Animal
Model Basics
|
|
In the summer of 1989, the animal model
became the basis for genetic evaluations for Jersey bulls and cows in the
United States. The model provides joint and simultaneous evaluations for
bulls and cows based on female records. The basic animal model includes
three factors that combine to result in a cow’s record:
Figure 1. Animal Model Components
The Lactation Calculation To take this concept one step further, the animal model considers the parts of a cow's lactation as follows: Yield = M + C + P + A + EIn this equation,
When we add all the parts together, we get the production for that lactation. The Genetic Calculation Dairy bulls do not have records themselves but are tied by genetic relationship (1/2 of genes) to their daughters that have records. These daughters are also tied back to their dams. These ties allow for the prediction of transmitting ability for sires from daughter records or from their son’s genetic evaluations. This cow family information is incorporated directly into all genetic evaluations through genetic relationships. It also provides for a correction of the daughter’s records for how good or bad the mates of a particular bull. For example when a bull is mated to the best cows he should not receive credit for genetic effects that the cows transmit to the daughters. A bull’s Predicted Transmitting Ability (PTA) is calculated from the PTAs of his daughters and his pedigree influence. The extent of the pedigree influence is decreased as the bull adds more daughters. The PTAs of the cows are based on a deviation from their contemporaries which is based on a mature equivalent value and again a pedigree influence. In both the case of the bull and the cow, the pedigree influence adds reliability to the PTAs. The pedigree adjustment is especially valuable to a bull’s proof early on as it can help temper the effect (good or bad) that one individual can have on a proof with minimal daughters. Specifically, three pieces of information are used to calculate PTA values on females:
To be able to understand how these pieces work together to calculate a Predicted Transmitting Ability, we must fully under-stand Yield Deviation (YD). The YD is a weighted average of a cow’s production, after accounting for different management and environmental factors. Basically it’s what the cow herself contributes to her PTA based on her performance. The components used to calculate YD are:
If a cow has one completed record, Yield Deviation is calculated with the formula: YD = ME - M - P - C The YD is expressed as a deviation from the U.S. Jersey population (in contrast to the herd). This is done so that the genetic level of the herd can also be adjusted for in the animal model when it is calculating the M, P and C. Thus, the YDs are not the same as herdmate deviations that are calculated by the DHI processing centers. Remember however, the differences between Yield Deviations for two cows in the same herd is the same whether expressed on a population or herd basis. Combining YD with information from relatives (Parent Average and Progeny information) then allows us to predict the PTA. Young Cow Calculation For a young cow that does not have progeny, her Predicted Transmitting Ability value is a function of the Parent Average and Yield Deviation. PTA = [X1 x PA] + [X2 x (YD/2)] where X1 and X2 are weighting factors that depend on the amount of information available. These factors can get rather complex, but they are .78 and .22 respectively for a cow with one completed record and no progeny. A quick example may shed some light on the process. If you have a Jersey cow with no progeny and one record as follows:
plugging into the formula, the Yield Deviation is:
Granted, this is an extremely simplified example, but hopefully it gives one an appreciation for how the computer goes about putting together the various pieces of information on millions of cows to obtain cow evaluations for genetic merit. |
||||||||||||
