University of Missouri - Columbia, S134 Animal Sciences Unit, Columbia, MO 65211

DHIA production records have been available to US dairy farmers for 85 years. Participation in DHIA has been associated with improved production efficiency and profitability. Michigan DHIA herds had 700 kg more milk per cow per year than non-DHIA herds (1). More recently, US DHIA herds had a 1360 kg advantage over non-testing herds (2). A survey of midwest dairy farms found a positive relationship between length of participation and production average (3). Profitability was also increased in herds enrolled in DHIA. Economic return of DHIA increased as the number of consecutive years of participation in production testing increased (4). McCaffree et al. (4) also reported increased economic benefits with larger herd sizes but also reported that dairy farms with as few as 40 cows could realize returned profits.

Simply being enrolled on DHIA does not assure higher milk production. The utilization of the records to support management decision-making is essential for the success of the dairy enterprise. Managers that report using DHI records to make feeding management decisions had herds with higher production, greater percentage cows in milk, and fewer days to first service (5). Improved profitability through the use of DHIA records is associated with a long term commitment to the use of production management records. Ohio dairy producers who strongly agreed that DHIA was cost effective had increased production per cow over 770 kg during a 7 year period compared with those who were neutral (230 kg) or those that strongly disagreed (63 kg) (6). Therefore, the use of DHIA production management records, if used correctly and consistently, can make an economic difference for the dairy enterprise.

In using production management records for analysis and decision aid, the records must be complete and accurate. Incomplete records only give half answer or may raise more questions. Inaccurate records can be and likely would be an economic liability due to poor decisions made using poor records. Visible and complete animal ID is required. A review of the records on individual cows can be an effective audit and helpful in maintaining quality records. This is especially true for reproduction management information such as breeding dates and pregnancy status. These two aspects, completeness and accuracy, of production management records are often "assumed". When working with production management records, review and evaluate this aspect of the records before proceeding with the analysis.

Using the records to develop a management action plan is step two. To develop a management action plan, the farm must identify the management areas that require improving. This step is often referred to as a weak link analysis. By taking this step, farm management can prioritize efforts to improve weak areas that need the most improvement. In order to identify management areas that require improvement, managers and service providers/consultants must review the records. One easy way is to compare past and current status of the herd's performance with production management targets or goals. The first level of goal setting are those factors that influence milk sold per cow per day. DHIA factors that contribute or influence milk sales include peak milk yield, persistency of milk yield, days open, days dry, and culling rate. There are a large number of factors that influence each of these primary goals or targets. The remainder of this paper will describe production management goals for these primary management factors and how they can be and are influenced by a set of secondary management factors.

Targets (management goals) provide a measuring stick for comparison. Comparing the herd to regional or state average performance values (Table 1) provides the basis for assessing herd performance (7). Another approach that can stimulate improvement is assessing performance on a percentile basis. The percentile approach compares the herd being evaluated to others contained within a database. The Score 202 developed at the University of Missouri utilizes all herd processes by DRPC Raleigh to develop its database (8). Based on the level of milk production (RHA-fat corrected milk), the program compares individual herd performance to the herds in the database with similar levels of milk production. A median value is the 50th percentile. A value at the 75th percentile indicates that 25% of the herds in the database have better values than the comparison herd. Table 2 presents selected herd performance measures at various percentile ranking for different levels of milk production.

Source: Generated using SCORE202 program and data supplied by DRPC@Raleigh. Raleigh, N.C.

 Within herd comparisons provide a historic representation of herd performance and allow for a comparison of current performance to past performance. When evaluating herd performance on a monthly basis it is important to recognize parameters that are influenced by season. When within herd comparisons are made for parameters influenced by season it is helpful to compare to the previous year.

When developing herd performance targets the emphasis should be placed on those parameters that have the most economic impact or represent a bottleneck to herd performance. Herd performance measures having the greatest impact on RHA milk are, peak milk production, persistency, calving interval (days open), somatic cell count and involuntary cull rates.

