Saturday, December 6, 2014

The Process of Mating and Selection for an on-site Herd Sire


                 (December 2009)             by   Greg Palen                 Netherhall (polled, grazing) Jerseys


Why Raise any Bull?

Artificial Insemination was first organized in the USA during the 1940s, to accomplish several goals of dairy cattlemen:  (a)  Make dairying safer  (bulls can be dangerous, both to people and cows);  (b)  Eradicate venereal diseases (bulls can spread trouble, like brucellosis, vibriosis, and trichinosis, all causing infectious abortion);  (c)  Improved herd management  (have more exact dates for breeding, thus dry off and due dates);  (d) Put a productive cow in the bull’s space (incremental income gain, lower net reproductive expenses);  (e)  Provide genetic variety (avoid having all your “eggs” in one “basket” of unknown genetic value).

After six decades of commercial AI activity, however, many dairy farms continue to use herd bulls, or have reverted to their use after trying AI and not always succeeding.     Problems that lead to natural service reproduction include:  (a)  No one on the dairy has good insemination skills,  (b)  Commercial AI service is not available in that area,  (c)  Personnel lack the skill for effective heat detection,  (d)  Facility design does not acccomodate AI easily,  (e)   Preferred breed of cattle is not available within any AI system,  (f)  Selection traits of interest to herdowner are not considered important by those in AI who are selecting which bulls enter AI service,  (g)  Dairy suffers under a skilled labor shortage,  (h)  Herdowner has a passion for genetic selection and wishes to develop his own bloodline(s).

While many of these problems can be solved, not all herdowners wish to invest the time and money, or have elected to defer the introduction of AI to a later time, after more immediate needs are met.   It is also not uncommon for many dairies to use AI in tandem with natural service, for example:

Seasonal calving windows under grazing management

In this scenario, where animals live in paddocks (rather than barns) and calving is not desired all year long, progressive graziers will focus AI on two to three cow cycles, first introducing the service bull to the virgin heifers, and then moving him with the cows to “clean up” any still open after 4-6 weeks’ AI.
This program focuses heat detection on cows being milked twice daily and thus closely observed, and gets heifers (under less observation) to calve in the same season as the cows.

High group – Low group feeding in confinement

In this scenario, the AI activity is focused on the fresher cows, who are grouped in the same pens and thus are all “open”, most are cycling, and will be more likely to exhibit detectable heats (known as the “dormitory effect”) from a consequence of “gang” cycling activity.

Pregnant cows are moved from the “fresh” group to a “bred” [low, ie, not as fresh] group, thus hardly any cycling activity is going to be detected.    Putting a bull in such a group insures “someone” is still doing heat detection, to catch cows that reabsorb, or who are moved there after repeated AI attempts.


Gang bull breeding
Some very large herds will insert multiple bulls into large cow groups to stimulate earlier repro.
 
Why not buy such bulls from a higher-profile breeder?

In most cases, this is what we do, and it comprises a steady source of added income to the purebred breeders in dairy communities.    As these are guys investing a lot of time and money into the type classification and official milk testing of their cows, maintaining accurate ancestral identity, perhaps utilizing some added repro technology (embryo transfer, cloning, genomic testing, etc) and focusing their semen purchases on the “elite” ranked sires of their breed, we just assume their cattle have more transmitting ability (capability of genetic improvement) than our own.    After all, these are the herds producing bulls for the various AI systems, who promote higher “genetic value” sires.

The dairyman not utilizing AI but who believes the AI sires are clearly superior, is highly likely to do just that—in fact many ET full brothers to AI sires end up as “jumper” bulls in dairy herds.   This is a natural economic consequence of the propogative technologies (super ovulation and embryo transfer) as well as the further level of screening AI systems now do from the availability of Genomic testing (DNA mapping) which is reducing the number of “full brothers” being sampled by AI studs.    While more pedigree (sire x dam) combinations are resulting from Genomic screening, this also means more surplus sires are being propogated.     [Studs expecting to sample 240 young sires annually are writing 1200 sire contracts, but only taking one out of five bull calves produced, based upon which one has the highest Genomic estimates of “genetic value”.]

The addition of Genomic testing to the sorting of ET sires may imply the bulls left for natural service do not possess the desired DNA markers.    This should give us pause—do we wish to use a “reject”?

Reasons you may prefer to raise your own herd sire


First, there is the herd health issue.    You do not have a “closed” herd (closed to any possible disease exposure that could come from new animal contact) if you keep bringing in bulls from other farms, in which the health status and disease exposure is different from yours.    Raising your own herd sires is a way to more fully reach a “closed” herd health status.

