Culling: Nomenclature,
definitions and some observations
John Fetrow, VMD, MBA, College of Veterinary Medicine, University of Minnesota
Ken Nordlund, DVM, School of Veterinary Medicine, University of Wisconsin
Duane Norman, PhD, USDA Animal Improvement Program Laboratory, Beltsville, MD
In advance of the Discover Conference on Culling in Dairy Herds in October, 2004, a subcommittee was formed for the purpose of laying out a proposed set of definitions of terms relating to culling on dairy farms. This paper is the product of that effort. In addition to the specific charge, the committee has chosen to make some observations on the general topic of culling in dairy cattle and on appropriate ways to examine the underlying factors surrounding the exit of cows from a dairy. The committee hopes that the contents of this paper will encourage the dairy industry to standardize the use of some terms and discontinue the use of others. Further, we hope that this discussion will contribute to a wider re-examination of long-held dogma regarding culling in dairy cows.
SIMPLE DEFINITIONS
Culling
Culling (exiting) is the departure of cows from the herd due to sale, slaughter, or death. Thus the term cull includes all cows that leave the dairy, regardless of their destination or condition at departure. There is some reason to object to including cows sold for dairy purposes as part of a general “cull” category, since the word “cull” generally means to separate off for undesirable reasons. This categorization may cause problems of interpretation for dairies that market adult cows for breeding or milk production, confounding their use of industry benchmarks regarding the number of animals that exit in comparison to dairies that do not market adults. At the same time, a general term is needed and “cull” remains the term in general use in the dairy industry.
The most unambiguous classification of cows removed from the herd is based upon destination of the cow after removal.
Dairy Sale
Sale (sold) in this context means that the cow was sold to another dairy alive, generally with the express goal of continuing to provide some income, such as producing milk, calves, or embryos.
Slaughter
Slaughter as a destination means that the cow left the dairy alive to be slaughtered for human consumption. This would include cows slaughtered on-farm for family or employee consumption.
Salvage
Salvage refers to those animals that leave the dairy alive, but not intended for human consumption, e.g. rendering, non-human food, etc.
Death
Death (dead) means the cow left the herd through death on the dairy. With the new rules regarding downer cows, more cows can be expected to be culled as dead cows.
Recommendation
The four mutually
exclusive destinations of a cow removed from a herd are sale, slaughter,
salvage, or on-farm death. We would
encourage dairy record system designers to adopt these four destinations as
standard terms in their systems. To complement
this system, users could add reasons for each cull event (see below for a
discussion on reasons). In that way,
the dairy could track the fate and relative numbers of cows that were culled by
destination.
DEFINING THE MAGNITUDE OF CULLING ON
A DAIRY
There is a generally perceived desire to quantify the amount of culling on a dairy.
From a precise epidemiologic perspective, culling is a specific event (an “incident”) in a cow’s life on the dairy. Measuring the occurrence of incidents is usually done by measuring the rate (incidence) of the events over a period of time in an “at risk” population. A simple count may be useful to a specific dairy (how many cows were culled last year?), but if the magnitude of culling is to be compared between dairies, then some standardization (e.g. percent of the “at risk” population) is needed to account for different herd sizes, i.e. an estimate of the “risk” in a defined population. The calculation should also specify a fixed time period, e.g. one year.
Culling Incidence Rate
From an epidemiologic point of view, the ideal measure of the amount of culling is a “culling incidence rate” (Dohoo: page 68; 2003).
The number culled is straight forward; simply count the cows that exited over the relevant time period. A specified time period is typically a year for culling, but could be per lactation, per month, or for some other specified time period (e.g. the first 60 days of lactation). The population at risk is the problematic question that plagues discussions of culling.
Formally, one way that population at risk can be determined is by following a pre-defined cohort of cows over time until they were all culled. A cohort is a group of individuals with some common characteristic when they are assembled, and who are then monitored for some period of time. For each cow, the years of life from start to cull would be determined, the sum of those years across all cows calculated, and the result expressed as “cow-years” at risk. The resulting culling incidence rate would exactly define the risk per cow per year of being culled. These sorts of studies and calculations are appropriate for prospective epidemiological or clinical trials. Alternatively, all cows on a dairy at a point in time could be followed for a year, counting the number in that cohort that was culled. Like all prospective cohort studies, consideration must be given as to whether the starting cohort (e.g. all cows in the herd on July 1st) is representative of the general population or demographics of interest.
