Cow longevity economics: The cost benefit of keeping the cow in the herd

It has been shown that a large number of cows leave the herd early in lactation largely due to metabolic health reasons, and the risk of death is greatest early in lactation. Involuntary culling of cows early in lactation is expensive, in the order of $500 to $1000 (380 to 760 EURO) per cow (US data) not including losses in milk yield due to disease and delayed replacement etc. Efforts to reduce death rates and improve early lactation health can therefore be profitable.

(This article belongs to the proceedings from the Cow Longevity Conference 2013 that took place at Hamra farm, Sweden in August 2013)

Take home message

1. Cow longevity depends on intrinsic and extrinsic factors.

2. Survival analysis has shown that a large number of cows leave the herd early in lactation largely due to metabolic health reasons.

3. The risk of death is greatest early in lactation.

4. Forced culling of cows early in lactation is expensive, in the order of $500 to $1000 per cow.

5. Efforts to reduce death rates and improve early lactation health, and therefore intrinsic cow longevity, are to be profitable.

6. The average cow longevity is determined by extrinsic economic factors unrelated to an individual cow’s health and performance. To extend the average cow longevity, dairy farmers would need to be motivated to bring fewer heifers into the herd and therefore cull fewer cows.

  • This could be accomplished by creating fewer dairy heifer calves, for example through delayed inseminations in heifers and/or cows, or not breeding non-pregnant cows that are late in lactation.
  • A portion of the heifers and cows could also be inseminated with beef semen and the crossbred calves could be sold.

Introduction

Cow longevity refers to how long a cow stays in the herd. It is convenient to break the total lifespan into the time before first calving and the time after first calving. The average age at first calving in U.S. dairy herds in 2007 was 25.2 months (USDA, 2007). This was down slightly from 25.4 months in 2002 and 6 months shorter than the average age of 25.8 months in 1996. In 2012 it was still around 25.5 months (Heinrichs and Jones, 2013). The age at first calving slightly overestimates the expected lifespan of a new born dairy calf because it excludes the approximately 15% of heifers that were culled before a first calving.

Culling is the departure of animals from the herd because of sale, slaughter, salvage, or death (Fetrow et al., 2006). Productive life is the time from first calving to culling. It is calculated as the reciprocal of (cow) cull rate. For cows, the annual cull rate in 2013 was approximately 38% (DRMS, 2013) and has been fairly constant for at least 2 decades (USDA, 2013; Figure 1). This is the equivalent of a productive life of 2.63 years, or 31.6 months. The average dairy cow longevity is therefore approximately 57.1 months or 4.8 years in the U.S. According to USDA data, productive life has decreased from 35 months for cows born in 1960 to about 40% for cows born in 2000 (USDA-AIPL, 2013). Around 1930, the average annual cull rate was approximately 25% (Cannon and Hansen, 1939). The natural life span for cattle is reported to be up to 20 years when they would die of old age. Fetrow et al. (2006) point out that this shorter lifespan is primarily the result of an economic decision making process by dairy farmers. Dairy producers cull cows because they are no longer profitable or they are replaced by more profitable cows.

Culling decisions are the result of intrinsic cow factors such as health, milk production, and reproductive status, and extrinsic factors such as the availability of replacement heifers, parlor capacity or land availability, and prices. This paper first briefly reviews some risk factors for culling and presents some cost estimates of forced culling for individual cows. Secondly, the paper explores cow replacement patterns for closed and open herds when various intrinsic and extrinsic inputs are changed and when the herd constraint is parlor capacity or a manure quota. The literature on risk factors for culling and economic implications of cow longevity is extensive. The review presented here is necessarily not complete.

Figure 1. Number of dairy cows present and annual number of dairy cows that were culled or died, as well as the annual cull rate, from 1989 to 2013 in the U.S. Source: USDA (2013)

 ,

Within-herd risk factors for culling

The risk of culling among cows within the same herd varies greatly. In general, pregnancy, higher milk production, younger age, and the absence of health issues such as metabolic problems, lameness or mastitis reduced the risk of culling. Figure 2 is the typical bath tub shaped risk profile for culling of non-pregnant (open) cows by days after calving in 727 U.S dairy herds with at least 100 cows from 2001 to 2006 (De Vries et al., 2010). The annual live cull rate in this study was 25.1% and annual death rate was 6.6%.

Figure 2 shows that the daily risk of culling increased with parity, except compared to fresh first lactation cows. The medium expected remaining lifetime for cows calving in parities 1 to 6 was 907, 697, 553, 469, 423, and 399 days, respectively. Noticeable is the increased risk of culling in the first 2 months after calving. The risk of culling at day 5 was typically lower than the risk of culling at day 30 after calving, except for first lactation cows. This could be related to greater risk of dystocia in first parity cows. Later in lactation, the risk increased again if the cow was not pregnant. This risk started to increase later for first parity cows (around day 300) compared to older cows (around day 220). This is likely due to a flatter lactation curve for first parity cows and therefore these cows are given more opportunities to get pregnant. The bath-tub risk profile was also shown by Fetrow et al. (2006) for Minnesota dairy herds and Dechow and Goodling (2008) for herds in Pennsylvania.

