (.5 to .672) affected fertility and hatchability of eggs but had a slight effect on production and average egg weight.
Turkey research was terminated in 1939 when the turkeys were shipped to ARS Headquarters in Beltsville, Maryland. The original crosses produced at Fort Keogh led to development of the Beltsville White breed, the first meat-type turkey breed developed in the US.
Sheep Improvement Programs
Sheep research at Fort Keogh centered around nutrition studies such as winter supplementation, pre- and post-lambing feeding regimes, and alternative protein and energy supplementation. Individual measures of fleece weights, fineness scores, character, staple length, and density were recorded to determine relations between methods of feeding and management on increasing the yield and quality of wool. There is no record of sheep breeding experiments taking place at LARRL. The only mention of genetic improvement came in a 1935 research report:
"During the 11 years these sheep investigations were in progress at the U.S. Range Livestock Experiment Station, all rams used in the breeding work were bred by the Bureau of Animal Industry and furnished from the U.S. Sheep Experiment Station, Dubois, Idaho. The Rambouillet rams were all purebreds and were selected for freedom from excess skin folds and for heavy fleeces of wool having a good length of staple. As a result of several years of progeny testing and selective breeding, there was developed a band of uniform, smooth-bodied Rambouillets, especially useful for breeding experiments and grazing investigations."
All sheep at LARRL were transferred to the U.S. Sheep Experiment Station in 1941.
Equine Improvement Programs
In 1912, Fort Keogh became a remount post for the U.S. Army. More horses were processed through Fort Keogh than any other post in the country. In 1925, after establishment of the Range Livestock Experiment Station, the second Government Morgan Horse breeding program was founded (the first was established in Middlebury, Vermont). Between 1925 and 1935, LARRL produced 93 purebred Morgans, 77 Morgan-grade crosses, and one Morgan-draft horse. Draft horses and thoroughbreds were raised at the station as well. Morgan x thoroughbred crosses were considered the "ultimate cow horse" at the time because it combined the stronger legs of the thoroughbred (for covering long distances in a day) with the easy upkeep of the Morgans, which didn't tire as quickly.
Horses were primarily used at LARRL to perform the necessary work at the Station. Thus, experimental studies were made to fit into this situation. Much of the early pioneering work in development of semen collection and artificial insemination techniques took place at Fort Keogh until the early 1940's.
The breeding of draft horses at LARRL amply demonstrated the importance of selecting suitable sires in producing horses adapted for use on the Northern Great Plains. For example, a research summary from 1945 reported that:
"The Belgian sire 'Rowdy' showed unusual ability in transmitting to his progeny (a) uniformity; (b) desirable temperament; (c) draft confirmation; (d) desired shape and soundness of feet and legs; and (e) straight balanced action even though the mares used differed greatly in type, breeding, and quality."
The Morgan stud was slowly dispersed after 1935. It is not known what happened to the Belgian horses. The thoroughbred breeding herd was maintained until 1964. The Laboratory now has a herd of about 20 horses that are used exclusively for working cattle.
Swine Improvement Programs
Experimental work with swine at LARRL began in 1925. Initial studies were nutrition-based (e.g., barley vs corn for fattening pigs). Animals used in these studies were crosses between Yorkshires and Chester Whites. In late fall of 1935, the entire swine herd was liquidated and in 1936 foundation herds of Danish Landrace and black (beltless) Hampshires were established with animals from the Agricultural Experiment Station, Beltsville, Maryland. Thus began one of the first composite breed development studies in the U.S., as the objective of the study was to "...combine genetically the best traits of each of these breeds into a single strain of the same color as the black parent strain of the Hampshires." Different intensities of inbreeding, in combination with selection of progenywere used to "fix characters such as prolificacy, economy of production and conformation."
Landrace white color proved to be dominant over the Hampshire black. Thus, F1 animals were inter-mated and backcrossed to Hampshire to produce black pigs for future generations. A recessive gene for spotting was detected in 1945. It is interesting to note that no selection for performance was begun until the spotting factor was eliminated although initial estimates of crossbred superiority proved promising. Due to limited resources, breeding herds of this black strain were established at Montana State University, Oregon State University, and the University of Nebraska. In spring of 1947, the name of "Hamprace" was adopted for this new breed of hogs. The Inbred Livestock Registry Association renamed this breed Montana No. 1 in 1948 from its geographic origin and numerical order of adoption as a breed.
A 1957 USDA Bulletin (no. 528) reported development, characteristics, performance, and distribution of Montana No. 1. Six one-sire lines and one two-sire line were developed between 1936 and 1954. Due to high inbreeding, some lines were lost. Inter-crossing of remaining lines occurred in 1951. Cumulative selection differentials (the annual superiority of selected individuals over those available for selection) were .34 for number of pigs farrowed per litter, .42 for number of pigs weaned per litter, 13.8 lbs for litter weight at weaning, and 1.8 lbs for individual weight at weaning. Feed lot performance (record-of-performance) was measured from 1942 to establish postweaning gain differences. Slaughter data was also collected. Live backfat measurements were first taken in 1956.