Unfortunately, there are practical problems for either of these methods of calculating the magnitude of culling on an operating dairy for management use. For the first technique, cohorts of cows generally do not conveniently arrive together (although this is not strictly required for determining cow-time at risk, but it does make it easier). If one simply starts with a cohort that is determined by all of the cows in the dairy on a given day, then that day’s particular demographics (by lactation and by stage of lactation, by season, etc.) will tend to influence the observed culling incidence and thus reduce the comparability to other herds’ data. In each successive month, fewer and fewer of the original cohort of cows will remain, and those cows will be mingled with other cows that have joined the herd over time. Finally, it takes a long time to track the cohort and collect the data, so by the time the results are in and tabulated, the information derived may not be relevant to current management needs. Suffice it to say that culling incidence rate is often the proper research tool to use regarding culling, but it is usually too difficult or historic to compute to be used for practical managerial purposes.
Herd Turnover Rate
When the literature and current working record systems are considered, there are a number of terms in practical use to describe culling. Terms like “herd turnover rate” (Minnesota DHIA, among others), “culling rate” (Brett 2003; Hoekema 1999, among many others), “proportion removed from herd” (Smith 2000), “percent left herd” (Gangwer), “percent leaving” (AgSource), and “cows left herd, %” (Wisconsin DHIA Herd Summary Form 202), “replacement rate” (Allaire 1981), and others are used to quantify the extent of culling in a population of cows. The method of calculation of these indices can vary both between indices and for the same index for different authors or record systems (e.g. culling rate: Brett 2003 versus Hoekema 1999). This means that the reader or user of an index must be wary when interpreting a value or comparing it to other dairies or studies (Radke 2001).
Again, the numerator of these calculations is fairly straight forward. There are some cases where the calculations exclude dead cows, which is a mistake if overall culling is being considered. There are several versions of denominators, but two are used most often. The first is to calculate the average number of cows (lactation 1 or older) on the dairy for the year. A simple approach is to calculate: (starting inventory + ending inventory)/2, but a better approach would be to average the cow inventory at each monthly test occurring over a 365 day period (e.g. DRMS). This better reflects the general inventory, since it accounts for changes in population size across the year and the calculation comes closer to actual cow years in the herd. If herds are fairly stable in size, the result is a fair approximation of cow years at risk, even if the cows represented in the starting cohort are not all the same cows in the ending cohort. Computers make it possible to actually calculate cow-years, cow-months or even cow-days in the herd and this is the ideal.
Herd turnover rate is defined as follows:
Number of culls over a period (year) Average inventory of cows during the period (or if
available, cow-years at risk)
This ratio, times 100, produces a familiar measure of the magnitude of culling from a dairy, expressed as a “herd turnover rate (%), but is also called “cull rate”, “culling rate”, “percent exiting”, and “proportion leaving” in various publications.
An alternative denominator described by various authors and in use in some record systems (“Herd turnover rate”: Radke 2001; “Percent cull rate”: Minnesota DHIA) is calculated by adding the culls to the current or average herd inventory. The resulting definition of herd turnover rate or culling rate by this method becomes:
Number of culls over a period (year) Average inventory of cows during the period + number of
culls during that period
We wish to argue that this definition is misleading and ought NOT be used to describe culling. The verbal justification for adding the culls to the inventory is that if one lists all of the cows “at risk” of culling at some time during the year, then both the current herd inventory and the culled cows were all “at risk”, at least at some point. This is true as far as it goes, but in epidemiologic terms, this population of cows was not all “at risk” for the defined period of interest. More likely, each culled cow was replaced at some time during the year by a new cow that filled her place. Considering both the culled cow and her replacement across the year, the farm had one cow in place for most of the year, resulting in (at most) one “cow year at risk”. Adding the culled cow and the replacement together in the denominator makes the calculation seem like there were two “cow-years at risk”; an error in epidemiologic terms. The sum of the average inventory plus the cows culled overestimates the animal time at risk. If two months into the year a cow is culled and immediately replaced, then there is still only one “cow-year” at risk, not two. As an extreme example, consider a herd that has only a single cow that milks for 11.99 months, is culled, and immediately replaced by a newly calved heifer. Using the first definition above (the preferred one), the herd turnover rate is 100%. Using the second definition (the one we recommend against), the culling rate is only 50%. We believe that the first parameter (100%) far better reflects the “herd’s” reality. In fact, the second definition will consistently underestimate the risk that a cow will be culled during a specified time period.