De Vries et al. (2010) also calculated a 3 to 7 times lower daily risks of culling for pregnant cows compared to open (non-pregnant) cows. The risk profile of pregnant cows was fairly even during the time of gestation, but with a sharp increase towards the estimated due date. This is likely due to culling around the time of calving while the actual calving date was not reported. Other reproductive factors that increased the risk of culling were greater calving difficulty (hazard ratio (HR) ≤ 1.95, where HR =1 is the baseline), and birth to males or twins (HR ≤ 1.36). Similar effects of reproduction traits on culling were shown for Canadian dairy cows (Sewalem et al., 2008). The importance of reproduction status was also reported by many others (e.g. Gröhn et al., 1998; Rajala-Schulz et al., 1999bc; Scheider et al., 2007)

Longer days to pregnancy in the previous lactation was also associated with increased risk of culling around the subsequent calving (Pinedo and De Vries, 2010). The risk of death from 14 days before the calculated due date to the first 60 days after calving increased from 2.5% to 5.8% when days to pregnancy increased from approximately 68 days to more than 301 days. The risk of live culling increased from 5% to 8.1% with the same increase in days open.

Early lactation cull risk varies greatly between herds. For example, in a study of Pennsylvania herds, Dechow and Goodling (2008) reported a cull risk for the first 60 DIM of 6.8% with a standard deviation of 4.6%. Considering that death most likely occurs in early lactation, Stone et al. (2006) reported a range of annual death rates between 3.5 and 16.8%, with an average of 8.1% in 20 New York herds.

 

Figure 2. Daily risk of culling in non-pregnant cows by days after calving and parity. Source: De Vries et al. (2010)

Relative higher milk yield within a herd is generally protective against culling (Gröhn et al., 1998; Beaudeau et al., 2000; Weigel et al., 2003; Pinedo and De Vries, 2010). However, the effect of milk production on culling risk varies with many factors. For example, Rajala-Schultz and Gröhn (1999b) found that in Finnish cows, milk production at the beginning of lactation did not have any effect on culling decisions, but later in lactation the highest producing cows were at the lowest risk of being culled. The effects of parity and late lactation pregnancy status on cull risk increased after adjusting for milk yield in their statistical model. The risk of culling decreased after adjusting for milk yield for diseases that clearly lower milk yield. In Chilean dairy herds, short and extended dry periods were associated increased overall culling when compared with the reference dry period, mostly because of lower milk production in the subsequent lactation (Pinedo et al. 2011).

Disease increases the risk of culling but the strength of the effects are not clear. In a review of the literature, Beaudeau et al. (2000) found that dystocia and udder disorders (mastitis and teat injury) had the most clear direct effects on the risk of culling. The effects of other diseases were less clear due to large variations in the results reported by the reviewed studies. The association between actual diseases and culling risk are often obscured because disease often leads to lower milk production which then is associated with increased risk of culling. Beaudeau et al. (2000) found that dairy farmers consider the most current disease events in their culling decisions, but gradually ignore past disease events. Studying Finnish Ayrshire cows, Rajala-Schultz and Gröhn (1999a) reported that all 15 studied diseases affected culling decisions mostly at the time of their occurrence with decreasing impact with time from the diagnosis. However, milk fewer, dystocia, and metritis also affected the risk of culling at the end of the lactation. Mastitis, teat injuries and lameness increased the risk of culling throughout the lactation. Following their review of the literature, Beaudeau et al. (2000) concluded that at least 50% of all culls are primarily declared as health-related.

Functional composition of the cow may also explain her risk of being culled. For US Jersey cows, Caraviello et al. (2003) found that cows with high scores for udder traits were less likely to get culled. Intermediate scores were desirable for rear leg set, dairy form, and strength. Stature, rump angle, and rump with had negligible effects on survival. Inbreeding slightly increased the risk of culling. There is probably less culling based solely on a cow’s physical appearance, such as udder and legs, than about 30 years ago. Much of this is due to genetic improvements in physical appearance over time in the cow population, such that poor conformation is less likely to exist (Dechow, 2013).

Culling is sometimes considered either voluntary or involuntary. Voluntary culling occurs when the primary reason for disposal is poor milk production but the cow is otherwise healthy and fertile. Involuntary culling occurs when the dairy farmer is “forced” to remove a productive, otherwise profitable cow, due to illness, injury, infertility of death (Weigel et al. 2003). Fetrow et al. (2006) discourages this distinction because disposal reasons are often multiple, including low milk production with other issues, and is an economic decision (although sometimes a straightforward one).