Next, there is the issue of “selection trait focus”.    You may not agree with the trait selection priorities of those breeders accessible to you and raising jumper bulls.     This is an issue that is growing bigger each year.    For example:

Organic dairy production.    Certified organic producers need “healthy” (disease resistant) and self- reliant cattle.     Some of the higher genetic value animals, bred to more extreme performance levels, demand a higher level of feed supplementation and daily labor care to maintain their productivity, or in more cases, their reproductivity.   In an organic production model, where antibiotics and hormone therapies are not allowed, such genetics may result in higher than desirable cull rates.

Polled heads.    Dehorning as a dairy farm practice is already an “animal rights” issue in Europe, thus an increasing veterinary expense for EEC producers.    Likewise, in subtropical dairy areas, where no winter occurs to break parasite cycles, polled heads avoid a lot of wound infections and growth rate setbacks.    Graziers calving outside in tight seasonal windows, raising spring calves in paddocks, find the dehorning job to be onerous, and occurring in the “fly season”.    Thus, demand for polled cattle increases, while the supply of polled AI sires grows very slowly (and is mostly heterozygous polled, ie, still producing 50% horned calves in horned herds). 


Specialty milk production.     One of the developing market niches is for “A2A2 Beta Casein” milk, a genetically-determined milk quality claimed to have health benefits to those recovering from cancer.

Likewise, a dairyman who has entered on-farm artisan cheese production for retail marketing of his milk, may wish to select for “BB Kappa Casein” milk, a genetically-determined milk quality that will raise cheese yields 10% to 15% over conventional blend milk at the same levels of bf% and pr%.

Why are premium gene traits for milk composition mostly ignored by AI?

While most AI systems test for Beta Casein and Kappa Casein genotypes as part of the screening tests for Genomic evaluation, the data is neither routinely published nor does it have much impact on which sires are chosen to enter AI service.    Likewise, sires rarely get the nod for AI due to being polled, as long as some other bull has a higher ranking on some screening index.    Commercial AI selection is driven on “ranking indexes” rather than trait selection matrixes, virtually worldwide.    

The justification for ignoring unique selection traits is mostly driven by the commodity focus of milk cooperative marketing, where milk from different (and unique) farms is “pooled” – first by the milk haulers, next in the silos of the receiving plant--  any market demand for a specialty milk composition is only provided if it can be accomplished at the balancing plant by separation and reconstitution.   As local AI systems merged into regional and national entities, their interest in providing all sires for all local tastes (breeds) (bloodlines) (trait mixes) has declined, looking only at broad market statistics to make decisions as to future sire selection preferences.

Most purebred breeders raising bulls are thinking about the AI market, not you


Purebred breeders using OvSynch, superovulation and embryo transfer, Genomic testing their cows, are focused on the AI sire paradigm.    They wish to recover those costs from premium bull sales to AI studs and premium embryo sales to other breeders wishing to enter this “index” driven market.

Thus, buying a bull from them to breed your cows is like using year from certified bank-run seed to plant a wheat crop.    You did not get the “best” genetics but you got a close approximation of those genetics.    Thus the question is as follows:   Do I want to breed cows just like the commercial AI cows for my herd?    If so, save time and buy your purebred neighbor’s left over ET bulls.

If, on the other hand, you have more specific selection goals, or a more unique focus in milk quality production, and your management environment and/or milk marketing differs from the commercial, commodity definition, you either have to raise your own herd sire  or  seek breeders who already are focused on producing the sort of genetic mix you wish to gain.

In saying this, I am firmly identifying myself as a dairy industry contrarian, in that I do not have a blanket belief that AI sires are automatically superior to what you could raise.    I remain a firm believer in AI as a useful tool that can help you become a better stockman, as well as a more profitable businessman.   When all the costs are accounted, basic AI technology saves you money over average natural service results.     We sometimes just have trouble recognizing what those costs are.

The large number of custom semen collection businesses (and not just in areas driven by beef breed cattle breeders) suggests that many today are utilizing AI, but from their “own” stable of sires.
 
You really have to have sound reasons for raising your own herd sires, to bear the added costs.   In some cases, specific (unique) genetic goals justify the added costs.    In other cases, the opportunity to develop a niche market for bulls from your farm can justify the added costs.    In either case, you may still find that you still wish to have access to AI technology as well as sires you can source via AI, to get all of your genetic “bases” covered.      

Where do I start to produce a useful herd sire?

Dairymen milk cows.     98% of all dairymen produce an income stream based entirely on what cows produce (milk, replacement heifers, deacon bull calves, and cull salvage).     Only 2% of all dairymen generate meaningful farm income from breeding bulls, and only a fraction of that 2% are able to sell bulls into AI systems on a regular basis.

A useful herd sire thus has to be produced from the perspective of what is a useful cow.    (This should be the first point at which you start to question the current market fixation over “sire index” rankings.)  

