Sheep Production Metrics

We use both locally calculated metrics on sheep productivity and growth, as well as metrics from the National Sheep Improvement Program (NSIP). These numbers are a much more effective way of choosing the most productive breeding stock than by “eye appraisal.” Below is an explanation of how to use the numbers and details on each metric.

What to Do With All These Numbers

NSIP metrics can be really confusing at first, seeming like an overwhelming sea of conflicting data that’s hard to sort through. Here’s how to get started using the numbers to your advantage. Think of things in your flock that you want to fix or improve: it might be lambs too big or small at birth, it might be their growth after weaning, it might be survivability. Focus first on the metrics that contribute to that improvement you want to make. Focus next on keeping traits you already have and value: if you have good weaning weights, don’t backslide by introducing a whole bunch of animals with low values here that pull your average down.

Some traits might be a “don’t care” for you. For example, if you do intensive lambing and are present to help with every birth and carefully manage lambs that have trouble post-birth, you may not need to worry much about birth weight or Number of Lambs Weaned (NLW) at all. Your management system will compensate for any problems in lower metrics here (and your subsequent NLW metrics won’t mean much, because you are helping the ewes a lot). But if you pasture lamb and are away from the farm long hours each day during lambing season; you may not want to compromise in the slightest on these numbers, as they would have huge impact on your lamb (and ewe) survival rates.

But don’t get too wrapped around the axle with the numbers. The irony of these numbers is there are some “inverse correlations”- when you choose more of one thing, you get less of something else you also want. You rarely get everything you want in a single animal. And, as we select for a certain trait, over time it becomes more and more extreme, and may cross the threshold of being too far in one direction, creating new problems. So it’s always a balancing act. Often, staying near zero in the NSIP metrics can be very safe: keeping you in the middle of the road with healthy, twinning ewes that produce moderate lambs and give you little trouble.

Genetics are complicated, and we must avoid the assumption that blending two opposites gives us something in the middle- it usually doesn’t. If you breed a huge animal to a tiny one, you don’t generally get a mid-sized baby; instead you have a 50/50 chance of getting a huge or tiny baby. With NSIP numbers, it’s best to think of them as a bell curve, where you are constantly trimming off the bottom, concentrating on the middle, and judiciously using the upper extremes for improvement and helping to shift the bell curve.

The numbers are not absolute: the numbers I choose as best for my flock might be completely different than what you’d choose for yours. It’s ideal to try to buy breeding stock from management systems similar to your own. If you grass feed, it’s a gamble buying sheep from a grain-fed system; as they may not be optimized for grass. Lambs reared on high-protein creep will always grow faster than lambs on grass alone. Even within grassfed systems, some lambs might be on bright green, boot-high grass in fertile pastures, others might be on meager, desert ranges where they have to walk a lot to take in only marginal forage. Often, people who feed alfalfa or grass fines still use the term grass-fed; but these animals are taking in high protein, dry-concentrate feed, so are more comparable to a grain-fed system.

And, though comparing NSIP numbers across flocks is useful, it does not always give you the full picture. NSIP is designed to be “system agnostic”- if enough related animals from different farms are in the database, then the algorithms can rule out the variables of rearing environment. It can then highlight animals which can do well anywhere. But in many cases, there isn’t yet enough cross-flock data in the system for this to work well. So your best information will be from intra-flock data: looking at a single flock’s bell curve spread, and choosing the best animals from inside that system for the traits which are important to you. A 1.2 WWT score might be the top animal in a pasture of 100 sheep, where it might be a low number in the spread of a Midwest feedlot operation. So always take into account what the number implies for that animal’s performance inside his own system, versus the nationwide pool of very different systems.

We must also not forget about management systems and how they influence animal performance. If you are having trouble with tiny lambs at birth, it may be genetics. Or, it could be that if you increase nutrition in late pregnancy, you’ll correct the problem that way. Most of the traits we value are only lowly heritable, and largely influenced by management choices. So, it’s important to concentrate on both.

Basic Metrics

Birth weight (BWT): The “raw” value of the lamb’s weight at birth. Birth weight is critical for hardiness and survival in pasture lambing situations. Too-small lambs chill easily in cold, wet weather; and too-large lambs can cause birthing difficulties. We shoot for lambs in the 7-11 pound range. People who barn lamb may prefer smaller lambs to reduce the incidence of difficult births, since chilling is not such a factor in that environment.