In a study of 2054 herds in the US, the primary reasons for culling (total = 100%) were: death (20.6%), reproduction (17.7%), injury or other reason (14.3%), low milk production (12.1%), mastitis (12.1%), feed and legs (8.1%), diseases (6.9%), udder (3.2%), and reason not reported (5.0%) (Pinedo and De Vries, 2010). Using 1.2 million records from 31,478 herds, USDA (2012) reported similar frequencies of given reasons for culling in 2012: reason not reported (30.3%), low milk production (20.7%), death (15.8%), reproduction (15.7%), mastitis or high cell count (12.5%), locomotion problems (4.8%), and bad behavior (0.1%). In the USDA data, the annual risk of death increased from 2.9% in the first parity to 10.5% in the 6th parity. Similarly, the risk increased from 0.8% to 4.0% for locomotion problems, from 5.5% to 9.3% for low production, from 4.1% to 6.9% for reproduction, and from 1.9% to 8.7% for mastitis of high cell count. In both datasets, death, low milk production and reproductive problems were the main factors that rank dairy cows for culling. The main reasons varied by stage of lactation as Pinedo et al. (2010) showed. In their study, death, injury and diseases were more often reported as main reasons for culling early in lactation. For example, the risk of death was at least 4 times greater in the first 30 days after calving than around 150 days after calving. This association was also reported by others (Thomsen and Houe, 2006). Hadley et al. (2006) reported that 42% of mortality occurred in the first 60 DIM. Low production and reproduction were more often the main reasons late in lactation. Only 57% of herds reported all 8 different disposal codes. Reported reasons for culling varied widely between herds (reviewed by Milan-Suazo et al., 1989; Beaudeau et al., 2000; Pinedo et al., 2010).

The association between primary culling reason and the actual health disorders recorded on the farm, including failure to get pregnant, was often biologically plausible (for example, cows with clinical mastitis were culled with reason “mastitis”), but in general weak (Beaudeau et al., 2000). In summary, Beaudeau et al. (2000) found the impact of diseases (adjusted for milk production) on longevity to be weak, compared to the stronger impact of low milk yield potential and poor reproductive performance.

Between-herd risk factors for culling

Several herd factors are associated with cow longevity, in addition to the ones presented later in table 1. For example, herds that were trying to expand (Hadley et al., 2006) or used crossbred cows had lower cull rates (Weigel and Barlass, 2003). In the study with larger dairy herds in the eastern U.S., the culling risk increased for cows in herds with shorter days to first insemination (HR ≤ 1.10) and herds that used a synchronized breeding program (HR ≤ 1.08) (De Vries et al., 2010). Hare et al. (2006) reviewed other studies that documented increased culling in extensive grazing (vs. moderate grazing), and confinement (vs. seasonal calving on pasture).

Within herds, culling reasons might shift. For example, an increase in on-farm mortality (euthanasia and death) in dairy herds has been reported in several countries in the last decade (Alvåsen et al., 2012). Death rate in Sweden was 6.6% in 2010. In Swedish dairy herds, higher mortality was associated with larger herd size, longer calving intervals, and herds that had Swedish Holstein as the predominant breed. Lower mortality was observed in herds with a higher herd average milk yield, during the fall and winter, and in organically managed herds. In a review of the literature, these authors (Alvåsen et al., 2012) reported that in general death risk (mortality rate) increased with an increased proportion of purchased cows, no pasture-grazing, larger herd sizes and lower herd average milk yield. These associations were confirmed by some other studies but not all (Thomson and Houe, 2006). Table 1 shows 19 dairy statistics compiled from 13357 herds primarily located in the eastern US and stratified by the annual cull rate (DRMS, 2013). The categories used were from 13% to 62%, with 7% increments. The average annual cull rate was 37.3% with a standard deviation of 11.8%. Smaller herds tended to have either a low or high cull rate where herds with average cull rates tended to be larger. Herds with lower cull rates increased in number of cows. On average 5% of the cows died per year, with an increase in death loss for herds with greater cull rates. Milk production was the highest for the herds within the 42 to 49% cull rate category. Herds with the highest cull rate had the most calvings per cow present. On the other hand, reproduction was slightly better in herds with the lowest cull rates. Herds around the average cull rate also were less likely to use natural service sires.

Table 1. Statistics for 13357 U.S. dairy herds on DHI milk test, sorted by % cows left per year 

 .

Source: DRMS (2013). Available at www.drms.org Accessed May 9, 2013

These results show that there is not a strong association between annual cull rate (cow longevity), and herd factors such as level of milk production, milk quality, and reproduction. The preceding section briefly described that relatively high milk production, pregnancy, low parity, and the absence of disease are protective against culling and therefore within a herd extend a cow’s longevity. These risk factors do not directly relate to the herd level cull rate. For example, herds with good fertility do not necessarily have lower cull rates as table 1 shows.

Dairy farmers also consider factors other than a cow’s performance or health when deciding to cull the cow. It suggests that dairy farmers make active, economic culling decisions that are not necessary forced upon them (Fetrow et al., 2006). These results are in agreement with Beaudeau et al. (2000) who report that herd characteristics such as the availability of heifers, quota, and farmer’s attitude towards risk and uncertainty change the risk that a cow is being culled for a given disease. Therefore, cow longevity is determined by (economic) factors external to the cow itself. In the next sections we will explore how culling decisions and longevity might change when some of these factors such as prices and availability of heifers are varied.

Economic principles: replacement cost

Currently, USDA values one month longer productive life at $35 in their genetic selection indices lifetime Net Merit, (NM$) Cheese Merit and Fluid Merit (Cole et al., 2009). This economic value is largely determined as (calving heifer cost - salvage value) / productive life, although the exact inputs have not been published. For example, ($1800 - $600) / 36 months = $1200 / 36 = $33. Figure 3 shows how these values change depending on starting productive life and the difference between calving heifer cost and cull value. These variables will vary from time to time and from farm to farm.