What do you mean by a “useful” cow?     Simply put, a cow is “useful” if she can successfully do all the functions you ask of her each season.    Can she calve a live calf?   Does she get up and get going as a milk producing cow after calving?    Will she rebreed at the time your calendar prefers?    Does she avoid added costs that reduce the profitability of her production (vet assisted births, milk fever, metritis, ketosis, displacements, hoof trimming, chronic lameness, mastitis, induced repro after failure of natural heat insemination, long dry periods, fatty liver post-calving)?    Are her calves vigorous or prone to pneumonia, scours, finicky eating habits, poor at making transitions?     Does she have the type to enhance your herd equity and make daily milking easier?    Does she have the physique and behavior to adapt to seasonal or structural or nutritional changes in her environment?

It should be evident from such a list, that our traditional measures of cows—DHIA milk test records and Breed Society type scoring—only scratch the surface of answering the “useful” question.   But a good on-farm record keeping system designed to observe and record “usefulness” will make a sound basis for the selection of cows who can produce “useful” herd sires.

Analyze your “useful” cows from a “lifetime performance” basis

The dairyman just starting out can have difficulty determining whom his most “useful” cows are, as it takes more than a single lactation to answer all those questions posed above.   It is for this reason that many traditional stock breeders preferred the oldest successful cows to be “bull mothers”.    This flies in the face of the “newest generation is best” fast turnover of pedigrees in commercial AI selection, a market that routinely buys bull calves from first lactation cows.   But bear with me on this.

Each year is different.    Different weather produces differences in forage qualities and quantities, can impact on daily comfort of cows, can affect the prices we are receiving for milk and produce years in which we tighten our belts and reduce input uses to those we can produce internally.     If you judge a cow on a single year of performance, you will overlook factors of adaptability—was it a good year to be a cow at your place, or a tough year?      Basically, analyzing cows over an accumulating lifetime of performance is a better guide to answering all the questions posed regarding usefulness, even if you are not convinced of the economic advantage many of us assume for cows with longevity.

If you find you are only saving bulls in years where performance came easily, you will not be putting enough selection pressure on adaptive and longevity characteristics.    Such bulls will produce milky heifers but may not possess the genetic qualities to make them equally useful in difficult years.   This is ultimately why we sometimes see most of our mature cows culling out in a single year—we did not produce them from selections (and matings) that took longevity and adaptability into full account, and they could not survive all the challenges thrown at them in that year.    Lesson one:  Save bull calves from “survivors” – not just from pretty heifers.

Cow families develop around more useful cows.

For the dairyman who has had his herd awhile, has thought about breeding quality and done some trait selection, a symptom of success will be the existence of cow families in the herd.    A “cow family” is an extended line of maternal relationships.    It can be a multiple generation of cows descending from a single older cow; it can be a group of cows all descendent from maternal siblings.    The point that we need to see is this—a cow who has successful fertility genes and normal herdife survival will produce more heifers in her lifetime than a cow of mediocre fertility and average herdlife.   These heifers in turn, if bred to the same or higher level of ability, will also produce more heifers.    “Cow families” accumulate in soundly managed dairy herds as a result of superior genetic fertility and longevity.   In the experience of those breeders who were concerned about this, it is consistent that these “maternal” traits would follow cow lines directly, and thus sire lines indirectly (as cow fertility genes are easier to verify from the maternal side of pedigrees).

Bull fertility is pretty simple compared to cow fertility.    All the bull has to do is develop a healthy set of testicles, not be overconditioned when young to avoid fatty tissue in the testes, learn to jump-mount and have enough libido to either serve cows or collect semen.   In the case of AI, his ejaculate must have the volume of sperm cells, and those sperm cells must have the vigor to survive the freezing and thawing process under conventional (1/2 cc straw) packaging.    These are the relatively simple factors of bull fertility.    AI studs feed high energy rations to bulls on full collection schedules so as to keep a stable level of body condition as a support to their libido, and restrict their collection to a two or three day per week schedule determined by the volume of sperm production.

Cow fertility is more complex.    Cows, on the other hand, have to fit their reproduction around cycles of ovarian activity that are driven by hormones produced in various glands.   Cows do not “produce” eggs like bulls produce sperm cells—they are born with a lifetime supply in their ovaries, and instead their cycles are designed to mature and then release an egg according to their body schedules.   All of this has to occur around the physiologic demands of milk production and the post-calving recovery of the uterus.    Periodic calving alters and ultimately ages the cow’s body in ways that sperm production never could for any bull.     This is true for all mammals, as the fetus is incubated within the body of the mother of the species.   (The father is just a passive spectator by comparison).