Table 1 illustrates the calculation of herd turnover rate by both approaches and for four types of herds: herds with moderate and high herd turnover rates and herds with stable and expanding herd sizes. The example demonstrates that the alternative definition (adding the culls into the denominator) underestimates the risk of culling in the herds. It also demonstrates that the preferred calculation works as an estimate of the risk of culling even in rapidly expanding herds, as long as the denominator calculation of average inventory is done on a monthly basis or more frequently.
Alternative Estimates of the Magnitude of
Culling
There are other, more indirect approaches to estimating the rate at which cows leave a dairy. It is possible to calculate backward from the demographic information available on a dairy to an estimate of the rate of cows exiting. The formula to do so is (all times are in years): estimated herd turnover rate
1 / ((average age of milking cows - average age at first
calving)*2)
The rationale for this approach is as follows: Average age of milking cows – average age at first calving = the length of time the average cow in the herd has spent thus far in her adult life in the herd. Double that value would estimate the average total time (years) that cows spend in the herd as adults. Dividing one by that total time (in years) would yield an estimate of the rate at which cow exit per year (modified from Durr 1997).
There are problems with this approach. The numbers used in the calculation again depend on events that happened long ago, so the results may not represent the conditions in the herd today. If the calculation does not include cows that have already died as a part of the cohort being considered, then the longevity of cows is overestimated and the estimated herd turnover rate is too small. Finally, the data themselves are often suspect; age at first calving on some dairies may only be a convenient estimate (often simply entered into record systems as 24 months when a heifer first calves), making age unreliable on some dairies. For these reasons, we do not recommend that this approach be used as a routine way to estimate turnover rates in a herd.
Another measure of culling magnitude can be calculated on a dairy by determining the lactation specific survival (culling) rate, i.e. the probability that a cow that enters a particular lactation will also subsequently calve on the dairy. These estimates of risk of being culled by lactation can be derived from the dairy’s records by following a retrospectively selected cohort of cows for each lactation from calving to either the next calving or culling. If a full year’s calvings are to be included to account for seasonal effects, this determination requires data from a minimum of two years ago to assure that the outcome is known for all cows in the cohort made up of a year’s calvings. Given these lactation specific lactational culling rates, a hypothetical lactation distribution for the herd can be calculated and compared to the herd’s actual distribution. Studies have described these rates across the U.S. dairy industry and elsewhere and the resulting lactation demographics (Nieuwhof 1989, Durr 1997). Given the lactation specific culling rate, simple spreadsheets can predict the demographics of the herd at steady state (Table 2).
Recommendation
As the term to represent the magnitude of removals from a herd, we prefer the term “herd turnover rate”. Turnover rate is a traditional term used in business inventory monitoring (US Bureau of Labor Statistics), and also reflects that the number is based on herd performance. Turnover rate, unlike culling rate, also avoids the negative connotations of “cull” for cows that leave for dairy purposes. It seems likely that the other terms will remain in use, however. We propose that all of the various terms in use be considered synonymous, i.e. the preferred definition above be used for all of these terms. Thus “herd turnover rate”, “culling rate”, “percent exiting”, etc. would all have the same meaning and should be calculated in the same manner.
Subsets of herd turnover rate
In the analysis of culling history on dairies, it often becomes useful to consider the rates in subsets of the total population, e.g. in first lactation cows only, in animals that had a dystocia or other health event, animals calving in July, culling in the first 60 days of lactation, etc. While the above calculation of herd turnover rate works just as well in a subset of cows, care must be taken to assure that the denominator cohort being considered is correctly followed and the resulting rate is properly described.