Figure 3. Valuation of 1 month extra cow longevity per cow, calculated as (replacement cost – cull value) / productive life. These values do not include the value of the cow performance between first calving and culling.

The $35 used in the NM$ selection index is strictly the value of the long term replacement cost. It does not include the additional cost or revenue from keeping the animals longer. For example, a reduction of the incidence of metabolic diseases might increase milk yield and fertility, which have value on their own in addition to their effect on culling and replacement cost. Thus, decreased replacement cost is just one component of the full value of extending cow longevity. The effects on milk sales, feed cost, calf values, and other variable cost need to be considered as well. The $35 also does not help when ranking cows currently in the herd for replacement decisions.

Ranking cows for culling decisions

Ranking cows for replacement decisions, as well as determining the value of cow longevity, is not a trivial task. Assuming a constant herd size, the marginal future profit of a present cow has to be weighed against the average profit of a replacement young cow that could be obtained in that period (Zeddies, 1972). Thus, the economic criterion of the replacement decision in a herd with constant size is: a cow of a particular age should be kept in the herd as long as her expected marginal profit is higher than the expected average profit during a replacing young cow's life (Renkema and Stelwagen, 1979). The value of extending cow longevity, either through cow factors such as fertility or external factors such as prices, therefore needs to include the opportunity cost of postponed replacement. This opportunity cost is defined as the profit sacrificed on an average replacement cow by keeping the current cow in the herd instead of the replacement (Van Arendonk, 1991). Lehenbauer and Oltjen (1998) point out the difficulty in determining the appropriate opportunity cost of space because dairy cows are often grouped in pens or replacement animals are not available, or other constraints apply. Results from a linear programming analysis shown later illustrate that these factors indeed affect cow culling.

A large number of computer models have been developed that optimize culling decisions as well as rank cows for future profitability, also known as retention pay-off (RPO). Examples are Smith (1973), Van Arendonk and Dijkhuizen (1985), Kristensen (1989), De Vries (2004, 2008), Nielsen et al. (2010), and Cha et al. (2010). Optimized culling decisions are assumed to give better estimates of costs of for example diseases. These optimization models commonly assume that profit per cow per time period is maximized in a herd of a fixed number of cows with an unlimited supply of replacement heifers. Often the optimal cull rate (and hence cow longevity) was not the main focus of these studies.

The referenced models calculate RPOs for individual cows that are used to rank cows for culling. The RPO is defined as the net present value of keeping the incumbent cow in the herd until the optimal time of replacement, compared with replacing her now with a replacement heifer. For both the incumbent cow and the challenging heifers and her replacements, future cash flow projections need to be made. An RPO of <$0 means that the incumbent cow should now be replaced. If the RPO is for example $500, then replacing the incumbent cow now with a calving heifer amounts to a loss of $500. Figure 4 shows RPOs by days since calving for first parity cows with low (90% of average), average, and high (110% of average) milk production throughout the lactation. The RPOs for both open cows and cows that became pregnant on day 61 after calving are shown. The patterns in figure 4 suggest that the RPOs of open cows decrease as they fail to become pregnant over time. This decrease is primarily a function of the shape of the lactation curve. The low producing cow reaches a negative RPO 6 months earlier than the high producing cow. Pregnant cows are more valuable than open cows (except for early gestation in the high producing cow which suggests delayed breeding is advantageous) and the difference increases with stage of lactation. The RPO is generally also lower for older cows and pregnant cows that conceive later in lactation.

Figure 4. Retention pay-offs (RPO) by days since calving in monthly increments for first parity cows with low (90% of average), average, and high (110% of average) milk production throughout the lactation. The cow should be replaced when her RPO < $0. The RPOs for both open cows and cows that became pregnant on day 61 after calving are shown. Results were calculated with the model of De Vries (2008) with inputs for a typical dairy farm in the US in 2013.

Figure 5. Retention pay-offs by days since calving for average first parity cows for 3 levels of cull prices: $1.20, $1.60, and $2.00 per kg live body weight. This equals $658, $890, and $1124 per culled cow, respectively. The retention pay offs for both open cows and cows that became pregnant on day 61 after calving are shown. Results were calculated with the model of De Vries (2008) with inputs for a typical dairy farm in the US in 2013.

Figure 5 shows how the RPO decreases when the beef price for culled cows increases. The graphs for the average cow in figure 4 and the cow with a $1.20 beef price in figure 5 are the same. Cull prices of $1.20 per kg body weight are more traditional in the US, but lately prices as high as $2.00 have been paid.

Assuming a dressing percentage of 55% (warm carcass compared to shrunk live weight), $2.00 per kg live body weight equals $3.64 per kg warm carcass weight. It is expected that cull prices will remain high in the foreseeing future because of the historical low beef cow population in the US. Higher beef prices clearly reduce the RPO and make the RPO for open cows become negative earlier, thus shortening cow longevity.