Natural service bull fertility is different in execution from AI service fertility.    Over decades of the advancing technology for semen collection and sperm preservation, we have advanced from chilled, liquid semen (good for a few days) to frozen semen (storable indefinitely)—we have advanced in the freezing knowledge from fat ampules (that killed a majority of sperm cells) to thin straws (that save a slim majority of sperm cells), with two consequences:  (1)  Straw technology enhanced the conception rates of marginal sperm quality sires;  (2)  Volume semen production created a preference for sires who will serve the same animal repeatedly and deplete their semen reserves at collection frequency.
 
In a natural service environment, the preferred bull behavior would be to serve the cow in heat  once and then walk away, seeking the next cow in heat.    This means the bull holds semen in reserve for a potentially next cow in heat the same day, and does not expend his systemic energy in those repeated mounts that, in a herd setting, lead to a loss of body condition followed by a loss of libido.

AI is not fond of such sires, because they do not want bulls that have to be collected daily to harvest all the semen they can produce.    This interferes with their full schedule of collection and processing of a larger stable of bulls.    Has this had negative fertility consequences?     To date, no one knows.    But the documentable fact is that sire conception rate (SCR=  bull semen fertility) and daughter pregnancy rate (DPR= cow fertility) are not very highly correlated.     
   
Are such bull behaviors genetically linked?      Yes, because to some extent all points of deviation (measurable difference) among animals within a herd are definable as genetic differences.    But as these differences were not important to AI success, they have never been summarized and evaluated.   Thus they remain within the realm of observational knowledge—the sort of data scientists distrust on the basis of limited sampling sizes (most people are not making useful observations of their cattle, or at least fail to write them down for future collation, or fail to submit them to an “official” summary).

The beef breeding industry routinely tests bull behavior by summarizing the percentage of cows each bull “covers” in a specific time period when turned out for range service.    Generally, we know that beef breeds seem to have more fertility than dairy cattle.    Part of the real reason would be that more attention is placed on fertility as a desirable selection trait, and that a multiple of measures both male and female are collated to estimate genetic fertility ability.

Successful fertility implies a living calf.     The beef guy knows this, as his income is based on the size of his calf crop first, how it grows after birth is a secondary level of income stimulation.    Likewise it is true for the pig farmer (litter size, livability of the piglets, sow acceptance of piglets for nursing) and for the sheep farmer (live lambs, even when twinning, means more income than dead lambs).    But for some perverse reason, an earlier generation of dairy geneticists (active in the formative era of AI and the key promoters of composite index ranking) seemed to have overlooked the link from reproduction to production—choosing to focus purely on comparative lactation production.   This is where most of our issues with unsuccessful fertility in modern dairy cows originated.

Thus ease of calving in an indirect sense, as it correlates to both cow survival and calf survival, is also linked to cow fertility.     You now have several points from which to decide how “useful” your cows are in a genetic sense of potential transmitting ability.    I would not save a bull from any cow who has a history of difficulty in calving or of presenting stillborn calves.    In all of animal agriculture, income derives from reproduction, as the precursor to measured production.      Keep that in your focus.

Successful fertility is positively linked to production profitability.     This is where the statistical data on which sires have been ranked leads us astray.     Think about your own dairy operation: is it actual pounds per cow per day that determines your milk check income, or is it predicted lactation pounds in 305 days?      You might think the two are essentially the same—but they are not.

I have a friend with purebred Jerseys whose selection focus has consistently sought two things: higher component %s (butterfat and protein) to drive his income, and annual calving intervals over multiple lactations (cow lifetime totals) to drive his profitability.    We will analyze his results:


He currently has a herd average just shy of 17,000 pounds of milk with 5.6%bf and 4.0%pr—thus 950 pounds fat and 680 pounds protein (on an ECM basis, this matches a 26,000# Holstein herd).   But the amazing thing is he averages nearly a 12 month calving interval on sixty milking cows at this level of nutrient energy conversion—a truly “elite” level of performance, as we will demonstrate.

He has a neighbor milking 120 Jerseys, with a 20,000 pound DHIA herd average, whose annual milk shipments are nearly exactly twice what Phil ships.     Why, if Phil’s cows are producing at the same level as a 20,000-pound herd, is he only getting lactation credit for a 17,000-pound herd?

The difference is—his neighbor accepts a 14 month calving interval as a consequence of a preference for “high peak day” genetics over “persistency” genetics.     He not only uses AI sires, he uses those of the “high PTA milk” persuasion—those who daughters are predicted at higher lactation pounds in 305 measured days, but whose reproduction is delayed (thus total lactation is 365 days).

17,000 pounds of milk in 305 days equals  55.74  pounds of milk actual per day.
20,000 pounds of milk in 365 days equals  54.79  pounds of milk actual per day.

For three decades we have been taught the rolling herd average was your measure of success in dairy. 
RHA is calculated as average test day pounds times a 365 day year.    But we kept on calculating bull evaluations on a 305 day lactation value – and in Mature Equivalent, not actual pounds.

