THE WHY , HOW-TO , AND COST OF PROGRAMED AI BREEDING OF DAIRY COWS

Management of the estrous cycle is now more practical than it was a decade ago because of our understanding of follicular waves. With availability of three gonadotropin-releasing hormone (GnRH) products and two prostaglandin products, the cycle can be controlled for fixed-time inseminations with little loss in conception rate compared to inseminations after detected estrus. Various systems are effective for programming first inseminations with or without some heat detection. With the incorporation of transrectal ultrasonography for early pregnancy diagnosis 28 to 30 days after insemination, routine heat detection programs could be eliminated by reprogramming each cow after an open diagnosis. The most limiting factor in the control of the cycle is the proportion of missed heats in estrus-synchronization programs that rely partly or solely on heat detection. Pregnancy rate (the proportion of cows that become pregnant of all cows programmed for insemination) is the best measure of an estrussynchronization program, because it measures total number of pregnancies achieved per unit of time rather than simple conception success at any given insemination.; Dairy Day, 1998, Kansas State University, Manhattan, KS, 1998;


Introduction
Improving dairy herd reproductive management requires an understanding of the basic principles of getting cows pregnant. It is critical to understand each component of the estrous cycle as well as the annual reproductive cycle (calving interval) and determine where limited time and resources might be best concentrated to reach A.I.-breeding goals. A calving interval consists of four major components. The first component is the rest period or elective waiting period (EWP). The duration of this period is partly a management decision. This period varies from 40 to 70 days on most farms. Part of its duration is based on the physiological need of the cow, in which the reproductive tract must undergo an involution process (return to its nonpregnant size and function). Research indicates that when cows calve without complication, this healing process requires no more than 40 days. This process includes macro-and microscopic processes that prepare the uterus for another pregnancy.
The second component is the period of time between the end of the EWP and when the first estrus is detected for the first AI breeding. The duration of this period is a function of the heat detection rate as well as whether or not some hormonal regimen is used to bring cows into estrus after the end of the EWP (e.g., PGF 2" ). Whether or not PGF 2" is used to bring cows into estrus for first services, the percentage of cows detected in estrus depends on the rate of heat detection or the efficiency of detecting estrus in all cows.
The third component of a calving interval is the active AI breeding period for each cow and represents the number of days required for the cow to conceive after the first AI service. If a cow conceives at first service, then the third component is nonexistent. Otherwise, it is a function of the heat detection rate and the level of herd fertility. The level of herd fertility depends upon a number factors, including sire and cow fertility, correct thawing and handling of semen, AI breeding technique, and timing of insemination. Fertility and heat detection rates are very important to establishing pregnancy in a timely fashion.
The fourth component of a calving interval is gestation. The duration of gestation is fairly constant. It can't be shortened significantly without adversely affecting the health or viability of the newborn calf.
Based on these component parts of a calving interval, an EWP of 40 to 50 days is probably sufficient for essentially all cows. With a rate of heat detection of 65% and a conception rate of 65%, the average period from the end of the EWP until pregnancy is established in 95% of the cows should be 35 days. This means that some cows conceive immediately following the end of the EWP and others remain open for 100 or more days. With an EWP of 50 d, estrus and conception rates of 65%, and a gestation period of 280 d, an average calving interval of 365 days (50 + 35 + 280 = 365) is attainable, when it is desired that 95% of the cows conceive.

