Wednesday, June 14, 2017

Are you getting what you want when you mate cows on “index rank”?

From the November December 2015 Dairy Route Letter

Since the introduction of Genomic testing, most AI systems have geared up their selling of linear mating to capture all your business, utilizing Genomic sires whenever possible so that their averages can appear better against competing sales efforts.

Only one new (of many old) problems with this—not only has University research continually proven that linear-based mating adds no value, the data on how Genomic ranked sires perform once daughters are milking in your herd indicates that while the “index” rank averages out, the performance on many desirable traits does not.    Thus, if you are buying into an “index ranking” mating service, you could easily go backwards from the resulting heifers it produces.     Here is why this can happen:

Example:   Lifetime Net Merit
The formula for this index has changed dramatically through its time, given its first format was simply 70% Milk (lbs) + 30% Fat (lbs)—a “single trait selection” approach that set commercial dairymen up for decades of “inbreeding depression” (loss of fertility and longevity).    The latest version of $NM is still heavy of “health and fertility” and is calculated this way:
Butterfat (lbs) 22% + Protein (lbs) 20% + Productive Life (mo’s) 19% + Udder Composite 8% + SCS* 7%  + DPR 7% + Body Size* 5% + Calving Ease 5% + Foot & Leg Composite 3% + Milk (lbs)* 1% + HCR 2% + CCR 1%.    [Traits marked * are preferred negatively]

In other words, it is possible to go multiple generations in Net Merit rankings and never use a sire who improves milk volume (this relates to how multiple component pricing determines our pay prices).    It is also possible, and perhaps more long-term detrimental, to never use a sire with normal cow size.

Dr David Kendall, Geneticist for ST Genetics, is suggesting (and we agree) that dairymen should use their own personal “index” based upon their current weak points in genetic traits, rather than pay extra for any sire just because he ranks highly (today) on any “one size fits all” index.     Simple is better.

Examples of the “Kendall approach” that fit real-world herd situations

Your herd averages 90 pounds per day (on 3x milking which is supposed to improve udder health).   In spite of that, your SCC is costing you quality premiums.   It takes 4 straws per cow to get pregnancies so calving intervals are stretched out past 400 days.    Milk price is low due to low bf% and lower pr%.

The selection traits to focus on are:

SCS – 3.00 is average, so seek out bulls well below 3.00 and you will see lower SCS in new heifers.
DPR-- +0.0 is average, but DPR runs counter to PTA Milk yield, so seek out bulls better than average
(-1.00 above +1500m) (-0.25 above 1000m) (0.50 above 500m) (1.25 below 500m) to gain fertility.
bf% and pr% -- cows in a negative energy state will convert protein to energy in the rumen trying to meet body demands.    PTA pr% should be emphasized whenever you see cows below 3.0% (HO) to 3.3% (JE) in early months of lactation.    PTA bf% should be emphasized on any cows below 3.6% (HO) to 4.4% (JE) as these levels are below the averages of the Federal Milk Order for components.

You will find that cows with higher test% (pr and bf) have less trouble with negative energy and thus will also have shorter calving intervals than higher yield, lower test% cows.              
Looking beyond index rank to the “numbers” undertneath

To keep this fair, I am only going to use Genomic sires we offer (from International Protein Sires and ST Genetics) to do these comparisons.

EXAMPLE:                         2371 GTPI  vs  2370 GTPI            (both $ 18.00)

566HO1217 Synergy PULASKI                            54HO 754 Mr Sunview Coin SUNFISH *RC

                    2371 G-TPI     (aAa 351426)                                 2370 G-TPI       (aAa 231456)
PTA Milk:  + 523                                                     PTA Milk:  +1834                 (1311 pounds more)
PTA Fat + Protein:  +105                                         PTA Fat + Protein:   +109     (essentially equal)
SCS   2.81                                                                SCS  2.75                               (both look good)
Udder Composite  + 1.43                                         Udder Composite  + 1.72      (statistically better)
Foot & leg Comp   - 0.07                                         Foot & Leg Comp + 0.84      (clearly better)

While both of these are from deep pedigreed cow lines, there have been more successful AI sires from cows behind “Sunfish *RC” than so far behind “Pulaski”.    Our choice to stock for your use remains “Sunfish *RC” given more people want high milk gains with equal high dollar value protein gains and understand that heifers who milk like that take one more cycle to breed back (- 1.1 DPR) but cow family longevity (at lower SCS) suggests these will be fertile cows over the long haul.

EXAMPLE:                    $798 G-Net Merit  vs  $723 PE-Net Merit 

151HO 696 Mr Shot DOZER 1491                      552HO2451 De-Su RANSOM

                 $798 G-NM      (aAa 423156)                                   $723 PE-NM        (aAa 342165)
PTA Milk:  +1452                                                   PTA Milk:  + 365                  (1087 pounds less)
PTA Fat:       -.02%  + 49                                        PTA Fat:       +.14%  + 50    (higher bf test %)
PTA Protein:+.01%  + 48                                       PTA Protein: +.08%  + 34     (higher pr test %)
DPR:          + 2.9                                                     DPR:           + 2.9                   (possibly equal?)
PL:             + 9.3      (breed elite)                           PL:              + 8.6                   (insignificant diff)
SCS             2.59                                                    SCS              2.80                   (both promising)
+1.52 Type  +1.65 UDC  +1.01 FLC                     +1.32 Type  +1.05 UDC  +2.20 FLC
calving ease  6.7%                                                  calving ease  5.5%                 (better or real calves)

What is the key difference between these two?    Progeny data on “Ransom”.   Two areas in which Genomics has not been too successful: (1) predicting calving ease, (2) assuming high milk bulls can still be really good on DPR (daughter pregnancy rate).     Exceptional FLC on “Ransom”.    Genomic type is based on a more limited set of markers; the trend is for type data to be lower in reality than in the theoretical G-based world.      The promising Genomic sire is higher for $NM, but the progeny data coming in on “Ransom” indicates he is a well-rounded improver, his $NM not as biased by assumptions in the index model for specific traits.



The calculation systems favor the newest “Genomic” bulls over the older “High Reliability %” bulls with extensive progeny data.     Have reasons beyond just “index” for any Genomic sires you use.

No comments:

Post a Comment