F. R. Allaire
Department of Animal Sciences
The debate continues on how much genetic selection emphasis should be applied to support cow production longevity. While most breeders consider the pillars (e.g., efficient reproductive function and sound udders, feet, and legs) of cow longevity vital to their quest for more efficient milk production, scientific indexes still reinforce the paramount importance of annual milk yield per cow. This article identifies some genetic and economic conditions that make productive-life2 even more important in an economic index, such as the USDA's Net Merit.
Numerous terms have been used to describe cow longevity. One is length of herd life, measured in months from the first calving date to the cull date. The current industry-wide average is about 36 months of herd life following first calving. A similar longevity measure and useful in genetic selection is the PL index that the USDA calculates to assist breeders in improving longevity. The reported PL value is given in months of life up to 84 months and is expressed as a difference from the average bull whose daughters were born in 1990. A typical range for bulls is -2.2 to +3.4 months.
Those familiar with the USDA's Net Merit index will recognize that 4 and 10 are the relative weights for PL and production, respectively. The weights combine the two evaluations into an index for economic merit. For the sake of this article, no reference is made to the third genetic evaluation, somatic cells score, which is also in the USDA's Net Merit index. The weights of 4 and 10 have been justified by research summarized by VanRaden et al. (1993). These weights combine production and PL so the ranking of sires or cows for in choices that support overall economic merit.
Breeders often are surprised to learn that this scientifically constructed index does not place more relative weight on longevity than it does. This article summarizes our efforts to identify some economic and genetic conditions that could lead to justify the use of a relatively larger weight for PL. For example, what conditions need to exist for the weight to be 6 or 8, instead of 4? Once breeders understand the conditions under which longevity becomes more important, a greater appreciation occurs about why the indexes are as constructed. Recognizing what different genetic relationships in cow populations or what different herd cost structures lead to a greater weight of 6 for PL would help us to better understand the importance of the background conditions on constructing the Net Merit index.
The debate on the proper emphasis for PL always may continue. The application of population-based selection policy is very difficult to reconcile with day-to-day experiences of breeders. For example, a selection policy that values PL at 4 compared to a 10 for production is difficult to accept when so many breeders experience the loss of a large fraction of cows for reasons other than low production. The loss from these non-production reasons is about three times more frequent than for loss from low production. This article places the construction of economic indexes within the broader context of current herd production cost relationships and cow genetic relationships so PL evaluations will be understood better and applied.
Selection emphasis is a finite resource in the pursuit of genetic change in two or more traits. Consider the need to select 50 young sampling bulls. The 50 bulls highest in PL will not be the same 50 highest in MFP$3, or the same 50 highest in a combined index of PL and MFP$. Less than perfect correlation between PL and MFP$ results in a different set of 50 bulls, and the choice of a specific set of 50 bulls depends on the best allocation of emphasis between PL and MFP$. The best allocation is determined from the genetic responsiveness to selection and to the economic consequences of any genetic change in each trait toward enhancing milk production economic efficiency. Genetic responsiveness is influenced by the underlying genetic relationships, including correlation and heritabilities. The downstream economic consequences of genetic change also are influenced by how combinations of genetic changes in each can lower the cost of production most rapidly. The best economic interests of dairy herd operations should be considered when implementing simultaneous genetic change in production and PL.
A change in the PL weight from 4 to 8 may not appear to be significant on the surface, but proportionately, a greater weight will allocate more of the finite selection resource toward further increasing the genetic basis of US dairy cow longevity and lead simultaneously to a comparatively slower rate of genetic gain in production. The background genetic and economic conditions must justify a change in selection emphasis if gains in dairy cow economic efficiency are to be continued.
It has been well documented (Allaire and Gibson, 1992; Rogers et al., 1988; Van Arendonk, 1985) that the most important dairy herd cost factors that influence the importance of further change in cow herd life, beyond the current average of 36 months after first calving, are those related to 1) the cost difference between average heifer replacement rearing cost (about $1000 to $1200) and the salvage value (negative cost of about $600) of the cow culled to make space for the entering replacement and 2) annual fixed cost (about $900) of keeping a productive cow in the herd. The later expense factor (annual fixed cost) for this study included expenses associated with maintaining cow herd facilities, general labor, and the average annual expense of keeping each cow productive. Fixed cost included feed for cow maintenance but did not include the feed expense that was used directly to produce milk. The feed expense necessary to produce milk was assumed to be broadly proportional to the amount of milk produced.
The different herd cost structures and genetic relationships under which the emphasis would be more or less than the current ratio of 10:4 used by the USDA are shown in Table 1 (adapted from Lobo and Allaire, 1995).
The ratio, net replacement cost (or the difference between the cost of a replacement heifer and the value of the replaced cull cow) to the annual fixed cost per cow, is central to allocating selection emphasis between PL and production (Allaire and Gibson, 1992). The cost values themselves, although obviously important to herd net income, are not important to how production and PL are combined in an economic index. The ratio of these costs is what is critical in deriving the proper selection emphasis.