For subsets of animals routinely present in the herd (e.g. first lactation cows), one can calculate a reasonable estimate of risk of culling by calculating the herd turnover rate across the year the same as for the herd as a whole, e.g. the total first lactation cows culled in the year divided by the average inventory of first lactation cows. This would give a “herd turnover rate per year for first lactation cows”. For other subsets, the process probably should be more like an epidemiologic cohort study. To consider the cull rates of cows that suffered a dystocia, for example, one might better start with all cows with dystocia and follow them to a specified end point (probably either for a year or for a lactation). The resulting rate would be expressed as “turnover rate in the year (lactation) following a dystocia”. These calculations may also lead to extreme variability when the population being considered becomes small. For example the number of cows that suffer a dystocia in a 100 cow herd may be as few as six. Any generalizations about causes or outcomes in such a small starting population are suspect at best.
When considering turnover or culling rates for times other than full lactations, attention must be given to the fact that the risk of culling is not a consistent level across all stages of lactation. In fact, cows experience their highest risk of culling early after calving, then the risk of culling (per day or per month) drops and then rises toward the later stages of lactation (Figure 1, Godden). This tends to even out if calculations are done in a herd of stable size across a year because over that time all seasonal effects and stage of lactation effects are included and probably are representative of the dairy’s general culling risk profile. In contrast, if one were only to consider the culling experience of a newly arrived cohort of first lactation animals over the past 6 months, the results are not likely to reflect the general pattern of culling in first lactation animals for the dairy on a continuing basis because the time period over which these heifers are followed may have different risks of culling than for heifers in general across a full lactation.
Using the Estimates of the Magnitude of Culling:
Issues arising around the magnitude of culling arise regarding both the national herd and individual herds. Perhaps it is useful to consider some of the issues related to each level.
Use at the national herd level
On a national basis, the annual national herd turnover rate will necessarily be:
If the national dairy herd had 8,900,000 cows at the beginning of the year and 8,800,000 at the end and 3,000,000 heifers calved, the years cull rate would be:
As this calculation shows, culling rate at the national level is inevitably determined by the change in the national dairy herd size and the supply of available heifers. Unlike the beef cattle industry, almost all dairy heifers are reared for herd replacement purposes because the economic value a dairy replacement heifer far exceeds the alternative value of a heifer in the beef slaughter market. If annual herd turnover rates in the nation have risen over the past decade, it can only have come about by reducing the size of the national herd or by increasing the supply of new heifers to calve, either by reducing heifer mortality, increasing cow reproductive productivity (shortening the average calving interval), importing heifers, or shortening the time it takes for heifers to reach calving age. Viewed broadly, the increase in the national herd turnover rate over the past decade has probably come about in part from many of these factors. Note that an increase in the national cull rate does not mean that the dairy industry has been damaging cows at an increased rate or that production stress is driving cows to slaughter. Nationally, higher culling rates may reflect nothing more than improved heifer rearing.
Use at the individual herd level
At the individual herd level, there are several basic questions that typically are addressed using these estimates for the magnitude of culling.
1.
The capacity of this
dairy in N adult cows. How many
replacements will be needed in a typical year?
This can be ably answered by multiplying the herd turnover rate times the capacity N. For example: assume a herd capacity of 1,000 adult cows and a herd turnover rate of 33 percent. The estimate of the number of replacements that will be needed in the coming year is 1,000 * .33 = 330 replacements.
2.
A dairy calves 100 animals. How
many will start another lactation after this one?
This can be estimated by converting an annual turnover rate to a lactational rate (in this case, lactation refers to the inter-calving interval, not just the milking phase). The annual rate, divided by 12, provides an estimate of the monthly culling rate of cows on the dairy. Multiplying this number by the average lactation length gives an estimate of the lactational turnover rate experienced by the dairy over the past year. If one presumes the conditions on the dairy will remain similar, then this projection of culls per lactation is a fair estimate of what will occur. For example: assume a herd of 100 cows, annual herd turnover rate of 36 percent, and average lactation length of 14 months. The estimate of the lactational turnover rate would be (36/12)*14= 42 percent. Alternatively, one could simply determine the lactational culling risk be following a cohort of animals (retrospectively) that from calving in that lactation until they either calved again or were culled. Unfortunately, this approach would necessarily include some fairly distant history in the herd.