In 1953, an experiment was initiated to test the usefulness of reciprocal recurrent selection (RRS), a breeding program that allowed the selection of boars based upon combined purebred and crossbred performance. This type of breeding system had been used successfully by corn breeders in the development of improved hybrid corn varieties. In RRS, randomly selected individuals from two lines or breeds are progeny tested in crosses with each other. Those individuals of each breed having the best cross progeny were then inter-mated to perpetuate their respective breeds. Offspring from these within-strain matings were again progeny tested in crosses with the other and the cycle repeated. Montana No. 1 and Yorkshire breeds took part in this test. The project ended in 1963. General conclusions were that there was some decrease in fertility and increase in growth rate in Montana No. 1 due to RRS. The two cross-line groups were intermediate to the inbred lines for most traits studied.
Federal funding for swine research at LARRL terminated in 1968. Work involving Montana No. 1 and the Yorkshire breeds was terminated in 1971 and a crossbred herd was established to supply animals for studies directed by MSU nutritionists. The last of the swine research unit was moved to Bozeman in 1986.
Beef Cattle Improvement Programs
According to a 1927 research report:
"A herd of registered Hereford cattle was purchased by the State of Montana on November 1, 1924, from George Miles and Sons, Miles City, as a foundation for range beef cattle operations. This herd comprised 96 breeding cows, 19 yearling heifers, 69 calves, and 4 herd bulls."
During this period, corn breeders were achieving great success in developing hybrid corn varieties that out-produced their inbred parents. In livestock, crossbreeding was the antithesis of acceptable production practices. Thus, the objective of initial research was the development of inbred lines of Hereford cattle particularly adapted to range conditions. Selection within each line was based upon performance rather than subjective visual standards. Fertility, rapid and economical gain, conformation, and high dressing percentages were the major traits of importance in the breeding programs. Development of these lines began at LARRL in 1934. By 1945, 3 lines were either established or just underway; in total 14 lines were developed by 1957.
Linebreeding program. Beginning in 1935, methods of genetic evaluation of beef cattle were pioneered at LARRL. All beef performance testing programs now active in the U.S. and much of the remainder of the world are built on this foundation. The concept of progeny testing steers of promising sires and lines of breeding was established. Individuals were stall fed postweaning. Progeny groups differed in weaning weight, postweaning gain, feed efficiency, and dressing percentage! This led, in 1946, to computing the first heritability estimates for performance traits in beef cattle, and established the separation of the influence of heredity and environment on performance. In general, it was concluded that much improvement could be expected from selection for growth rate, weight for age, and carcass quality, but only little improvement may be expected from selection for higher dressing percentage. As early as 1954, researchers at LARRL were recommending use of low birth weight bulls on yearling heifers, based upon studies of genetic variability (differences in sire groups) of birth weight. The determination that traits of economic importance were rather highly heritable changed beef cattle selection from visual appraisal to performance standards and affected cattle breeding more than any other event in its history.
Use of ultrasound for live animal evaluation of fat and lean tissue began in 1960 when LARRL embarked on a study to improve beef carcasses through more accurate live animal evaluation. Bulls were chosen from feedlot tests based upon a selection index that included weight for day of age and fat thickness. Remaining yearling bulls were slaughtered for carcass appraisal. The conclusion was that selection index was ineffective in applying selection pressure against fat thickness, although interpretation of the data was limited due to the small numbers (387) of animals sampled.
All inbred lines developed at LARRL other than Lines 1, 2, and 3 were established by purchasing related bulls and heifers from individual herds. Each line was maintained as a closed herd with no outside introductions. Sire selection was based almost exclusively on performance testing for postweaning growth. By 1962, six of 11 lines had been culled for poor growth or other problems. At this time, line crosses were made among five of the lines, based upon success of previous crosses, which showed line-cross progeny expressing heterosis through growth traits. Of the 14 lines that were eventually produced, only Line 1 exists today, and is undoubtedly the longest-running beef cattle selection experiment in the world. In a recent publication (MacNeil et al., 1992. J. Anim. Sci. 70:723-733), data from 1934 to 1989 was summarized in terms of estimated breeding values (EBV's; expected progeny differences, EPD's are one-half the EBV), which takes into account performance and pedigrees of all animals used in the line. For example, the graph below displays the genetic trend in yearling weight in Line 1. It is obvious that there has been a steady increase in yearling weight since 1934.
Figure 1. Genetic trend in yearling weight of Line 1 Hereford cattle at Fort Keogh.
The building blocks of performance recording that were established so early in the beef cattle industry obviously have had a profound influence on genetic improvement of all livestock. This could never have been achieved were it not for the foresight of past researchers, who developed and refined performance recording techniques currently with the development of inbred lines of cattle. It is interesting to note that by 1983, almost 68% of all purebred Hereford calves produced in the U.S. have some Line 1 breeding in their pedigrees.