When the RPO of open cows reaches $0, the milk sales minus the variable cost for that cow that day are typically still positive, in the order of $1 to $3 (De Vries, 2009). This illustrates that cow longevity is reduced by perhaps one month under an optimal replacement policy compared to a policy where open cows are kept as along as their milk income minus variable costs are positive (they pay for themselves”). However, when milk prices are temporarily very low (resulting in negative profitability), the RPO can still be positive when milk income minus variable cost is $0. Such cows do not generate enough milk sales to cover their variable costs any longer. But there is still a chance that they get pregnant and reach the next lactation. Although these pregnant cows would have to go through an extended period of low milk production, at a very low milk price, this option is still less costly than culling the cow and replacing her with a heifer. The farm would have to decide if it wants to remain in business until the milk prices increase again.

This general RPO is an estimate of the value of keeping the incumbent cow in the herd, accounting for future risks of forced culling, and assuming that an average replacement heifer will take the place of the culled cow as soon as her RPO becomes negative. The RPO also represents the expenses that could be made to keep the cow in the herd if she faced a health problem. The economic loss of a dead cow is equal to her RPO at the time of death plus the missed beef income if the cow was culled alive.

The negative RPO represents the opportunity cost for keeping the cow one decision stage too long (one month in the model of De Vries (2008)). Therefore, negative RPOs are generally greater than -$200 for open cows. This is also a reflection of the milk sales during that month. If a cow was kept for example 5 months too long, the opportunity cost would be approximately the monthly negative RPO multiplied by 5. Groen et al. (1997) explained that the cost of extending herd life of the present cow is given by the profit of the (best) alternative use (which is the opportunity cost), which is keeping the replacement heifer.

The model of De Vries (2008) does not consider specific diseases, but for example Cha et al. (2010, 2011) studied the RPOs and treatment/cull decisions for cows with different types of lameness and mastitis. Because lameness and mastitis (temporarily) decrease cow performance, the RPOs of ill cows are lower than for healthy cows. Their expected cow longevity is also reduced.

When the herd is subject to a milk quota, the optimizing criterion is profit per unit of milk produced. Kristensen (1989) showed that under a milk quota, low producing and late lactation open cows would be kept much longer, thus extending herd cow longevity significantly. Further, differences in milk yield and stage of lactation would have less effect on the RPO of cows, essentially reducing the variation in RPOs that would exist among cows a herd. The ranking of cows also changes compared to a situation without milk quota. Given these theoretical results, one would expect that the cull risk for low milk production is lower in herds under milk quota (for example Europe) compared to in herds where no milk quota exist (for example US).

The RPO can be used to estimate the cost of early lactation mortality. Given that most death occurs early in lactation and no income is received, the cost of a death of an early lactation open cow is the cost of a replacement heifer minus the calf value for a first parity cow, say $1500. For an older cow, the value is slightly less depending on age, for example $800. The loss is at least equal to the salvage value. Assuming the loss is $1000 for an average death, a reduction in annual death rate from 7% down to 3% would be a savings of $40 per cow per year.

The referenced optimization models assume that there is no shortage or surplus of replacement heifers and optimize profit per stall. The assumption of fixed herd size is common. All mentioned optimization models also allow for simulation of herd statistics, such as herd profit or cull rate or productive life, given a (non)optimal replacement policy. Table 2 is an example of herd statistics calculated with the model of De Vries (2008) including updated prices under typical US conditions and an optimal, unrestricted, replacement policy. Sensitivity analyses were performed with cull prices ($1.20 to $2.00 per kg body weight) and 21-day service rates (45% to 75%) as a measure of reproductive performance. Key findings are the sharp increase in annual cull rate from 35% to 57% with higher cull prices, and a small decrease in cull rate (35% to 31%) with increased reproduction. Thus, higher beef prices reduced cow longevity and better reproduction increases cow longevity. These findings are consisted with results from sensitivity analysis found in the reference optimization models. However, in practice improved reproductive efficiency is generally not associated with lower cull rates (extended cow longevity).

Table 2. Herd statistics for an optimal cow replacement model under US conditions depending on cow cull price or 21-day service rate.

 .

Source: model of De Vries, 2008 with updated prices for 2013

Continuing with the updated model of De Vries (2008), a reduction in forced cull rate for the first 3 months only by 50% and 100% resulted in gains in profit of $37 and $77 per cow per year. These gains likely underestimate the true benefits of reducing fresh cow culling through improved management and cow comfort because effects on milk production and fertility are not included in these gains.

In all sensitivity analyses in table 2, there is a discrepancy in the number of heifers that could be raised (if all heifer calves were kept, including a 15% loss from calving to age at first calving) and the number of heifers that enter the herd to replace culled cows. This situation resembles a herd which sells all heifer calves and purchases calving heifers (or raises all heifers and sells surplus heifers not entering the herd). At least at a national scale in the US, virtually all dairy heifer calves are raised and enter the milking herd because the alternative, selling dairy heifers to feedlots for beef production, has traditionally been less profitable (Fetrow et al., 2006). Also at the farm level, most US dairy farmers raise all heifers, with some willing to sell surplus heifers and others purchasing these surplus heifers. Dairy farmers who manage to raise many dairy heifers often will cull more low producing cows to make room when these heifers calve.