Estimating the average lactation length can be done directly, by computing the average time from one calving either to the next or to a cull event. The problem with this approach is that in reaches very far into the past for historical calving data that may no longer represent the herd. If the calculation is done from one calving to the next, it only considers that portion of the herd that have had two or more calves (first lactation animals and animals culled at the end of lactation have not had a second calf). A second alternative would be to sum the herd’s average days open plus 280 days (gestation). This parameter suffers as well from only looking at the subpopulation that has a confirmed or projected pregnancy. Perhaps a better approach is to double the average days in milk of the herd, (thus estimating the average lactation length) and then adding the average dry period, although this approach may also include distant historical data and may not include culled cows.
3.
If a dairy is in the
midst of an expansion, does the calculated turnover rate describe what is
happening and does it predict removal rates into the future?
The above defined “herd turnover rate” works in expanding herds as well as stable sized herds in so far as describing what has occurred. It does not necessarily serve as a reliable estimate of what culling rates will be experienced in the herd in the future, however. There are two main reasons for this. First, in order to fill new facilities, expanding herds often voluntarily choose to limit culling of cows that under other circumstances might exit the herd. These cows may not be desirable in the long term, but serve short term needs, particularly in terms of cash flow. As the dairy fills to capacity, there may be significant financial incentive to replace these cows with better animals, raising the cull rate in the dairy above earlier levels seen during the expansion. Second, expansions typically fill the barn with first lactation cows. At least in the first year, fewer of these cows will be culled, in part because they will not yet have completed their first lactation. In the second year of expansion there may be a sudden rise in culling in this cohort as well, either because they were poor producers and did not warrant a second lactation or because of the routine risks around the time of their second calving. For both of theses reasons, expansion herds typically experience an important surge in culling in the second year of expansion compared to the first year. Beyond that, things tend to settle into a culling pattern more indicative of stable operating conditions for the dairy.
4.
How does culling on my dairy
compare to others? Am I culling too
much or too little?
Even though this is probably the most common question asked of culling statistics, it is also the most difficult to answer. Given the charge of our subcommittee, addressing this question in any depth is probably not within our purview, but a few general statements might be in order.
There is a general consensus within the dairy literature that lower annual turnover rates are more profitable and generally the studies suggest optimal turnover rates of 30% or less, based upon modeling or surveys of dairy farm financial records (Allaire 1981, Congleton 1984, Van Arendonk 1985, Williams 1987).
However, there is not an optimal culling rate for all herds or for all years. Culling rates are the net result of a series of culling decisions made each day on individual cows. Those decisions are derived ideally from considerations of economics (milk price, cull price, replacement costs, etc.), farm capacities, health and productive status of the individual cow, disease and death rates within the herd, available replacements, and biosecurity considerations, among others. If the dairy has made optimal culling decisions cow by cow, then the resulting culling rate is ideal for that dairy at that time. However, sustained high turnover rates should stimulate an investigation to identify herd risk factors that devalue cows prematurely within the herd, i.e., mastitis, infertility, lameness, etc.
Herd summary statistics, including herd turnover rate, should not be used in isolation to evaluate herd performance. Looking at a culling statistic for a dairy in isolation and concluding that the rate is either too high or too low is fraught with grave peril. For example, a turnover rate of 25% does not reflect “good” herd management without knowing the productive, reproductive, and health status of the herd and the economic conditions under which that culling statistic was achieved. It might, for example, reflect the herd’s inability to keep replacement heifer calves alive or inability to finance the purchase of needed replacements. Considerations of culling are necessarily a retrospective or historical activity. In the case of annualized rates, the events that ultimately lead to a cull often happen as much as a year or more before the culling event itself. Thus a case of ketosis, fatty liver, and left displaced abomasum may lead to poor production and subsequent culling after 16 months of lactation. If that culling event happened 11 months ago (and thus is included in the current herd turnover rate), then the actual management breakdown in the pre-fresh transition program that is now being included in the statistic happened more than 25 months ago.