Source: Generated using SCORE202 program and data supplied by DRPC@Raleigh. Raleigh, N.C.

Peak milk yield is the culmination of numerous factors. One obvious factor is the relationship between the genetic potential of the herd and peak milk yield. It is important to note that higher producing herds with higher peak milk yields have genetically superior cows (PTA$, all cows). The expression of the genetic potential is dependent on the animals' overall environment and health. But what do we expect peak milks to be and when or how do we know something is wrong?

A first analysis would be evaluating peak or summit milk yields reported by lactation number on the Herd Summary. As a general statement, a normal distribution of peak milk yields would include the highest peak yields for third+ lactation animals. The younger cows would have lower production, with second lactation cows averaging 90% of mature cow performance and first lactation animals having 75% of mature cow performance. Evaluating the lactation curves of these groups of animals by age group is also a useful tool in analyzing the peak milk production of the herd. Two scenarios commonly observed on farms include poor performance of first lactation animals (relative to mature cows) or excellent performance of second lactation cows that is equal to the mature cow group.

In the first scenario, the lack of peaks in first lactation cows can indicate several possible problems. A number of these problems may be affecting the whole herd (i.e., Poor hoof health, high SCC during early lactation, etc.) and are discussed in more detail below. Our experience has been that poor performance of just the first lactation animals indicates poor nutritional management. For example, poor plane of nutrition of the heifer during development will result in the failure of first lactation animals to reach optimal body size at calving. Small heifers that experience significant growth and development during the first lactation have repartitioned nutrients away from milk production to muscle and skeletal growth. This shift in nutrient utilization results in low peak milk yields.

If the first lactation animals are at target weights at calving, but still do not peak at proper levels relative to mature cows, this can indicate poor nutritional management of the heifers during the transition and in the milking herd. The problems associated with transition can also be a whole herd problem and are discussed below. In the absence of metabolic problems, if heifers fail to reach peak milk production, the feeding management system should be investigated. In many cases, limited feedbunk space allows older, larger, more dominant animals to force the smaller, more timid heifer away from the feeding area. As herd sizes continue to get larger, the use of grouping will allow the early lactation heifers to be grouped and fed as a single group with more opportunity to reach potential peak yields.

A myriad of factors or events can lead to poor peak milk yields of the older cows. In the event that mature cows are not reaching peaks compared to second lactation animals, the records of these animals should be reviewed as a group. It is important to realize that older cows are more susceptible to metabolic disorders during the onset of lactation. In addition, mammary gland health in the previous lactation can also lead to diminished performance in the following lactation. Poor culling decisions or high involuntary culling may actually force producers to keep lower producing mature cows. In addition, the older cow can experience lower peak milk yields due to the general problems discussed below.

Optimal peak milk first requires minimal metabolic disorders during the transition period and early lactation. Cows suffering metabolic disorders during the periparturient period have lower dry matter intake for as long as 30 days postpartum.

Body Condition Score (BCS) is an estimate of subcutaneous adipose tissue stored by the animal for mobilization during periods of negative energy balance. As Kertz and coworkers noted, milk yield is accelerated during early lactation with a concurrent delay in feed intake (9). Due to the difference in feed energy intake and milk energy production, we expect animals to lose body condition. In fact, Pedron and others reported the loss of 1 body condition score corresponded to 430 kg increased milk yield (mature equivalent basis) (10). Body condition score at calving was also found to be related to FCM yield the first 90 days of lactation (11). Therefore, adequate body condition at calving supports peak milk yield.

Peak milk production can also be influenced by animal health. For example, cows suffering clinical mastitis during the early postpartum period will likely not achieve peak milk potential. In addition, cows suffering from feet and leg problems that limit mobility and time spent eating will also have lower peak milk production. Heat stress can also cause animals to fail to reach high peak milk yields. Season of calving and animal environment can combine to limit production of genetically capable dairy cows. A thorough analysis of the peak milk yields can help determine the management area needing attention.