This created a selection preference for a higher peak lactation curve (in which that peak could extend as a consequence of slow or defective fertility) over a flatter, more persistent lactation curve (which is ultimately the most profitable, ie, persistency of production meant fewer below-cost stale days at the end of the lactation, thus a controlled-length dry period).    

Earlier feeding approaches also expressed a preference for delayed fertility.    In the 1970s and into the 1980s (a very important transition period in genetic evaluation concepts) the standard nutritionist’s   approach to feeding cows was to “challenge feed” fresh cows—an added pound of grain for each three pounds of milk, and see how high she will “peak”, on the assumption that a higher peak will then carry her steadily declining daily yield further into the end of the lactation, when actual production did not always cover total feed costs.    We were taught that a “good” cow makes half her total lactation in the first 130 days, and that pregnancy would make her yields drop, eventually into a “loss” column, at which point you dry them up (no matter how many days until the next calving).

This was a strategy designed to sell more grain (and grain was cheap and plentiful in those days, so why not?   The grain ration could make up for a deficient forage base).     The sort of cow that makes milk out of grain, of course, had to be sorted from the cow that gains weight from added grain.   Thus in genetic selection, we redesigned type standards to prefer the more “angular” cow over the easier conditioning cow—and along with that, accepted a longer period of “negative energy deficit” in the early months of lactation, thus delaying the reproductive response.    These changes in cow structure and performance began slowly (usually masked over by a steep decline in bf% tests, not considered important in that time period as “fluid” milk was preferred over “manufactured” milk products in a society obsessed by “low fat” —the loss in bf% a consequence of internal energy rationing by cows).

On the genetic side, while thinking in PD pounds, we were actually selecting genes that regulate the conversion of energy intake into four bodily uses:  production, reproduction and health maintenance.
 
The genetic anchor to fertility was lost when the high-indexing young heifer replaced the long-lifetime production cow as the “bull mother” of choice.    Long lifetimes were not possible without fertility – but a “hot” first lactation is more easily made in the absence of reproduction.    Thus, if reproduction is no longer a drag on energy intake, then we could select for higher levels of protein %, and still have an increasing volume yield plane.     As long as we introduced OvSynch, we could continue to get at least some cows rebred, and stay focused on pushing the envelope for individual cow yield volume.

The purebred sector focused on producing AI sires uniformly adopted “induced” fertility—ie, super ovulation and embryo transfer.    The natural fertility of these “hot” young first lactation cows was never tested, because “induced” fertility could produce more than a lifetime of calves within two years of embryo marketability.     (No one ever considered this may not be the fertility preference of typical dairymen—as usual with newer technologies, we assumed they rendered traditional ways “obsolete”.)

For too long, the genetic community patted itself on the back for increasing the “base change” for milk yield in every breed (every five years, as adjusted by USDA) – until it was clear that dairymen were not always happy with the higher costs of reproduction, slower success of reproduction, producing a faster herd turnover, and in many cases a shortage of replacements.    Some tried crossbreeding as the “solution” to restoring “vigor”—others just quit AI and reverted to natural service.    Neither of these have to date altered the overall cost of production, thus milk production profitability remains marginal.

Genes truly lost can never be replaced.     One of the consequences of the “index” era has been a rapid destruction of bloodlines, followed by a decline in sireline variety.     All AI sires today are related to each other on a passive level, but the “linebreeding” of any sire line to develop a more homozygous pattern in trait transmission is avoided due to phobias over “inbreeding”.    Thus it is more difficult to create heterozygous variation within a purebred population, otherwise known as “hybrid vigor”—and it is more difficult to sustain “hybrid vigor” within a crossbreeding scheme, due to the lack of patterns of homozygosity unrelated to other breeds (able to produce heterozygous variation) when crossed.

Basically, “genes” are either present or absent in the DNA.    Genes do not “dilute”.     If the cow lines most superior for fertility or health maintenance have been culled from the population (on the basis of being “noncompetitive” under prior single-trait index rankings, as bull mothers) (on the basis of plain old “old age” culling them form the breeding herds) – their genes are lost.     Thus the blithely stated assumptions that “all you need are plus DPR and plus PL sires to restore prior levels of fertility” both overstates the genetic ability of current AI sire lines and understates the time and culling rates needed to recapture prior levels of performance in these traits.

But the lesson of my friend Phil’s Jersey herd—in which homebred sires comprise half of the ongoing sire selection and their selection is based upon rigorous adherence to desired trait levels—proves that it can be done…  IF you stay focused.