REASONS FOR CULLING
The issue of culling reasons is perhaps the other aspect (besides the issue of the amount) that stirs the most controversy relating to culling. There are a variety of ways this issue has been dissected; we will address them only in summary.
Voluntary Versus Involuntary
Traditionally, culls have been classified as voluntary (presumably sale of cows for dairy purposes or live culls deemed to be normal except that they are poor producers) or involuntary (presumably culled by coercion due to mastitis, lameness, poor reproduction, disease, death, etc.). Despite criticisms leveled at this classification for a long time (Fetrow 1988, Dohoo 1993, Leslie 1994, Radke 2000, Radke 2001), this classification has remained a firmly ossified part of the culling mythology. Created for appealing reasons (wouldn’t every dairy prefer to cull for “voluntary” reasons?), the classifications do not reflect the reality of culling decisions or the characteristics of why cows are culled. While sales of healthy, productive cows to other dairies certainly occur (less than 5% of all culls according to NAHMS 1996 data), almost no cow exits the dairy as a pregnant poor producer with a disease-free, mastitis-free and trauma-free history. Cows that are culled as “open” are sometimes truly hopelessly infertile, but in most cases a voluntary, economically based decision was made that the cow was no longer worth breeding in comparison to replacing her with a new heifer.
Recommendation
The industry should
discontinue the use of “voluntary / involuntary” terminology to characterize
culling.
Economic Versus Biological (or “Forced”) Culling
An alternative conceptual distinction for culling has been to distinguish culls that exit due to “forced” or biological reasons from “economic” culls. Biological culls are those cows for which no possible productive future exists, e.g. hit by lightning, permanently sterile, irreparably injured, positive for Tuberculosis, etc. These would be a small minority of all culls on most dairies. Economic culls are those for whom a decision has been made that replacing them with a new cow is the best economic option for the dairy under the circumstances. This distinction remains useful for culling discussions because it underscores the reality that culling is an overwhelmingly economic decision making process. The distinction between the two categories is not, however, a useful day-to-day managerial distinction.
Characterizing Specific Reasons for Culling
Dairy record systems have long offered the opportunity for dairymen to describe “why” a cow exited the herd. Usually limited to one or two choices, the list usually includes “reproduction, mastitis, udder conformation, died, low production, injury, dairy sale, feet and legs, disease”, and sometimes miscellaneous reasons.
The process of categorizing culls with a reason is subject to many forms of bias and error. Sometimes the reason is clear and proximate to the culling event (hit by skid-steer, euthanized), but more often the reasons are distant, less evident, and multi-factorial. Consider a cow that was poorly managed in the transition from dry cow to milking cow, developed ketosis and a left displaced abomasum, developed a chronic mastitis infection due to impaired immune status, was too rapidly transitioned to a high starch diet and developed sub-acute rumen acidosis and subsequent laminitis, was slow to return to estrus and hesitant to stand to be mounted, was never successfully bred and finally was culled 14 months after calving as an open cow with poor production, chronic high somatic cell count, lame, and in poor body condition. What does one expect of the dairyman when he is asked to provide the one reason the cow was culled?
The problem of offering a single reason for culling is borne out in a study of a small number of New England dairy farms where the dairyman was allowed to provide up to three reasons for culling (Bascom 1998). Given the opportunity, 35 percent of cases were given two reasons, 11 percent three reasons. In addition, farmers have been shown to alter their culling criteria and decision making based on sociological variables in addition to economic or biological ones (Beaureau, 1996).
Despite the problems with classification, it remains clear that several problems contribute importantly to the risk of a cow being culled. In the study that most effectively characterized the entire U.S. dairy industry, the National Animal Health Monitoring Service characterized the reasons for culling into seven categories, not including death on the dairy (NAHMS 1996). Reproductive failure was the leading reason given for culling, followed by mastitis and udder problems, lameness, poor production, disease, aggressiveness, and others. Studies and reviews have described these types of patterns and attributed the risk impact of a cow suffering from particular maladies (Smith 2000, Groehn 1998, Beaudeau 1996, Milian-Suazo 1988, Fetrow, 1988).