60 Day Adjusted Weight (60Adj): we weigh all the lambs on the same day near their sixty day birthdays. Since most are not weighed exactly on the 60-day anniversary of their birth, the weight taken is adjusted, assuming a linear growth curve from birth, so that they are all fairly compared whether they were weighed a little early or a little late.

60 Day Corrected Weight (60corr): the lamb’s 60-day weight adjusted again to account for it’s dam’s age and the lamb’s rearing type (single, twin or triplet). The Katahdin breed has customized adjustment factors which account for the fact that yearling and two-year-old ewes are not as productive in milk as mature ewes; and that single births grow faster than twins and triplets. This extra math gives a yearling or two year old ewe “credit” for rearing smaller twins as compared to a mature ewe who might raise one very big yearling. After applying this adjustment, the “spread” between the weights becomes larger, so you can really see which lambs are at the top and bottom tiers. Whereas with the “raw” 60 day weights, they are all close together and it’s hard to see much differentiation.

120 Day Adjusted Weight (120Adj): Same as 60Adj, the weights are taken all on one day around the time the lambs are four months old. The numbers are adjusted on a linear scale so that they reflect weights as if the lambs were each weighed exactly on their 120-day birthdays.

NSIP Metrics

The NSIP metrics are all expressed as a percentage of the mean: so larger positive values indicate you will get “more” of this trait, negative values will give you “less.” They are like control knobs on a stereo synthesizer. Zero values imply the animal is “average” for that trait. NSIP calculates “expected breeding values” or EBVs (also called EPDs, or “expected progeny differences”), which predict the future productivity of an animal based on its past performance and genetic correlations with related animals’ performances. This extra mathematical regression analysis based on genetics helps identify animals with superior genes, giving much more information than breeders could do “in their heads” based on memory of accumulated data over the years.

Birth weight (BWT): estimates direct genetic potential on weight at birth. Picking bigger numbers here will give you bigger lambs at birth. But if you tend to have big lambs already, or even too-big lambs, you will want to avoid big numbers here. Birth weight is one of those fine lines we walk. Bigger birth weight is statistically correlated to higher weaning and post-weaning weights; so selecting for it is proven to help us produce bigger butcher lambs sooner. But too-big lambs increase the incidence of difficult births (dystocia). So bigger is not always better with this metric.

Maternal birth weight (MBWT): estimates effects of the ewe on the birth weight of her lambs- reflects the quality of the uterine environment and genetic effects of length of gestation. Here again, bigger is not always better- it depends on whether you need to adjust up or down here within your own herd performance: it’s a control knob for you to use.

Weaning Weight (WWT): estimate of pre-weaning growth potential. Usually we select for big numbers here. But concentration on this metric brings two opposite things: some lambs which are early maturing (that’s what we want!) and some lambs which are actually late maturing, but are destined to be very large adult animals (which we don’t want!). So, avoid getting the “big eye” when you see big numbers here: they should be used with prudence to improve early maturation tendencies, while avoiding producing a herd of pony-sized adult sheep!

Maternal Weaning Weight (MWWT): estimate of the genetic merit for mothering ability which contributes to pre-weaning growth. Some component of this metric is general mothering instinct: good mothers keep their lambs close and make nursing convenient, promoting growth. But most of the metric is about milk; thus it was called “maternal milk” in the old NSIP system. This metric is more important for pasture lambing operations where lambs get no creep feed; because the lambs are relying solely on milk for growth during most of their first sixty days. A lamb can have awesome growth genetics, but if the milk isn’t there, he won’t be able to express them if his only other food is grass, and his rumen isn’t yet mature. On the contrary, systems where creep feed is used may care less about maternal milk, because the lambs can compensate for a weak milk supply by taking in more feed concentrate. As with all metrics, it’s another one to use with caution. We want lots of rich milk, for sure, but ewes with big udders can be hard to dry off and tend more towards mastitis, especially in early weaning situations. So choose based on your system: if you wean early or use creep feed, go easy on these numbers. If you pasture lamb and let ewes self-wean their lambs, you can afford to let this number go bigger in your flock.