Follicular Growth during the Estrous Cycle
A follicle is similar to a fluid-filled water blister and contains the egg. The follicle is composed of an outer layer of cells (theca cells), which are exposed to blood capillaries. Blood delivers gonadotropic hormones (FSH and LH) from the anterior pituitary to the follicle, which stimulate its growth, production of gonadal steroid hormones, and growth and maturation of the egg. Inside the follicle, another group of cells (granulosa cells) surround a fluid-filled cavity that forms the antrum of the follicle. These cells take the androgen precursors (stimulated by LH) produced by the thecal cells and synthesize estrogen (stimulated by FSH). Deep in the antrum, surrounded by specialized granulosa cells, is found the microscopic egg cell. Hundreds of thousands of these follicles are found in the ovaries of the heifer at birth. Once she reaches puberty, these follicles grow in a cyclic fashion from diameters of <1 mm to ovulatory sizes of 16 to 18 mm in diameter .
For many years it has been known that as follicles grow, some eventually ovulate, whereas others become atretic (die). Earlier, it was thought that follicular growth was either bimodal or continuous. More recently, it was assumed that whatever follicle had reached ovulatory size at the right time during the cycle would be the one that would eventually ovulate. Although this concept is probably correct, it was based on the fact that at least one large follicle can be palpated in the ovaries on almost every day of the estrous cycle.
With the use of the real-time, B-mode ultrasonography, the same type of equipment used in hospitals by physicians to monitor development of human babies within the uterus of their mothers, we can examine the growth of follicles in cattle. This same technology is used to measure backfat and loin-eye areas in finishing cattle and pigs. The probe is inserted into the rectum with the gloved hand just above the reproductive tract as if the cow were palpated. Placement of the probe in this position allows visualization of the ovaries, uterine horns, and cervix. The probe emits ultrasound waves that are absorbed by fluid-filled cavities and appear on the viewing screen as images in various shades of grey or black. Follicles appear as round black circles, and the corpus luteum (CL) looks like a peppery elliptical structure.
Using this technology on a daily basis, several patterns of follicular growth have been described, along with new terminology to describe the dynamics of follicular growth. These terms were borrowed from similar studies performed in monkeys. Figure 1 shows the diameters of several follicles during the estrous cycle of a cow. Two groups or "waves" of follicles developed during the cycle. On days 1 and 2, four follicles were visualized, but only one continued to grow (dominant) from this group (cohort) and "dominated" the other (subordinate) smaller follicles. The subordinate follicles underwent atresia (death) and were no longer useful. The first dominant follicle underwent a growth phase (d 1 to 6), a static phase (d 7 to 9), and a regressing phase (d 11 to 12 or longer). The second wave of follicles visualized appeared around days 9 to 11, one of which dominated the other follicles and became the second dominant follicle that eventually ovulated after luteolysis (death of the CL).
Although any number of follicles can make up a wave of follicles, usually only one to six develop in a wave. The first wave and its dominant follicle always appear at the same time during the cycle in all cows. A "two-wave" cow has an estrous cycle of 21 days. Two-, three-, and four-wave cycles have been observed in cattle, with the appearance during the cycle of the second, third, or fourth wave being more variable than the first. Estrous cycles become longer with increasing number of follicular waves. Two-wave cycles are 19 to 20 d, and four-wave cycles tend to be 23 to 25 d in duration.
The growth of a group of follicles that make up a wave is initiated by a transient increase in blood FSH, which is observed 1 or 2 days before the beginning of each follicular wave. Estrogen in the blood also rises and falls with the growth and regression of a dominant follicle. The dominant follicle apparently dominates its subordinate peers by producing substances that inhibit their further growth.