Sunday, November 23, 2014

Further thoughts on the evolving science of genomic testing and evaluation


A year ago we introduced this topic with what was known at the time.    Since then, we are seeing varying levels of AI stud adoption, from using G tests only to choose young sires, to assimilating G tested sires right into their main sire lineups as if they were fully “proven”.       We have some observations drawn from recent conversations around the world.

So how have G tested sires fared in New Zealand (where G tested sires hit the market in 2005) ?

The two major AI systems in New Zealand are LIC New Zealand (who breeds 80% of the cows) and CRV Ambreed (who breeds 10% of the cows).     Looking at the top six rated sires for each AI system over five years of progeny produced (from 2001 through 2006 calvings), the following happened:

Overestimation on EBV Protein:    Holstein:   25% for LIC, 11% for CRV.     Jersey:   35% for LIC, 19% for CRV.
Overestimation on BW index:        Holstein:   30% for LIC, 11% for CRV.     Jersey:   20% for LIC,   9% for CRV.

These twelve bulls (each breed) sired a total of 500,000 progeny annually that were tested.   So Rel% for all sires (assigning 75% to the G tests) ended up at 99% Rel (progeny data replacing the G estimates).
These results (consistently overstating the superiority of the sires G tested as at “the top”) have driven the Kiwi’s back to their microscopes and computers to figure out why Genomics failed as a predictor.   One of their conclusions was that higher density SNPs were required… and always, more data…

So what are the latest developments incorporated into the August 2009 predictions here?

An individual cow of interest (who has produced a polled Red & White son) has the following history of Genomic evaluations:     April 2009:  GPTA  $251 Net Merit:      August 2009:  GPTA $453 Net merit!
Over $200 gained in “Net Merit” rank, without any change in her type scores or lactation records.

What changed?    After all, her Genomic SNP (direct measurement of gene enzymes and sequences) would be the same (genes never change from conception to demise).    But this particular cow has a German AI sire—foreign G test results were not included by USDA in April, but were in August.   In other words, although the G test is a “direct” look at the individual genotype, we still are interpreting the results in part, on the parentage of the animals tested.     

Should not the accumulating body of G tests, establishing “marker” genes for each evaluated trait, be the determinant of a Genomic value, rather than a continuing reliance on the simplistic (lower Rel%) Parent Average calculations?    After all, we had already determined “G” tests were 70% Rel on milk and 60% Rel on type, on the first go-rounds (January 2009). 

I posed this question to Dr Curt Van Tassel, of AIPL-USDA, who is deeply involved in the computer estimation of trait values from Genomic testing.    His point gave clarity to why we could see such a change on a cow:   “…we can only predict genetic values accurately for animals that are represented in the data.  What I mean… is that if we have not seen an animal with a genotype like the one we are trying to predict, we don’t have a great ability to predict that genetic value.”    The less common a pedigree, the more estimation falls back on Parent Averages.

This goes a long way toward explaining why we keep seeing “the same” pedigree combinations on those G tested young sires who have the “elite” PTA estimations.     Those pesky genes still do not wear name tags that say, “I am the milk gene…”     Thus USDA researchers are hoping to implement new genotype platforms with higher density (600,000 SNPs compared to the current 50,000 SNPs).

Dr Van Tassell’s  “quotes” come from the online “polled dairy cattle” discussion group within his responses to questions.
What conclusions did Dr Van Tassel offer for the market use of Genomic data ?

“I was asked after a presentation that I gave to the New York all-breed society meeting in January 2009 whether I would use bulls based [entirely] on Genomic testing.     My response (which I stand by here in August) was that if I had spent my life building a herd of cows, that I would look at any technology that had the ability to undermine those efforts with great skepticism.    I think that this technology has great promise, but as of yet, is still largely untested in the true application of predicting response to selection using genomic predictions.”

He earlier answered my query, “that I am absolutely correct that we have not seen any real evidence of the PREDICTIVE power of the Genomic PTA.   Everything up to now has been predicting genetic merit for animals that were selected using quantitative genetic tools that we then re-predicted using SNP data, OR animals that have had genomic predictions early in life that have not been validated by real progeny data.”

The current excitement over Genomics by those scientists involved at AIPL is… “we have recreated [the selection scenario] by using genomic data from historic bulls to predict future genetic merit using data of 5+ years ago, and then looked to see if the predictions were more accurate than the old [pedigree-based] evaluations, and they indeed were!”

What underlying problems plague accuracy in genetic evaluation?

Dr Van Tassel reminded us that based upon DNA samples of heifers given ID, anywhere from 10% to possibly 40% of the progeny lists of sampled sires are misidentified (by sire or dam or both).    Because this has an ongoing potential to cause fluctuation in progeny evaluations, it costs us all in accuracy of the data we attempt to use for genetic selection improvement.

How should we then adapt G tested sires in our breeding programs?