In 1961, researchers at LARRL were interested in knowing to what extent adaptation to a specific location or environment influenced the productivity and efficiency of beef cattle production. Thus, a study of genotype x environment interaction was begun between LARRL and the Subtropical Research Station at Brookesville, Florida, with the shipment of 65 Line 1 cattle. The herd was treated as a closed line at each location. A herd originating in Florida was also formed where females were mated to bulls of that line. One-half of the Florida herd was transferred to Montana concurrently with the shipment of Line 1 cattle to Florida. Thus, two herds existed that were selected for performance at both locations. It was concluded that cattle of Montana origin performed best in Montana and cattle of Florida origin performed best in Florida. Thus, each group of cattle was better suited genetically to their location of origin. The project ended about 1974.
From 1976 to 1988, a herd of cross-line Herefords was managed under a random selection procedure (the Selection Criteria Study, or SCS). The purpose was to examine as many growth and reproduction measurements as possible to determine direct and correlated effects of selection. This was unique in that most herds where genetic parameters were estimated had some previous history of selection, thereby potentially biassing the estimates obtained. With new techniques of variance component estimation, selection history can be accounted for when analyzing such data. Data from this experiment are now being analyzed using these new and more accurate and unbiased techniques.
In 1978, a new selection experiment began with Line 1 addressing concerns among beef producers over calving difficulty. The herd was "sub-lined" into two herds. Replacement sires in the Y sub-line were selected for high yearling weight. In the YB subline, independent culling levels are used where average or below average birth weight of sires is an additional selection criterion. This project ends with information collected from the 1993 calf crop.
Crossbreeding program. Breeding experiments involving a three-way cross of Hereford, Shorthorn, and Aberdeen Angus cattle were started in the 1938 breeding season when two groups of randomly selected Hereford cows were bred to Shorthorn bulls. This continued through 1939; Angus bulls were used on crossbred females from 1941 to 1942. Preliminary results showed that crossbred animals exceeded purebred animals in growth and carcass merit, and that three-way cross animals exceeded F1's. It is interesting to note that in a 1944 research report it was stated, "The use of crossbreeding has been criticized because of the belief that it would not be advisable to use the crossbred females for calf production." By 1959, performance testing had also included Charolais sires. Performance testing of steers included Charolais, Brown Swiss, and Brown Swiss x Hereford.
In 1961, a long-term crossbreeding project began by mating Hereford, Angus, and Charolais sires to Hereford, Angus, Charolais, and Brown Swiss dams. The first progeny from this phase were born in 1962 and the last in 1965. Rotational crossing among these breeds occurred at this time. The project ended in 1976. General conclusions supported the notion that two- and three-way rotational crossing schemes were useful in that they provided their own crossbred female replacements. The three-way system produced higher levels of heterosis than two-way.
With an increased influx of new breeds to the U.S., LARRL instituted a new project in 1974 to evaluate Angus, Red Poll, Pinzgauer, and Simmental breeds. Again, this project included an objective related to genotype x environment interaction. Sires used at Miles City were also used in the Germplasm Evaluation program at the U.S. Meat Animal Research Center, Clay Center, Nebraska. These breeds were chosen based on size of sire breed (medium or large) and level of milk production (medium or high) to provide a more general level of applicability of the results than inference to the breeds alone would allow. In 1977, a group of Hereford x Tarentaise females was added, and Tarentaise sires were included as a sire breed.
In 1979, a project was started at LARRL to develop a composite of beef breeds adapted to the Northern Great Plains. Red Angus females were mated to either Charolais or Tarentaise sires. Subsequent crossing of foundation animals produced a composite population that is ? Red Angus, ? Charolais, and ? Tarentaise. Analysis of the population to 1990 showed the herd to be well suited to the range environment, with high pregnancy and calving rates, fast gains, and acceptable carcass quality. In 1988, three lines were formed, reflecting two different methods of selection for biological efficiency plus a control line. In one line, bulls are selected on the index I = yearling weight - 3.2(birth weight). This index accounts for potential losses due to calving difficulty from heavier birth weights of calves, while in theory minimizing effects on growth. In the second line, bulls are selected on an index I = weaning weight/cow weight, resulting in a measure of production per unit of cow size. This index has been advocated as a measure of biological efficiency in terms of maternal productivity, assuming annual feed requirement is related to cow weight. The final calf crop of this study will be in 1994.
Recent collaborations with groups in Canada and New Zealand have led to a greater understanding of the development of breeding objectives and selection indices that optimize profit of beef cattle production. Research continues in testing the accuracy of genetic evaluation procedures, investigating sources of bias (e.g., inbreeding) in EPD's, understanding genetic antagonisms between biological traits and their effects on multiple-trait selection, and exploiting dominance effects in breeding programs. Just as researchers at LARRL first developed the foundations of performance recording for genetic evaluation, LARRL geneticists are cooperating with other locations in defining variability of DNA sequences that may lead to eventual discovery of major genes affecting traits of economic importance. The genetics program at Fort Keogh will remain a vital force in the development and application of techniques that improve beef cattle production.
Key Individuals Associated With the Genetics Program at Fort Keogh
This Historical Perspective was published as part of the 1993 Field Day Report and has not been updated.
USDA, ARS Fort Keogh Livestock and Range Research Laboratory