Variation in the Interval to Estrus after PGF 2" "
When PGF 2" was being tested as an estrussynchronization hormone in cattle, a common endpoint to measure its success was the proportion of cows observed in heat during a 2-to 5-day period after injection. That period reflected the proportion of cows that had a functional CL secreting high blood concentrations of progesterone at injection time. Any cow coming into estrus much before 48 h most likely had natural or spontaneous luteal regression before the PGF 2" injection. These cows showing estrus before 48 hr were likely on days 19 to 21 of their cycles when PGF 2" was injected. Approximately 2 to 5 days after the injection, cows would come into estrus because blood progesterone would return to baseline concentrations within 12 to 24 hr, and the CL would no longer be functional. Interestingly, regardless of how soon a cow came into estrus, concentrations of progesterone would return to baseline at nearly the same time.
We have learned that the variable part of this interval is the period of time during which the follicle matures and induces estrus by secreting high concentrations of estrogen. So it seems that interval to estrus after PGF 2" was not related to concentrations of progesterone during the estrous cycle but rather to the relative maturity of a developing follicle at the time of PGF 2" injection or luteal regression.
What would happen if PGF 2" were injected at various stages of the cycle? Intervals to estrus are dependent on the relative diameter (maturity) of the dominant follicle at the time of PGF 2" injection. Short, medium, and long intervals to estrus after PGF 2" are based on when PGF 2" is injected in the cycle. So if PGF 2" were given when either a first or second dominant follicle is quite mature (large in diameter), the interval to estrus would be much shorter than if PGF 2" were given at mid cycle between follicular waves or later in the cycle when the second dominant follicle is relatively larger in diameter.
Evidence exists for these different intervals based on studies conducted in dairy heifers when PGF 2" was given at various stages of the cycle. Assuming that most heifers have two follicular-wave cycles, then injections of PGF 2" at various phases of follicular maturity, whether given while the first or second dominant follicle was present, would produce the various intervals to follicular maturation, estrogen secretion, and the onset of estrus (Table 1).
Is conception rate affected when PGF 2" is given at different times? Apparently it is not, as long as inseminations were based on detected estrus. For example, if a PGF 2" injection is given on day 6 or 7, when the first dominant follicle is growing (Figure 1), the CL would regress and the first dominant follicle would ovulate and be normally fertile when AI breeding was performed after detected estrus. Similar results occur for any dominant follicle that is in its growing phase at the time of PGF 2" injection.

Pregnancy Rate
Several factors determine the number of pregnancies or the number of calves born (e.g., herd fertility, technician ability, sire fertility, and heat detection rate). One method to examine the success of the insemination program is to determine the number of cows that become pregnant during each 21-day period after the end of the EWP. This concept is suggested to be the best measuring stick for success of AI breeding.
The pregnancy rate (PR) equation can be simplified to two factors: heat detection rate (HDR) and conception rate (CR) or PR = HDR × CR (Table 2). Conception rate is determined by herd and sire fertility plus inseminator proficiency. Using the simplified method of determining pregnancy rate, one can evaluate the success of the AI breeding program, including whatever programmed breeding system is used to synchronize estrus for first services after calving. Let's assume a 50-day EWP. Because no cows are AI bred until after 50 days,

Heat Detection and Conception Rate
The downward slope of those generated curves (pregnancy rate) in Figure 2 are determined by the number of cows detected in estrus and their resulting conception rate after each AI breeding. In Table 3, a few examples of various heat detection and conception rates are shown to illustrate their effects on pregnancy rate during each 21-day period. In the first four examples, holding heat detection rate constant at 60%, conception rates were varied from 30 to 60% (low to high). The resulting pregnancy rates range from 18 to 36%, or in other words, 18 to 36% of the cows detected in estrus and inseminated became pregnant during each 21day period. In the last four examples in Table  3, conception rate was held constant at 50%, and heat detection rate varied from 40 to 70% (low to high). Resulting pregnancy rates ranged from 20 to 35%. The number of pregnancies achieved during each 21-day period, or after each additional estrous cycle, can be affected by various rates of conception and heat detection. For example, one herd could achieve a 24% pregnancy rate with better than average heat detection (60%) and an average conception rate (40%), whereas another herd could achieve a similar pregnancy rate (25%) with average heat detection (50%) and better than average conception rate (50%).

Percentage Open Curves
Using the first example (first four lines of data in Table 3), four curves are plotted in the Figure 2. The upper curve represents an 18% pregnancy rate for each 21 days. The remaining curves represent 24%, 30%, and 36% pregnancy rates. A horizontal line drawn across the graph at the 50 percentage open mark approximates average days open for each pregnancy rate curve. Using this method of evaluating the number of pregnancies established during each 21-day period after the end of the EWP, one can see how improvements in either heat detection rate, conception rate, or both can increase the number of pregnancies achieved in the AI breeding program. Because conception rate is determined easily, one also can estimate the herd heat detection rate by dividing the pregnancy rate for each 21-day period by the conception rate during that same period.