They remain, for practical purposes, “young sires for sampling”.    Using a group rather than a single bull choice;  using added data from their individuality, pedigree and aAa;  avoiding paying too great a premium for an elite G-based ranking level;  all this makes common sense.

Based on this advice from a scientist actively working in Genomics (rather than a magazine writer who needed an upbeat topic for the latest issue) – we offer G tested sires as “super samplers”, and we do not confuse the ranking of progeny-tested sires on our price lists alongside G tested young sires.

A final quote from Dr Curt:  “if this technology does work, then there is a huge opportunity for someone to use it aggressively as an ‘early adopter’.”     So do your homework and buy sires accordingly.

The breeder’s philosophy in the use of sire selection tools


We tend to believe that, every time a new technology comes along, it makes earlier practices obsolete.   In dairy cattle selection and mating, this is not supported by experience.    The more tools available, old or new, the better job we do at sorting the long-term useful (dependable) from the short-term novelty.  
Genomics is no different today than indexes were in 1970; they are a way to screen animals, while you seek the traits and qualities specific to needs you have in your herd.    The identification of those traits specific to problems in your herd, which can be solved by careful analysis and mating by heritability, remains more important to your profitability than any external ranking of genetic value.      

 
Are you ready to pass up progeny testing to rely completely on Genomic testing?

At least one major AI system is so convinced that Genomics is a “done deal” that it has replaced 99% Rel progeny-tested sires with G-tested sires on its active sire price list and is scaling back its young sire sampling program in favor of its own ET donor herd.

Inside you will find data of earlier G-test experiences around the world, an example of a current issue in debate on how G testing should be done, and a recommendation from a key researcher that cautions caution alongside optimism.

Our goal is that your herd continue to improve as fast as good genes and sound matings and competent heifer rearing allows.     We hope you find this information challenging and able to stimulate better-informed questions on sire selection.

Saturday, November 15, 2014

Selecting for “Type” does lead to “Production”—just not overnight


I have always found pedigree studies fascinating, and it is one of the reasons why I remain “contrarian” on breeding selection by indexes alone (or even primarily).

The three main Holstein “production” sire lines descend from three sires – Wis Burke Ideal born in 1947, Osborndale Ivanhoe born 1952, and Pawnee Farm Arlinda Chief born 1962.    All three sires developed in the context of earlier sire selection systems that combined a belief in “linebreeding” with the breeding worth estimations used prior to “Animal Model” and composite index rankings.

“Ivanhoe” was proven by three breeder partners in CT, MA and RI, started out controversial for type when first entering AI in 1958, but within a decade his “tall, style” physique became the Holstein type standard.    In 1966, at 99% Rpt, he was only +270m, with plus bf%, +21bf—hardly a “ranking” sire in an era that had AI proven sires as high as +1600m.    But those other “milk” sirelines are extinct, while “Ivanhoe”—the +1.66 “type” and +2.40 “stature” sire is the direct grandsire of Carlin M Ivanhoe Bell (born 1974), whose sons once dominated the ranking lists for milk and protein production, and whose descendants (in spite of *BL and *CV) are still pretty obvious around the world.

In three generations, “Ivanhoe” as a  type sire, produced  “Apollo” (the sire of +2000m “Wayne”) and “Ivanhoe Star” (+1000m in his own right), among many others…    His son “Mowry Prince” sired the first 50,000-pound cow, Mowry Prince Corrine… “Ivanhoe Star” then produced “Bell” whose +1700m with +60bf and +40pr put him at the top, at the beginning of the “indexing” era.    [Note: both bulls Mowry Ivanhoe Prince and Penstate Ivanhoe Star are the cross of Osborndale Ivanhoe onto Lauxmont Admiral Lucifer daughters.]

“Arlinda Chief” was proven by Wally Lindskoog in CA, entering AI on a two-herd proof in 1967, and thus was highly controversial even before we had the multi-herd AI sampling model thoroughly in place.
But he was +1622m and +79bf in those two herds, so many were willing to give him a chance.   By the time he was 99% Rel he had reached the +1805m mark—also climbing from +0.35 to +1.17 type.

What makes “Chief” so unique today was more common in his era—he was closely linebred, with three close crosses to ABC Reflection Sovereign, a Canadian “show type” bull—six crosses to Montvic Rag Apple Sovereign (sire of “ABC”)— and twelve crosses  to Johanna Rag Apple Pabst (maternal grand-sire to “Sovereign”).     Most of the “ABC” sons were more “type” than “milk”, but they were used in a similar way to “Ivanhoe”—as sources of more “modern” udder and stature traits.

In three generations, looking on the sire side of “Chief”, you go from a –1035m double grandson of old “ABC” (Rosafe Pearl Hannibal) to a +700m bull, whose dam also carried more “Sovereign” (Pawnee Farm Reflection Admiral), to “Arlinda Chief”, both of whose grandams were sired by the same old “Sovereign” son (Tabur Sovereign Man O War).     –1035m to +700m to +1800m in three generations.