The value of improving traits, such as cow longevity, is therefore dependent on the farm production system, such as a milk quota, a closed or open herd (Kristensen, 1989; Harris and Freeman, 1993; Groen et al., 1997). For example, Harris and Freeman (1993) found that the value of (genetically) extended cow longevity (reduced risk of forced culling) at least doubled under a milk quota system compared to under a fixed herd size constraint. One reason given was that under a milk quota, the value of genetic gain in milk yield traits is lower, therefore reduced culling does not slow down the importance of genetic progress in yield traits as much as when no milk quota is present. Genetic improvement in replacement heifers generally should not affect individual culling decisions much, however. Kristensen (1989) calculated a lower optimal cull rate for herds producing under milk quota compared to the situation where herd size is the limiting factor. One would expect cow longevity to be higher in herds in the European Union and Canada.

Linear programming to study cow longevity under herd constraints

The foregoing review and data analysis has made clear that the productive life of a dairy cow in a herd depends to a large extend on external factors such as production quotas, heifer availability, and prices, and assuming dairy producers make optimal profit maximizing decisions. Improving a cow’s ability to avoid forced culling through improved health, higher milk production, or better fertility does not necessarily translate into an equivalent reduced herd cull rate and therefore increased cow longevity.

To further study the effects of herd constraints such as quota and closed herds on cow longevity given optimal insemination and replacement decisions, we have developed a large linear program. The model is similar to the dynamic programming model referenced earlier in that it exists in over 45,000 states for which optimal insemination and replacement decisions are made. A key advantage of this model is that insemination and replacement decisions for individual cows are dependent on decisions for other cows and the number of dairy heifer calves that are born. For example, the model will find out if it is more economical to raise a dairy heifer and calve her into the herd, and cull a used cow to make room for the heifer, or sell the heifer and keep the cow longer. A constraint can be for example that heifer calves are never sold (unless for forced reasons such as health) but are all brought into the herd. This would likely increase the cow cull rate. A parlor capacity constraint or environmental constraint also affects optimal insemination and replacement decisions. Yet another type of constraint is allowing the use of sexed semen to be considered for optimal insemination decisions. These various constrains all affect annual cull rates and therefore cow longevity.

The effects of changes in inputs on herd annual cull rate (and thus average cow longevity) were studied for 2 x 2 main herd constraints. The first is a herd with 1000 milking slots which mimics a limiting parlor capacity (common in the US) or the use of milking robots. The second constraint is the amount of manure produced by the herd, including by youngstock and cows to mimic an environmental constraint. Within each main constraint, the first secondary constraint was the sale of all born calves. Replacement heifers were purchased one month before their calving date. This assumption is similar to the dynamic programming models referenced earlier. The second secondary constraint was a closed herd where all born dairy heifers calves were raised and entered the milking herd to replace culled cows. Within these 4 constraints, the default scenarios assumed breeding with only conventional dairy semen and unrestricted insemination and replacement decisions for cows. Youngstock was not sold (except forced culls). Thus, decision making for youngstock was excluded in the scenarios shown here. Five other scenarios were evaluated with each of the 4 constraints: The effect of 1) greater milk production, 2) higher cull price for culled cows, 3) no model-decided culling allowed if daily milk yield >15 kg/day, 4) availability of sexed semen (at a higher price and with reduced fertility) and beef semen, and 5) greater probability of conception in cows. Some main results are shown in tables 3 and 4.

Table 3 shows increased cow longevity with keeping cows longer, and increased probability of conceptions. Greater milk production per cow and higher cull prices (from $1.20/kg body weight to $2.00/kg body weight) reduced cow longevity but profit per milking slot was increased. These trends are similar as calculated with the referenced dynamic programming models (which assume a fixed cow herd size, including dry cows). There are no marked differences between the milking cow constraint and the manure constraint. Notice that the amount of manure produced in the default scenarios was the same, but the insemination and replacement policies were different. The chosen prices are such that when sexed and beef semen were allowed, all cows were inseminated with beef semen only. Some delayed breeding took place in high producing cows in the first parities (not shown). There were also low producing cows that failed to get pregnant that were not immediately replaced.

In table 4 all dairy heifer calves that are born were necessarily raised (excluding some forced culling) and brought into the herd when they calved. Cow longevity was shorter than when all calves were sold and replacement heifer purchased as in table 3, but longer under the manure quota than under the milking constraint quote. Higher cull prices did not affect cow longevity much. Days open was under the manure constraint markedly longer than under the milking slot constraint, again due to some delayed breeding in high producing cows. Scenarios where heifers were raised but surplus 8-month old heifers could be sold at a small resulted in cow longevity that was in between both main constraints shown here.

These results from the linear program illustrate once more that cow longevity should be extended for the relatively better producing animals in the herd, but overall cow longevity is strongly dependent on economic constraints that are external to the intrinsic cow performance. Extension of one cow’s longevity often should result in the reduction of another cow’s longevity in the herd if herd profitability is the optimizing criterion. The ranking of cows for culling decisions, such as by an RPO method, under these various herd constraints likely does not change much, but should be studied more.

Table 3. Effects of variations in dairy model inputs on cow longevity and other results for the constraints milking slots and manure quota. Further constraints were that all calves were sold and springing heifers purchased were one month before calving to replace culled cows.

 

 

Table 4. Effects of variations in dairy model inputs on cow longevity and other results for the constraints milking slots and manure quota. Further constraints were that all dairy heifer calves born in the herd were raised to replace culled cows and no heifer purchases were allowed.