Programmed Breeding
Most dairy producers appreciate the benefits and advantages of using an estrus-synchronization program. Synchronizing estrus in cattle simply makes occurrence of estrus more predictable and AI breeding more convenient. Dairy producers have benefitted from the superior genetics of proven bulls, which have increased pride of ownership in better-bred cattle, as well as providing a pay off in greater milk production. Although most are sold on the idea of using heat synchronization, one question most frequently asked by dairy producers and dairy veterinarians is: What is the best way to synchronize estrus in dairy cows and heifers for AI breeding?
The program used successfully on dairy farms is probably the one that is the most simple to execute. Although heat synchronization of large numbers of cows and heifers is not typical on most dairy farms, except in large herds or where seasonal calving is practiced, one needs to develop a system for identifying cows (based on days after calving) and heifers (based on age) that should go into each breeding group cluster.
The breeding cluster is one method that can be used. For example, if the EWP is 50 days before AI breeding, then a breeding cluster of cows can be organized that falls within a certain range of days in milk to fit the targeted first breeding date. These cows can be identified easily from a breeding wheel, computer records, or by simply keeping a chronological list of calving dates. In our herd of 200 cows, we cluster cows that calve during a 3-wk period so that the freshest cow in the cluster meets the minimum acceptable EWP. When the EWP is 50 days, then a cluster would consist of cows that are 50 to 71 days in milk during the targeted breeding week. Therefore, the average interval to first insemination is 60 days for the herd. Cows that fail to conceive should return to estrus during the breeding week of the next cluster of cows, which would be estrus-synchronized for AI breeding 3 wk after the first cluster of cows. This clustering method allows first services and repeat inseminations to occur during the same week, thus concentrating most inseminations during 1 wk out of every 3 wk. This same system can be employed for AI breeding of replacement heifers when they reach an acceptable age and weight to enter a breeding cluster.
In larger herds, grouping cows into a 1-wk cluster is necessary. This 1-wk cluster simplifies AI breeding of cows that meet the breeding criteria on a weekly basis. Therefore, during the period before the cows reach their targeted breeding date (based on days in milk and the EWP), estrus is synchronized to occur during each breeding week. Usually, the synchronization period is set so estrus or fixed-time insemination will occur in the Monday-to-Friday work week.

Choosing a Breeding System
Once a system is in place to identify cows and heifers that fit those criteria for inclusion in an AI breeding cluster, then the specific programmed breeding system is fit into a weekly management sequence. What successful programs are available? There are two general categories of programs from which to choose: 1) PGF 2" ; or 2) gonado-tropin-releasing hormone (GnRH) + PGF 2" . The first involves using either of two prostaglandin products that are available in the U.S. market (Lutalyse® and Estrumate®). The second category uses either of three GnRH products (Cystorelin®, Fertagyl®, or Factrel®) plus a prostaglandin product in combination with heat detection or a fixed-time insemination.

Targeted® Breeding Program
The Targeted Breeding program has been promoted by one of the PGF 2" manufacturers (Pharmacia & Upjohn) for synchronizing the AI breeding of lactating cows in a herd (Figure 3). Injections of PGF 2" are administered 14 days apart. This interval is simply based on the fact that sufficient time must pass after the first injection so those females responding to the first injection (their CL regresses and they come into estrus) have a new CL that is mature enough to respond to a second injection (at least on day 6 of the estrous cycle). In addition, those females that were not in a stage of the estrous cycle with a CL that could regress after the first PGF 2" injection should be responsive 14 days later. Targeted Breeding calls for the first injection (so-called set-up injection) to be given 14 days before the EWP ends. No cows are inseminated after the first injection, although about 50% show estrus in response to the first injection. The second injection (first breeding injection) then is given just prior to the end of the EWP, so first services can occur when cows are eligible for AI breeding. The Targeted Breeding program then suggests that if no estrus is detected after the second injection, a third injection (second breeding injection) is given in another 14 days. If no standing estrus is detected after this third injection, then one fixedtime insemination can be given at 80 hr after this third injection of PGF 2" .