A larger relative weight is assigned to PL as the net cost of supplying the herd with a cow replacement increases relative to the annual fixed cost of keeping a cow productive in the herd. The selection emphasis changes by only about one point over a very extreme set of economic conditions. The ranges examined were: net-replacement costs increase from $600 to $1200 and fixed-cow cost decreased from $900 to $450. This means that the different herd cost structures which favor more selection emphasis on PL are minor compared to the alternative background genetic relationships also shown in Table 1.
| Table 1. Different herd cost structures and genetic relationships. | |||||
| Net Merit index weight | Industry-wide dairy herd cost structures | Genetic Relationships | |||
|
Production:
productive-life |
Relative emphasis on
productive-life |
Net replacement cost/Fixed-cow cost | Correlation
Production & non-productive cull traits |
Heritability
Non-productive cull traits | |
| 10 : 7 | More | $1200 / $450 | zero | 10% | |
| 10 : 6 | . . . | $ 750 / $750
or $ 600 / $900 |
zero | 10% | |
| 10 : 5 | . . . | $1200 / $450 | -25% | 20% | |
| 10 : 4 | Current1 | $ 750 / $750
or $ 600 / $900 |
-25% | 20% | |
| 10 : 3.5 | . . . | $1200 / $450 | -25% | 10% | |
| 10 : 3 | Less | $ 750 / $750
or $ 600 / $900 |
-25% | 10% | |
| 1 Current weights used in the USDA's Net Merit index. | |||||
Results show that a great change - from current ratio of about 2:3 in the net-replacement cost to fixed-cow costs to about 3:1 - would be needed to increase the emphases on PL. A 3:1 ratio, for example, would indicate that net replacement and fixed-cow costs would need to approach those found in New Zealand.
Any potential to enhance economic efficiency must be routed through genetic improvements in those non-production traits that support longer herd life. Given what is known about selection's ability to change any combination of the non-production traits, little selection potential is available to place even more emphasis on PL. This is true, primarily, because the heritability of most non-production cull traits is very low and, secondly, because these traits have only a secondary role in influencing cost efficiency. This means that the route of influence is long and indirect between genetic change in non-production traits and change in efficiency of cow herd milk production costs.
There is strong evidence that as per-cow production increases for a herd, the expenses associated with veterinary care and breeding also increase. If some of this antagonistic relationship arises from a genetic basis, i.e., genes favoring production more frequently coexist with other genes that limit those traits that support long PL, then the most effective approach to increasing efficiency is by raising the production level in the first or second lactation when health care costs are least. On a cow-population basis, this circumvents the negative high annual production effects on the expected cow PL.
Does the rising cost of per-dairy-cow herd-veterinary services reflect an indication of a mismatch between the population's genetic complement and the care received in our modern dairy production systems? This question can be answered only by a careful analysis of the array of veterinary and health care provided within the production systems and the animal welfare goals of breeders as cows are managed for high production.
Some options to reduce the adverse relationship between production and health care costs might include the following. First, more sensitivity to pending health care needs of susceptible cows may help circumvent later health care costs. Cow care managers could provide facilities, nutrition, or other lifetime productive cycle design changes that achieve a reduction in health care needs as annual production level per cow increases. For example, perhaps high production cattle can benefit in cost efficiency terms when they have long calving intervals compared to average cows. A second possibility might be to raise the standards in disease avoidance. For example, "normal" animal welfare standards might be that health care or genetic bases might be designed to reduce the frequency of certain diseases, e.g., mastitis or metabolic disease, to lower the need for greater health care in the highest production-potential cows.
Prevailing and potential antagonistic genetic relationship between a production and non-production basis in PL will lead to even more selection emphasis on production as the industry pursues more cost efficient milk production. Production is more responsive to selection than is PL to selection. Breeders may need to sacrifice slower production gains in order to make gains in the non-production traits that lengthen PL. Alternatively, greater quality in cow care management schemes may become necessary if both production and PL levels are to be enhanced.
A great change in herd cost structures would be required to put as little as a +1 weight, e.g., from 4 to 5, on PL.
3 MFP$ denotes the USDA calculation of the genetic evaluation of predicted transmitting ability for combined revenue from milk, fat, and protein yields.
Allaire, F.R., and J.P. Gibson. 1992. Genetic value of herd life adjusted for milk production. J. Dairy Sci. 75:1349.
Lobo, C.H., and F.R. Allaire. 1995. The Effect of Alternate Economic and Genetic Covariation Structures on the Relative Economic Gain from Selection using Stayability Records. J. Dairy Sci. 78: (10) 2299.
Rogers, G.W., J.A.M. Van Arendonk, and B.T. McDaniel 1988. Influences of involuntary culling on optimal rates and annualized net revenue. J. Dairy Sci. 71:3463.
VanRaden, P.M., M.M. Schutz, and G.R. Wiggans. 1993. Calculation of USDA-DHIA Net Merit, October 8th. Anim. Improv. Prog. Lab., ARS, USDA, Beltsville, MD.
Van Arendonk, J.A.M. 1985. Studies on the replacement policies in dairy cattle. II. Optimum policy and influences on changes in production and prices. Livest. Prod. Sci. 13:101.