Take home message

1. Cow longevity depends on intrinsic and extrinsic factors.

2. Survival analysis has shown that a large number of cows leave the herd early in lactation largely due to metabolic health reasons.

3. The risk of death is greatest early in lactation.

4. Forced culling of cows early in lactation is expensive, in the order of $500 to $1000 per cow.

5. Efforts to reduce death rates and improve early lactation health, and therefore intrinsic cow longevity, are to be profitable.

6. The average cow longevity is determined by extrinsic economic factors unrelated to an individual cow’s health and performance. To extend the average cow longevity, dairy farmers would need to be motivated to bring fewer heifers into the herd and therefore cull fewer cows.

  • This could be accomplished by creating fewer dairy heifer calves, for example through delayed inseminations in heifers and/or cows, or not breeding non-pregnant cows that are late in lactation.
  • A portion of the heifers and cows could also be inseminated with beef semen and the crossbred calves could be sold.

References

1. Alvåsen, K., M. Jansson Mörk, C. Hallén Sandgren, P. T. Thomsen, and U. Emanuelson. 2012. Herd-level risk factors associated with cow mortality in Swedish dairy herds. J. Dairy Sci. 95:4352-4362.

2. Beaudeau, F., H. Seegers, V. Ducrocq, C. Fourichon, and N. Bareille. 2000. Effect of health disorders on culling in dairy cows: a review and critical discussion. Ann. Zootech. (Paris) 49:293-311.

3. Cannon, C. Y. and E. N. Hansen. 1939. Expectation of life in dairy cows. J. Dairy Sci. 22: 1025-1032.

4. Caraviello, D. Z., K. A. Weigel, and D. Gianola. 2003. Analysis of the relationship between type traits, inbreeding, and functional survival in Jersey cattle using a Weibull proportional hazards model. J. Dairy Sci. 86:2984-2989.

5. Cha, E., J. A. Hertl, D. Bar, and Y. T. Gröhn. 2010. The cost of different types of lameness in dairy cows calculated by dynamic programming. Prev. Vet. Med. 97:1-8.

6. Cha, E., D. Bar, J. A. Hertl, L. W. Tauer, G. Bennett, R. N. González, Y. H. Schukken, F. L. Welcome, and Y. T. Gröhn. 2011. The cost and management of different types of clinical mastitis in dairy cows estimated by dynamic programming. J. Dairy Sci. 94:4476-4487.

7. Cole, J. B., P. M. VanRaden, and Multi-State Project S-1040. 2009. Net merit as a measure of lifetime profit: 2010 revision. AIPL Research Report NM$4 (12-09). Available at http://aipl.arsusda.gov/reference/nmcalc.htm Accessed May 8, 2013.

8. Dairy Records Management Systems (DRMS). 2013. Dairy Metrics. Available at http://www.drms.org. Accessed April 10, 2013.

9. De Vries, A. 2004. Economics of delayed replacement when cow performance is seasonal. J. Dairy Sci. 87:2947-2958.

10. De Vries, A. 2008. Economic value of pregnancy in dairy cattle. J. Dairy Sci. 89:3876-3885.

11. De Vries, A. 2009. Ranking cows for culling decisions. Pages 20-28 in: Proceedings of the Southeast Dairy Herd Management Conference. Macon, Georgia, November 11-12. Available at http://www.animal.ufl.edu/devries Accessed May 10, 2013.

12. De Vries, A., J. D. Olson, and P. J. Pinedo. 2010. Reproductive risk factors for culling and productive life in large dairy herds in the eastern United States between 2001 and 2006. J. Dairy Sci. 93:613-623.

13. Dechow, C. 2013. Culling patterns shift productive life estimates. Hoard’s Dairyman, January 25, 2013.

14. Dechow, C. D. and R. C. Goodling. 2008. Mortality, culling by sixty days in milk, and production profiles in high- and low-survival Pennsylvania herds. J. Dairy Sci. 91:4630-4639.

15. Fetrow J., K. V. Nordlund, and H. D. Norman. 2006. Invited review: Culling: nomenclature, definitions, and recommendations. J. Dairy Sci. 89:1896-1905.

16. Groen, A.F., T. Steine, J-J. Colleau, J. Pedersen, J. Pribyl, and N. Reinsch. 1997. Economic values in dairy cattle breeding, with special reference to functional traits. Report of an EAAP-working group. Livest. Proc. Sci. 49:1-21.

17. Gröhn, Y. T., S. W. Eicker, V. Ducrocq, and J. A. Hertl. 1998. Effect of diseases on the culling of Holstein dairy cows in New York state. J. Dairy Sci. 81:966-978.

18. Hadley, G. L., C. A. Wolf, and S. B. Harsh. 2006. Dairy cattle culling patterns, explanations, and implications. J. Dairy Sci. 89:2286-2296.

19. Hare, E., H. D. Norman, and J.R. Wright. 2006. Survival rates and productive herd life of dairy cattle in the United States. J. Dairy Sci. 89:3713-3720.

20. Harris, B. L. and A. E. Freeman. 1993. Economic weights for milk yield traits and herd life under various economic conditions and production quotas. J. Dairy Sci. 76:868-879.