If managers record and track disease reasons for culling, presumably they do so to guide future management decisions aimed at avoiding problem that lead to undesirable culling outcomes. In addition, outside consultants may refer to these data or tabulations in an effort to better understand long-standing breakdowns in management. Pointing to prevailing reasons for culling may highlight long ignored problem areas and motivate change for the better. The data may also be useful as inputs into economic models of the costs of disease or for recommendations of control efforts. For these reasons, the small effort it takes to record the general reasons for culling cows is a worthwhile part of a thorough dairy record system.
Recommendation
We recommend that
dairy record systems categorize removals by destination first (sale, slaughter,
salvage, or on-farm death), and allow multiple selection of predefined specific
reasons to better characterize the removal of individual cows.
Inappropriate uses of removal reasons
Sometimes dairy managers attempt to use culling reasons to monitor the incidence of disease on the dairy. Those efforts are seriously misguided. First, much of the incidence and losses from disease occur at the subclinical or clinical level and does not result in the death or removal of the cow, so monitors of culling only may detect only the tip of the disease iceberg. Second, by the time the cull event occurs, the disease is usually long past and the need to correct management is extremely overdue. If the dairy manager wishes to monitor disease, it should be done in a more direct and timely fashion using records of clinical disease events and subclinical screening programs.
Recording diseases
While not directly related to the topic of culling, the issue of monitoring disease on dairies becomes immediately relevant when management asks about systems to avoid culling. By monitoring disease rates on a continuing basis, management breakdowns and deficiencies can be identified much more quickly and interventions made. This should result in fewer cows losing enough of their value that culling them is a wise economic decision. Defining what constitutes a “case” of a disease can be somewhat subjective (e.g. ketosis); suggested definitions are available in the literature (Kelton 1998). A second issue is what constitutes a new case of a disease (e.g. clinical mastitis or lameness) in a cow that has already had a case in the current lactation. In general, the recommendation is that clinical cases that occur more than 30 days after the first case should be counted as a new case. This presumes that clinical manifestations within a month of the first appearance of a clinical problem are just continuation of the original problem. For many diseases, the recommendation is that only one case can be recorded per lactation (e.g. dystocia, milk fever, retained placenta, LDA). These recommendations avoid duplicative recording of a single case, which would bias the apparent incidence upwards.
The USDA Animal Improvement Program Laboratory has recently drafted a set of standardized health trait terms for recording disease events on dairies (Table 4, USDA 2004). If widely adopted, these terms might help make recording of diseases more consistent and comparisons between dairies more reliable. For the most part, the diseases listed are those common to dairy cows and that can be fairly reliably identified by a clinical examination of the cow by lay staff on the dairy or at routine scheduled veterinary examinations. Thus diseases are identified by signs, not by causative agent (the one exception is Johnes Disease). In addition, the recommendations include three Management Traits that could be recorded as well on some dairies.
Recommendation:
We recommend that DHI
record centers adopt and record the recently developed USDA AIPL listing of
health trait terms and we recommend that dairy extension agents, consultants,
veterinarians, and producers encourage the recording of these health traits in
the national DHI record system.
REFERENCES:
AgSource: DHIA record services: http://cridata.crinet.com/pmeasure/herd.php access 12JUN2004
Allaire FR. Economic consequences of replacing cows with genetically improved heifers. J Dairy Sci 1981;64(10):1985-95.
Bascom S, Young A: A summary of the reasons why farmer cull cows. J Dairy Sci: 81-2299-2305, 1998
Beaudeau F, van der Ploeg J, Boileau B, Seegers H, and Noordhuizen J. Relationships between culling criteria in dairy herds and farmers’ management styles. Prevent Vet Med 25:327-342, 1996
Brett J: What is the ideal culling rate? In Dairy Herd Management on-line from October, 2003; http://www.dairyherd.com/ : Health news, accessed 14JUN2003
Congleton WR, King
LW. Profitability of Dairy Cow Herd
Life. JDS 67:661-674, 1984.