Ovsynch
The second method (named Ovsynch) is similar to the previous program, except it requires no heat detection (Figure 4). In fact, it is described more accurately as an ovulation synchronization program; hence the name, Ovsynch. An 100-µg injection of GnRH is given 7 days before a PGF 2" injection, and then a second 100-µg injection of GnRH is administered 36 to 48 hr after PGF 2" , with one fixedtime insemination given 8 to 20 hr later. (A recent study has found that 1 mL or 50 µg of Cystorelin is sufficient.) The first GnRH injection alters follicular growth by inducing ovulation of the largest follicle (dominant follicle) in the ovaries after the GnRH injection to form a new or additional CL. Thus, estrus usually does not occur until after a PGF 2" injection regresses the natural CL and the secondary CL (formed from the follicle induced to ovulate by GnRH). Therefore, a new group of follicles appears in the ovaries (based on transrectal ultrasonographic evidence) within 1 to 2 days after the first injection of GnRH. From that new group of follicles, a newly developed dominant follicle emerges, matures, and can ovulate after estrus is induced by PGF 2" , or it can be induced to ovulate after a second injection of GnRH. The GnRH injections release pituitary luteinizing hormone (LH), the natural ovulation-inducing hormone of the estrous cycle. Few cows will show heat in this program. About 8 to 16% may show heat around the time of the PGF 2" injection. If so, those cows should be AI bred according to the AM-PM rule and the second GnRH injection eliminated.
This program works in replacement heifers, but because of lower pregnancy rates than can be achieved with other programs, it is not recommended. For some unexplained reason, the first GnRH injection fails to ovulate a follicle about 50% of the time in heifers compared to about 17% failure in lactating cows. We have found that the fixed-time insemination (Ovsynch) produces slightly lower conception rates than are achieved when AI breeding is done after detecting a cow in standing estrus (GnRH + PGF 2" + heat detection). However, looking at the number of pregnancies achieved per unit of time, we find that the second program is very competitive.
When fixed-time inseminations are performed in cows that you are attempting to A.I.breed, then by definition conception rate is the same as pregnancy rate, because the heat detection rate (AI submission rate) is 100%. Therefore, PR = HDR × CR becomes PR = 1 × CR or PR = CR. For example, let's compare a traditional AI program that uses heat detection to Ovsynch in which no heat detection is necessary prior to first service (Table 4). If 70% of the cows in the traditional program are submitted for insemination (70% heat detection rate), with a 50% conception rate, 35% of the cows become pregnant in a 21-day period. With an Ovsynch program, 100% of the cows are inseminated, and with a similar conception rate, 50% of the cows become pregnant in a 10-day period. Therefore, 15 more pregnancies are achieved at a similar conception rate because all cows eligible for insemination are AI bred; or in other words, 30 eligible cows in the traditional program were not inseminated because they were not detected in heat. Therefore, more pregnancies can be established per unit of time.

Costs of Heat Detection
Programmed breeding systems not only provide an organized approach to administering first AI breedings to dairy cows or dairy heifers, but should be cost-effective in most herds. Can one determine whether or not the programbreeding system is cost-effective?
The biggest problem in estimating the cost of a programmed-breeding system is estimating the real dollar value of heat detection and the convenience factor of using a programmedbreeding system. If all cows were AI bred during one season of the year (seasonal calving and breeding), the value of heat detection could be determined more easily as a component of the total number of pregnant cows at the end of the breeding season. Perhaps a similar value could be determined by calculating the number of pregnant cows at 100 or 150 days in milk, or the number pregnant after one round of a programmed breeding system. In this way, the value of heat detection, as a component of the pregnancy rate equation, might be estimated.
Because programmed breeding systems basically are designed to synchronize estrus before the first A.I.-breeding, cows must be watched to observe the repeat estrus that occurs when they fail to conceive to first service. One way to eliminate heat detection almost completely would be to determine an early pregnancy status (for example, by day 15, which is not possible now) before an open cow repeats to estrus and then synchronize the repeat estrus so no heat detection is necessary. Another way would be to use ultrasound and diagnose pregnancy at 28 to 30 days and then reprogram the next estrus in the nonpregnant cows. Even with that approach, some cows will be repeating to estrus at 21 to 23 days after their first AI breeding that should have been detected in estrus and reinseminated before the pregnancy test.
Therefore, our current programmed breeding systems require daily heat detection to pick up the repeat estrus. That being the case, the cost of heat detection should be viewed as a fixed cost just as milking labor.