21. Heinrichs, J. and C. Jones. 2013. Age at calving is on the decent. Hoard’s Dairyman. April 25. Page 305.

22. Kristensen, A. R. 1989. Optimal replacement and ranking of dairy cows under milk quotas. Acta. Agric. Scand. 39:311-318.

23. Lehenbauer, T. W. and J. W. Oltjen. 1998. Dairy cow culling strategies: making economical culling decisions. J. Dairy Sci. 81:264-271.

24. Milan-Suazo, F., H. N. Erb, and R. D. Smith. 1988. Descriptive epidemiology of culling in dairy cows from 34 herds in New York state. Prev. Vet. Med. 6:243-251.

25. Nennich, T. D., J. H. Harrion, L. M. VanWieringen, D. Meyer, A. J. Heinrichs, W. P. Weiss, N. R. St-Pierre, R. L. Kincaid, D. L. Davidson, and E. Block. 2005. Prediction of manure and nutrient excretion from dairy cattle. J. Dairy Sci. 88:3721-3733.

26. Nielsen, L.R., E. Jorgensen, A. R. Kristensen, and S. Ostergaard. 2010. Optimal replacement policies for dairy cows based on daily yield measurements. J. Dairy Sci. 93:75-92.

27. Pinedo, P., C. Risco, and P. Melendez. 2011. A retrospective study on the association between different lengths of the dry period and subclinical mastitis, milk yield, reproductive performance, and culling in Chilean dairy cows. J. Dairy Sci. 94:106-115.

28. Pinedo, P. J., and A. De Vries. 2010. Effect of days to conception in the previous lactation on the risk of death and live culling around calving. J. Dairy Sci. 93:968-977.

29. Rajala-Schultz, P. J. and Y. T. Gröhn. 1999a. Culling of dairy cows. Part I. Effects of diseases on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41:195-208.

30. Rajala-Schultz, P. J. and Y. T. Gröhn. 1999b. Culling of dairy cows. Part II. Effects of diseases and reproductive performance on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41:279-294.

31. Rajala-Schultz, P. J. and Y. T. Gröhn. 1999c. Culling of dairy cows. Part III. Effects of diseases, pregnancy status and milk yield on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41:295-309.

32. Renkema, J.A. and J. Stelwagen, 1979. Economic evaluation of replacement rates in dairy herds. I. Reduction of replacement rates through improved health. Livest. Prod. Sci. 6:15-27.

33. Schneider, M. del P., E. Strandberg, U. Emanuelson, K. Grandinson, and A. Roth. 2007. The effect of veterinary-treated clinical mastitis and pregnancy status on culling in Swedish dairy herds. Prev. Vet. Med. 80:179-192.

34. Sewalem, A., F. Miglior, G. J. Kistemaker, P. Sullivan, and B. J. Van Doormaal. 2008. Relationship between reproduction traits and functional longevity in Canadian dairy cattle. J. Dairy Sci. 91:1660-1668.

35. Smith, B. J. 1973. Dynamic programming of the dairy cow replacement problem. Am. J. Agric. Econ. 55:100-104.

36. Stone, W. C., C. L. Guard, and D. A. Button. 2006. Why do cows die? Hoard’s Dairyman 151:117.

37. Thomson, P. T., and H. Houe. 2006. Dairy cow mortality. A review. Vet. Quart. 28:122-129.

38. USDA. 2007. Dairy 2007, Part I: Reference of Dairy Cattle Health and Management Practices in the United States, 2007 USDA-APHIS-VS, CEAH. Fort Collins, CO. #N480.1007.

39. USDA, 2012. Reasons that cows in Dairy Herd Improvement programs exit the herd (2012). Available at http://aipl.arsusda.gov/publish/dhi/current/cullall.html Accessed April 30, 2013.

40. USDA-AIPL, 2013. Trend in productive life for Holstein or Red & White. Available at http://aipl.arsusda.gov/eval/summary/trend.cfm Accessed April 30, 2013.

41. USDA, 2013. Dairy statistics. Available at http://future.aae.wisc.edu Accessed April 30, 2013.

42. Van Arendonk, J. A. M. and A. A. Dijkhuizen. 1985. Studies on the replacement policies in dairy cattle. III. Influence of variation in reproduction and production. Livest. Prod. Sci. 13:333-349.

43. Weigel, K. A. and K. A. Barlass. 2003. Results of a producer survey regarding crossbreeding on US dairy farms. J. Dairy Sci. 86:4148-4154.

44. Weigel, K. A., R. W. Palmer, and D. Z. Caraviello. 2003. Investigation of factors affecting voluntary and involuntary culling in expanding herds in Wisconsin using survival analysis. J. Dairy Sci. 86:1482-1486.

45. Zeddies, J. 1972. Ökonomische Entscheidungshilfen für die Selektion in Milchviehherden. Züchtungskunde 44:149–171.

Author

Albert de Vries

Albert de Vries
5 articles

Associate Professor, Department of Animal Sciences, University of Florida

Read more about Albert De Vries here

Link to Florida Dairy Extension

Read more »

University of Florida

University of Florida

Longevity Conference Proceedings Index

More information

Find more information about the Longevity Conference 2013 here

Summary of the Longevity conference

Read the conference summary here

Proceedings

Read the whole proceedings here (pdf download)

Videos from conference

Check out the videos from the Longevity conference here