Once achieving the target peak milk production, milk sold per cow will be influenced by the persistency of the lactation curve. The animal has a natural decline in milk production over time following peak milk production. An abnormally rapid decline in milk production (poor persistency of lactation) results in significant loss of potential milk production. Using a random sample of Missouri Holstein herds, the following table was developed to illustrate this relationship. Herds were downloaded using CTAP (a computer software package available from DHIA), and grouped by production. As illustrated in Table 3, the higher producing herds in addition to reaching higher peak milk yields, have a higher persistency during all stages of lactation.

Factors that influence persistency of lactation include age, body condition, nutrition management, animal health (ie, mastitis) and cow comfort. As a rule, primiparous cows have a more persistent (or flatter curve) lactation curve. A severe negative energy balance could also result in a severe decline in body condition. Poor energy balance could be due to:

• under feeding energy
• severe anorexia during early post partum resulting in thin BCS.
• loss of BCS during late gestation and pre partum transition period.

 Episodes of mastitis that result in tissue damage of the gland can also cause decreased persistency. Periods of severe heat stress can also result in significant declines in milk production. Evaluating records such as somatic cell count scores, body condition score changes, and seasonal patterns/trends can help explain problems of poor persistency of milk production.

Herd reproductive performance affects the average daily milk production (tank average) due to its influence on the average days in milk. There is a direct relationship between the calving interval and the average days in milk. However, shape of lactation curve, culling practices and extended dry periods can somewhat distort the relationship. In general as the calving interval increases, the average days in milk increases. Figure 1 shows a lactation curve for a herd producing 9980 kg of milk. It points out the effect that days in milk has on the average daily milk production (tank average). If the average days in milk is 165, the daily tank average would be 32 kg. If the average days in milk increases to 195 days, the tank average milk drops to 29.5 kg. Based on this relationship, the general recommendation is to have calving intervals of 12.5-13 months.

 When evaluating herd reproductive performance it is helpful to understand the factors that influence the herd performance measures so a management strategy can be designed and implemented.

The factors that influence minimum projected days open are:

• Average days 1st service.
• Services per pregnancy.
• Heat detection rate.

The factors that influence average days to 1st service are:

• Voluntary waiting period (VWP).
• Heat detection rate.

The effect of changes in VWP and heat detection rate on the average days to lst service are demonstrated in Table 4.

 Monitoring heat detection rate begins within the first three weeks post-calving with the recording of all observed heats. The routine post-partum exam performed at 20-40 days post calving is recommended as an early detection procedure designed to evaluate uterine involution and the re-establishment of cyclicity. Cows that have not been observed in estrus by the end of VWP should be re-examined by an experienced veterinarian and a determination made for the presence of pathologic conditions that prevent cycling. The two major conditions encountered are pyometra and cystic ovarian disease.

The number of cows diagnosed with pathologic conditions after 50 days post-calving should be less than 10%. The body condition scores at 50 days post-calving should range from 2.0 - 3.0 (1-5 scale).

The second factor influencing days open is services per pregnancy. Services per pregnancy is influenced by the accuracy of heat detection, ova quality, health of the reproductive tract, insemination technique, semen quality and infectious diseases. With so many factors involved it is critical that a systematic approach be followed to evaluate this area.

The first step is to look at the differences between services per pregnancy (pregnant cows) and services per pregnancy (all cows). The difference between these numbers provides insight into herd reproductive management. In most herds, approximately 50% of the cows are pregnant at any point in time. The number of breedings required to get these cows pregnant provides the numerator for determining services per pregnancy (pregnant cows). At the same time there are cows in the herd that have no breedings and cows that have been bred but not determined pregnant. Services per pregnancy (all cows) include all services more than 64 days before test day plus services for cows bred in the last 64 days which have not been diagnosed pregnant or open. Therefore, as the numbers of repeat breeder cows in the herd increases, the differences between services per pregnancy (pregnant cows) and services per pregnancy (all cows) increases. A rule of thumb is that the difference should be .5 services or less.