Costs of Programmed AI Breeding
Assessing the costs of using programmed AI breeding is not easy. Further, most producers assume that it is more costly because of the extra labor, semen, and hormones. Table 5 summarizes how programmed breeding pays for itself. Let's assume that you are using Ovsynch and want to compare that to AI breeding cows based on heat detection, perhaps coupled with tail chalk, tail paint, or even Kamar®or Bovine Beacon® heat-mount patches. The total cost of Ovsynch is about $38 ($13 for the three injections, $5 for labor to administer injections, $15 for semen, and $5 for AI breeding). That compares to $20 (semen + AI breeding) for the traditional approach. If we assume that conception rate is 40% in both cases, then at a 70% heat detection rate, the traditional program would produce 28 pregnancies (70 × 40) and Ovsynch would produce 12 more pregnancies or 40 in total.
What is the additional value of those 12 pregnancies? To determine this, we need to estimate the value of one pregnancy after the cow has already failed to conceive once. It takes about 63 days to get a cow pregnant after the first unsuccessful service, so at only $1 per day, the pregnant cow has a $63 greater value compared to the nonpregnant cow. On average, 2.5 more doses of semen + AI labor will be needed or $50 more per pregnancy. If we assume that 20% of the cows will fail to conceive, the cost of a replacement heifer is $1200, and the value of a cull cow is $500, then we must add $140 ($700 × 20% culls). So one additional pregnancy is worth $253 ($63 + $50 + $140). Because those 12 additional pregnancies cost us $200 each, we have a positive return on our investment of $53. Now if heat detection is closer to 50% as in most herds, then only 20 pregnancies are achieved in 21 days and that is 20 less that what is achieved with Ovsynch. Each of those pregnancies would cost only $140 ($3800 Ovsynch costs -$1000 traditional costs/20). Because of poorer heat detection, it will take one more estrous cycle or 84 days to get 80% of the remaining cows pregnant, so the value of a pregnant cow is $84 more than that of the open cow. The costs of semen, AI breeding, and culling are the same, so the value of one additional pregnancy at a 50%-heat detection rate is $274 ($84 + $50 + $140). That means the cost of $140 per each additional pregnancy gained by Ovsynch gives a positive return of $134. Clearly, Ovsynch or other programmed AI breeding systems can pay for themselves because more cows become pregnant per unit of time, so even though more costs are associated with their use, the return on investment is greater. Based on these cost estimates, as either heat detection, conception rates, or both decline, the programmed AI breeding, in this case, Ovsynch, pays for itself.
These differences between the two programs might be even greater, if the costs of heat detection and tail chalk, tail paint, or heat-mount detectors in the traditional program were included. We know that heat detection cannot be eliminated completely, so it leaves us wondering how to estimate the real costs of administering a programmedbreeding system. Of course, many variables determine the cost-benefit ratio of a given system on each farm, for example, the number of cows, type of housing, cost and availability of skilled labor. The selection of the best programmed-breeding system for an individual herd also depends on that herd's rate of heat detection. Those herds with excellent heat expression and(or) heat detection may be served best by programs with less hormonal intervention. 3. 01"""""""""""    Figure 4. 3 PR = HDR × CR.