Saturday, September 30, 2017

“Custom Indexes” or what we used to call “Matrix Selection”


The previous generation of dairy geneticists who gifted the world with the “selection index” must be rolling in their graves.    These were the fellows who actually told us “just breed on Net Merit ranking and all your other problems will go away.”    At that time “Net Merit” was based 70% on PD Milk and 30% of PD Butterfat—no protein, no type, no fertility, no health trait measurements.  All those things were added after the cow population began to suffer in all of those areas (some of which those same geneticists tried to claim were “management” traits rather than heritable behavior, as we now know them to be).

Consistent through all the formulation and recalculation and creation of new measurements (some of which are quite arcane, perhaps less than 5% heritable) – the advice was always, “don’t focus on some specific list of traits, just let the selection index do the job”.    Indexes were promoted as “more efficient” than a dairyman making his own list of desired traits and setting minimum levels of trait expression he would accept—a practice known as “matrix” selection.    

For the 55% of dairymen who still use a “matrix” approach, the latest advice – “calculate your own index” (according to the needs of your current herd) – is a vindication.   So as not to offend their old college profs, these new geneticists just avoid the word “matrix”.   But in fact they are now telling us, “composite index selection is obsolete” (even though these are the basis for all sire selection being done under Genomics, which are calculated purely to the various national index formulas).    

Just ask Nate Elzinga of Zeeland, MI, one of Michigan’s most milk productive dairymen, how he chooses his sires.   His office computer is programmed to recalculate the data from every sire AI offers him, to the weighting of traits he devised matched the heritable traits he wished to address in his herd of cows.   Or ask Josh DeHaan of Wayland, MI, how much he was able to accomplish in improving all fertility characteristics of their expanding herd by a focus on health and fertility traits alongside matings to create balanced physiques.   Many of your neighbors may be requiring plus bf% and pr% traits who wish to raise their milk price and find they are $2.00 per cwt above their breed average after only two generations.

The famous Wally Lindskoog of Arlinda Farms, Turlock, CA shortly before his passing told an interviewer that he could see the future of AI developing many specialty AI companies, whose sire selection would not be based on an ”index” but on the genetic preferences of like- minded dairymen of similar geography or facility design and milk market.    We have seen this process occurring ever since Alta Genetics purchased Landmark Genetics.

How do you design your own herd index?

First, you look at your own economic and management data.    What elements of cow care and management are costing you the most?   Focus on whatever heritable measures may have an impact on those issues.    For example, if conception rates are an issue, then establish a minimum trait value for daughter pregnancy rate (not semen conception rates, which have no impact on your cows’ natural fertility capability).    If you lose too many calves at birth look at stillbirth rate; if you lose too many heifers to hard calvings, look at both sire direct and daughter calving ease.    Establishing dollar values for each of these types of financial loss helps to put all of your herd data in perspective of where genetic change could help.

No comments:

Post a Comment