Services per pregnancy (pregnant cows) in general evaluates the level of fertility in the fertile cows. Infertile cows generally end up as repeat breeders remaining open for long periods of time, therefore influencing services per pregnancy (all cows). If the services per pregnancy (pregnant cows) is greater than your target, a systematic evaluation of all factors that influence herd fertility level should be done.

Infectious diseases tend to affect the level of fertility in all cows at a particular point in time, thus increasing services per pregnancy (pregnant cows). Poor insemination techniques or semen handling would produce similar results. The accumulation of older cows or nutritional problems related to energy balance often affects selected individuals and results in the difference between services per pregnancy (pregnant cows) and services per pregnancy (all cows).

Services per pregnancy numbers can be artificially influenced by culling practices. Strict culling of repeat breeders cows will adjust services per pregnancy (all cows) downward. It is also important to recognize that these numbers of averages, depending on the herd size, can be significantly influenced by one or two cows that have multiple services.

The next step in evaluating the herd reproductive performance is to look at the current breeding herd. The current breeding herd includes cows that have not been bred and are past the VWP, cows bred but diagnosed open and cows bred too recently to determine their pregnancy status. To affect change efforts should be concentrated on these cows. The cows in the current breeding herd are divided into two areas. Those that have not been bred and those that have been bred, yet not determined to be pregnant.

If your goal is to reduce the average days to first service then the number of cows with no service that are greater than 100 days in milk must be reduced. A thumb-rule would be to have less than 10% of the cows exceed 100 days in milk with no breeding.

The second category of cows in the current breeding herd are bred cows. This includes cows bred too recently for a pregnancy exam, along with cows bred and diagnosed open. This group of cows influences the average days open (calving interval). The majority of the cows should be bred between the VWP and 100 days in milk. Conception rate and early embryonic death will determine the number of cows that establish pregnancy. Those that fail to establish a pregnancy should return to estrus. However, the nutritional and health status of the cow will affect her return to estrus. Cows that become sick and lose body condition or develop uterine or ovarian pathologic conditions may not return to estrus. The reproductive examination for pregnancy post breeding is helpful in diagnosing most of these conditions. The major factor influencing this group of cows, reproductive performance is the failure to observe the behavioral signs of estrus. Understanding how environmental conditions, nutrition and cow health affect the expression of behavioral estrus, along with good heat detection techniques will help to improve this area.

Dairy cows leave the herd for voluntary or involuntary reasons. Involuntary reasons include death, mastitis, reproduction, feet and legs, and injury or disease. Voluntary reasons include those animals sold to other producers as productive animals and productive animals sold because their level of production does not meet herd standards.

The management decisions associated with culling practices have a major impact on profitability. For example, herds that are able to maintain low involuntary cull rates have the opportunity to merchandise cows, sell replacement heifers or sell off low producing cows, thus improving the overall herd performance.

When evaluating herd performance measures recognize that culling practices can influence the numbers. A herd with a large number of repeat breeder cows can sell the repeat breeder temporarily improving the reproductive records.

It is important that accurate records be kept on the reasons cows leave the herd. A cow that fails to establish pregnancy until 250 days in milk that drops in milk production may be sold. When recording the reason for leaving the herd, the choice is low production or reproduction. If the cow would have established pregnancy within a reasonable period of time, she would have been sold for low milk production, if not, the reason for leaving the herd should be recorded as reproduction.

Monitoring cull rates provides timely information so management can adjust to correct problems. For example, if the herd historically has a disease/injury cull rate of 4% and has recorded that most of the disease and injury occurs during the first 30 days of lactation, one would expect to lose 1 cow during the first 30 days of lactation. Based on historic cull rates, management can anticipate losses, thus monitoring prevention programs.