Post-weaning Weight (PWWT): estimate of the genetic merit for growth after weaning. This metric assumes a lamb was weaned at 60 days, and his resulting growth after that is all his own genetics, unrelated to his mother’s milk. Just like with WWT, it’s tempting to select for bigger and bigger here. But what we want to choose are early maturing lambs, not late-maturing sheep which will be huge as adults.

Number of Lambs Born (NLB): estimate of the genetic potential for prolificacy- number of eggs dropped, fertilized and brought to fruition as live births. Choosing sub-zero numbers here can lead you towards less twinning; big numbers here will take you down the road of lots o’ triplets. But don’t get too wrapped around the axle with this number: birth rate is lowly heritable, so as long as you stay out of the extremes with both sire and dam in a breeding, you’ll just get regular old twins most of the time, like with all Katahdins.

Number of Lambs Weaned (NLW): estimate of the combined ewe effects on prolificacy and lamb survival to weaning. This number gives credit to ewes that not only tend to have multiple births, but also manage to keep them alive to weaning. High birth rate does no good if a ewe forgets about her second or third lamb and leaves him behind, isn’t adequately protective or nurturing, or if she doesn’t have enough milk to feed all the lambs she birthes. This metric can be falsely harsh: it doesn’t know if a ewe’s lamb was killed by a dog, crushed by a gate or taken by disease caused by mismanagement. It also doesn’t know if the farmer purposely chose to bottle-raise one or more of her lambs when she could have raised them herself. It always assumes it’s the ewe’s fault that she lost a lamb, and we know this isn’t always the case. But we do know that if she loses lambs more than once, it may reflect that she is a sloppy mother that lacks protectiveness, or perhaps that her milk supply isn’t producing robust lambs that are resistant to disease and environmental challenges.

Total Maternal Weaning Weight (TMWWT): Formerly called “milk plus growth” in the old NSIP system, it combines the EBVs for MWWT and WWT to estimate the total anticipated contribution of an animal’s daughters to lamb weaning weight. This is a derived value from other metrics, an weights more heavily WWT. It gives you a wider “spread” of numbers so you can really differentiate your top and bottom animals for a ewe’s ability to crank out good butcher lambs.

Katahdin Ewe Productivity Index (EP): combines the EBVs for various traits into an index designed to maximize “pounds of lamb weaned.” This is a derived value from the other metrics, which weights heavily MWWT, credits WWT and NLW, but slightly discredits NLB. This number also gives a wider spread of a bell curve to help differentiate extremes in ewe populations. Selecting based on this trait helps focus on ewes that reliably produce robust twins with high survivability and growth genetics. If all these NSIP numbers are confusing at first, or if you don’t have a particular trait you are worried about, EP can be an easy number on which to focus for choosing flock ewes.

USA HAIR: This is a composite index calculated by NSIP. This index is a combination of WWT, MWWT, NLB and NLW. It tries to assess what I would consider to be all-rounder success at being Katahdin: good growth, milk, prolificacy and survival. The only flaw I find in this index is that it lets poor milkers score high if they have a lot of lambs and good potential for growth. In grain-feeding systems, this is ok, but on grass, it’s a problem. So, I always look at MWWT in conjunction with this index, and penalize heavily any ewe that has low milk scores.

Maternal$: This index is very similar to the USA HAIR index. The algorithm is proprietary to LambPLAN, so we don’t know exactly how it works. I believe this one emphasizes growth and milk a little more than our Katahdin index does. I tend to rely on this one more for quick “gut check” of whether a ewe will be a good pasture/grass ewe.

Accuracy (Acc) values: when available, these are represented in parenthesis after the metric, and they reflect the accuracy or confidence level in each metric. The accuracy value will be lower if there is insufficient data in the system to confidently say that the EBV is accurate. These numbers essentially tell you when to take an EBV “with a grain of salt”- meaning that it’s still useful information, but there is a chance it could be “off”. With low accuracy values, it’s ok to fall back to your personal judgment on the relative quality of a breeding animal as compared to another. When the accuracy values are high, however, you should have a darn good reason for choosing to ignore them!