“Chief” started breeding milky sons and never quit—first Glendell, then Conductor, then Valiant, then Milu Betty, all the way to Walkway Chief Mark (a fifteen year span from first to last AI success).    He sired a world record cow—Beecher Arlinda Ellen (55,661 pounds in 365 days before rBST).        

Milu Betty Ivanhoe Chief – never as widely used, thus easier to forget today, combined “Chief” on top and “Ivanhoe” on the bottom, and his grandam was Dunloggin-bred just like those old “Lucifer” cows behind “Ivanhoe Star” and “Prince”.    But he sired Cal Clark Board Chairman, who in turn is sire of ToMar Blackstar – so you can see what we sometimes consider “outcross” is more mating effect than ancestral exclusion.    (“Blackstar” is a multiple of Chief, Elevation, Ivanhoe and Burke crosses).     

“Selecting for Type does lead to Production—just not overnight”   (page two)


“Wis Burke Ideal” --  why leave the eldest to last?     Because his influence is more subtle, yet more extensive, than the credit he ever gets.     Holstein’s “Red Book” tells you the three biggest sire lines are basically  “Bell” (Ivanhoe) – “Chief” (Rag Apple) – and “Elevation”, and that is true—but old Elevation combines all three lines—the “Burke’s” through his inbred sire Tidy Burke Elevation, “Ivanhoe” from his dam—Round Oak Ivanhoe Eve—and the Rag Apples through the linebred dam of old “Eve”.    Just as “Ivanhoe” was the result of “linebred sire x inbred dam” from two unrelated lines, “Elevation” was a result of “inbred sire x linebred dam” from two unrelated lines.    


“Elevation” basically preserved the smoother Burke bloodline phenotype into modern breeding, as a mating balance for the “tall, dairy” Ivanhoe type and the “strong, style” Chief type.    Old Wis Burke Ideal was wide, deep, open ribbed, strong front-ended.    He was only +477m in his era, but he sired useful bulls like Tidy Burke Forty Niner (+835m at 99% Rpt, who sired +2000m Arlinda Jet Stream, who sired +3000m Browncroft Jetson).   He was # one bull for Feet under Holstein USA’s “descriptive type” system and lived over 17 years of age.   The Thonyma and Paclamar herds were deep into WBI breeding—he is part of the sire side of Paclamar Astronaut (considered by many the major modern source of “Protein” in Holstein breeding).

“Elevation” was noted early on for “milk” sons like Rockalli Son of Bova (in four generations you have today’s Net Merit leader “O-Man”) -- and later for “type” sons like Hanoverhill Starbuck (two major crosses to WBI—“Elevation” as sire, “Astronaut” as the dam’s sire).   Here again, the “three generation” rule seems to apply:              Hanoverhill Starbuck   (a “type” sire)

      [son]  Ronnybrook Prelude  (a “fat” sire)                 [son]  Madawaska Aerostar  (a “milk” sire)
    
      [son]  Carol Prelude Mtoto   (high Euro index)        [daughter]  Condon Aero Sharon  (EX-91)

                                              In this linebred progression, we produce
                                              Picston Shottle   (premier Genomic “sire of sons”)    
The “Elevation” influence should be heavy in “Shottle”, as he carries four crosses (Mtoto is doubled Elevation, Aerostar is doubled Elevation) – but he breeds more like old “Starbuck”.      So you could probably cross him on “O-Man” (strong, wide) as well as “Storm” (dairy, wide) and get good results, even if that means even more crosses to old “Elevation”  [and “Ivanhoe”]  [and “Chief” as well].

My final example—the dam of “Shottle” is a three generation progression as well, from type to milk, without sacrifice of type:   (dam three) sired by Browndale Commissioner, a “pure type” bull;  (dam two) sired by Hanoverhill Inspiration, a “type” pedigreed milk bull;  (dam one) sired by Madawaska Aerostar, a true “milk” bull, who actually was inconsistent in type—but the result was an EX-91 cow who produced a top record of 45,000m with 2340bf in fourth lactation, and a world-class “index” bull.

So  what  is  my  point ??

Quit worrying that every bull you use has to be +1000m.    It is more important that every bull you use is capable of adding more desired traits, than undesirable weaknesses.    It is also more important that your mating combinations produce phenotypic balance, than that they have a high pedigree index.    If we are learning anything from Genomic testing, it is that “pedigree” was an imperfect predictor of performance, and that “inbreeding” is not pedigree-driven, it is an increase in homozygous gene pairings, which may occur more often from mating similar phenotypes than it comes from passive pedigree relationship