Dohoo I, Martin W, Stryhn H: Veterinary Epidemiologic Research. AVC Inc. Charlottetown, PEI, Canada 2003
Dohoo I, Dijkhuizen A: Techniques involved in making dairy cow culling decisions. Compend. Contin. Ed. Pract. Vet. 15: 515-520, 1993
DRMS: Dairy Records Management System, Raleigh, 2004: Dairy Records Management Systems: DHIA Records: herd summary fact sheet: http://www.drms.org/dhia.htm accessed 14JUN2004
Durr JW, Monardes HG, Cue RI, and Philpot JC. Culling in Quebec Holstein herds. 1. study of phenotypic trends in herd life. Can. J. Animal Sci. 77:593-600, 1997
Gangwer M, Gamroth M, and Seldin R: Agritech Dairy Records Processing Center: Understanding Dairy Herd Performance Measurements from the Agri-Tech Analytics DHIA Herd Total Report: 10 Important Measures, .3. Percent left herd: http://eesc.orst.edu/agcomwebfile/edmat/html/EM/EM8540/EM8540.html#anchor109691 accessed 12JUN2004
Godden S., S. Stewart, R. Cady, S. Eicker, J. Fetrow, P. Rapnicki, W. Weiland, H. Spencer. The Relationship Between Herd rBST-Supplementation and other Factors With Risk for Removal for Cows in Minnesota Holstein Dairy Herds. Proceedings from the 4-state Applied Nutrition and Management Conference. LaCrosse WI. July 2003
Grohn Y, Eicker S, Ducrocq V, and Hertl J: Effect of diseases on the culling of Holstein dairy cows in New York State. J Dairy Sci. 81:966-978, 1998
Hoekema M: Guess what may be eating your lunch: the hidden costs of cull rate (Parts 1 and 2). Dairy Business Analysis Project, University of Florida, 1999: http://www.animal.ufl.edu/dbap: accessed 14JUN2004
Kelton, D.F., Lissemore, K.D., Martin, R.E. Recommendations for recording and calculating the incidence of selected clinical diseases of dairy cattle. J. Dairy Sci. 81, pp 2502-2509, 1998
Leslie K: Culling and genetic improvement programs for dairy herds. Chapter 7 in Herd Health: Food animal production medicine, Radostits, Leslie, Fetrow, Saunders, Philadelphia, 1994
Milian-Suazo F, Erb H, Smith R: Descriptive epidemiology of culling in dairy cows from 34 herds in New York State. Prevent Vet Med 6:243-251, 1988
Minnesota DHIA records calculations for percent cull rate and annual turnover rate: MN DHIA electronic newsletter for May 2003.
NAHMS: National Animal Health Monitoring System: Part I: Reference of 1996 Dairy Management Practices. USDA AHPIS VS at: http://www.aphis.usda.gov/vs/ceah/cahm/Dairy_Cattle/dr96des1.pdf accessed on 14JUN2004
Nieuwhof G, Norman HD, Dickinson F: Phenotypic trends in herdlife of dairy cows in the United States. J Dairy Sci 72:726-736, 1989
Radke B, Lloyd J: 16 dairy culling and replacement myths. Compend. Contin. Ed. Pract. Vet. 22:S36-S57, 2000
Radke B, Shook G: Culling and genetic improvement programs for dairy herds. Chapter 8 in Radostits: Herd Health 3rd Edition. Saunders, Philadelphia, 2001
Smith JW, Ely LO, Chapa AM. Effect of region, herd size, and milk production on reasons cows leave the herd. J Dairy Sci 2000;83(12):2980-7.
USDA Animal Improvement Programs Laboratory: Format 6: Health Record (May 2004 draft version), 10300 Baltimore Avenue, BARC-West, Building 005, Room 306 Beltsville, Maryland 20705-2350.
U.S. Bureau of Labor Statistics. Job Openings and Labor Turnover Survey. People are Asking page. Available at: http://www.bls.gov/jlt/jltask.htm. Accessed May 25, 2004.
Van Arendonk JAM. Studies on the replacement policies in dairy cattle. II. Optimum policy and influence of changes in production and prices. Livestock Prod. Sci. 13:101-125. 1985.
Williams CB, Oltenacu PA, Bratton CA, Milligan
RA. Effect of Business and Dairy Herd
Management Practices on the Variable Cost of Producing Milk. JDS 70:1701-1709, 1987.