Fat, resilience, reproduction and profit in Merinos

Why fat deserves a place in the breeding objective

Thanks Josh, and the Mumblebone team, for having us along today. My topic is the role of fat in resilience, reproduction and profit. There are a couple of key messages. I think fat is your friend, and clearly, if we look up here, it's been a bit of a focus for Mumblebone. But just how much of a friend it is depends on where you farm, what your attitude is to risk and to dealing with seasonal variability, and also what your attitude is to stocking rate.

So I'll rave on about it a bit, but you need to temper it. It's more important if you live in tougher environments, and it's more important if you're focused on stocking rate. The other thing, before I get going, is to finish where Ferg got to about balance. I don't want you to leave the room thinking I'm just one-eyed and only after fat. It's just another trait that we need to consider in conjunction with everything else.

What we've been trying to do is enable breeders and commercial producers to better understand the value of the trait of fatness. It's currently not considered in standard selection indexes. We all know, to varying degrees, that there is a trade-off between fat and fleece weight, and we have different ideas on just how big that is. We need a better understanding of these relationships so we can appropriately value fleece weight, fatness and so on in our breeding objectives.

This is some data from a colleague of ours at Agrarian Management, a farm consultant in the south-west of WA around the Kojonup and Katanning area. It's a small sample of his clients, collected over maybe 10 or 15 years, including not-so-good times like now. Right now in WA margins are well down, whereas over the last couple of years we've been at 50 and 60 dollars. So that's the first thing: this data represents a spread across time.

Why am I showing it to you? We've done a pretty good job moving from left to right. That's what you've been selecting sheep for: to be more productive, with higher fleece weights, finer wool or more lambs, whatever you choose. Ultimately you've been trying to breed sheep that generate more income. And take Ferg's message that it's not just about income, it's also about reducing costs, and therefore ultimately increasing gross margin per DSE, obviously helped by annual changes in market conditions.

The big factor going up and down the graph, though, is stocking rate. Dollars per DSE sets the upper ceiling; you can see the ceiling to the data set up there. But how close you actually get to that ceiling is stocking rate, that's all it is. Stocking rate multiplied by dollars per DSE equals dollars per hectare. And we've put less time into working out what traits we might select for that can actually enable us to run more sheep per hectare, because that's really what matters if that's what you're chasing.

It's a different story if you're a big cropper with thousands of acres of stubbles to run your ewes on; it's not going to be so important for you. But in permanent pasture country, in high rainfall, a lot of farmers will be chasing stocking rate. We think fat ticks a fair few of the boxes. So what is fat? It's basically energy reserves. Eighty or 90% of the energy in your body is fat.

The energy cost of fat and protein

What does it do? We think it provides buffering capacity, the ability to cope with poor seasons, and that's what we've been trying to get a handle on: the opportunities to increase stocking rates and/or reduce supplementary feeding. I'll talk a little bit about that as well. And finally, we've known for a little while that it's certainly beneficial in terms of reproductive performance and welfare. We'll work our way through some of these.

Before I get there, when you talk about fat it normally has bad connotations, if that's the right English. So I want to get rid of a few myths and check a few facts. There's no question fat is expensive to lay down in the first place. There are about 40 megajoules per kilo in fat, and when you take the efficiency with which it's deposited into consideration, it costs you, let's say, 60 megajoules of energy intake to put down a kilo of fat. By comparison, for protein it might be 12 or 15. So the requirement for energy is much, much higher for putting fat down.

We'll come back to that one. But once you've got fat, fat's cheap to keep. There's zero running cost. There's no metabolic cost of it just sitting there; it's very inert. By contrast, protein has a very high turnover rate and a high metabolic cost. Put it this way: about 10% of the protein in your body is burnt every week. You can do the sums, but it might be something like three-plus megajoules per day of a maintenance requirement of seven or eight, depending on how big your sheep are, just purely replacing and turning over protein. So you've got one that just sits there and costs you nothing, and one that costs you quite a bit.

The other part of the equation is when we go through tougher times, like summer, and we need to draw on our reserves. When we pull back our fat, it cost us 60 to put down and we get 40 back. Our protein might have cost us 15 to put down, but we only get 5 back. So recycling fat, putting on fat and losing fat, is about twice as efficient as doing that with protein. A fatter animal has lower running costs and is more efficient at going up and down with the seasons.

What we already know about the benefits of fat

Here are some of the traits where we've known about the benefits of fat for a long time in some cases. Some work from a good past colleague of ours, Sarah Blumer, almost 10 years ago, showed that fatter animals with a higher condition score have a lower maintenance cost. Over the summer, an animal that was half a condition score fatter going into summer could get by eating about 0.7 of a megajoule per day less, 10% less per day, and still achieve the same liveweight outcome. They needed less because they were saving on their maintenance cost.

In the work I'll talk about later in this presentation, with a different scenario of adult wethers, we're finding up to 1.6. So the fatter animals needed 1.6 megajoules per day less to achieve the same liveweight gain. That supports what I was saying before. Likewise, fatter animals will lose less weight over the summer. If we need to pull X megajoules out of storage, because fat is so dense in energy you're going to lose less weight doing that. In Sarah's work, from the same intake, there was about a 30 gram per day difference where animals were about half a condition score fatter.

Then there's increased reproductive rate. Ferg did a lot of this work, again more than 10 years ago. We know reproduction is mainly driven by environment, nutrition and so on, but there is a genetic component. On average in that work, I think one extra millimetre of genetic fat was about a 5% increase in reproductive rate, in scanning performance. The point was that it was zero in some years and 25 in others: zero in good years, when you don't need the extra genetic fat, and 25 in the tough years. That comes back to resilience, and consistency of performance year in, year out.

There are other benefits of having some extra fat in the system. Again this is from Ferg's PhD. What we're looking at here is the ewe breeding value for fat along the bottom and lamb birth weight up the side, which you can think of as an important part of lamb survival. Then there's a scenario where he had ewes on poor nutrition, lambing in condition score 2.5, and another group lambing in condition score 3. As you can see, on high nutrition the genetic fatness of the ewe wasn't so critical. They produced twins of around 4 kilos regardless of the genetics.

If we're running them a bit harder, maybe like this year, certainly in the west, where despite all our work things have just got tough, there's no question people have eased off on condition score targets and run them harder. In that scenario, having that extra genetic fat in the system could buffer them and make sure the foetus was nourished and resulted in a half-decent lamb.

Some of our own work, again 20 years old, from Lifetime Wool, and many of you would have done Lifetime Ewe Management, shows the well-known relationship between lamb birth weight and survival. The different curves represent the survival of lambs where the sires had different breeding values for fat. As you can see, going from a yearling fat of minus one at the bottom to a yearling fat of plus one, the fat of the sire was resulting in fatter lambs that were presumably more able to survive. We've duplicated that work with weaner survival as well. Particularly with lighter weaners, if they're from genetically fat sires they'll have better survival. Some of these effects are small, but they're cumulative, aren't they? So far at least, there are lots of small positive benefits of fatness.

Fat, ewe lambs and the gaps in genetic evaluation

Joining ewe lambs was mentioned earlier, and there's been a fair bit of work in that space. What we're looking at here is liveweight at joining and reproduction. This is Merino data: daughters joined at seven months, from sires with different fatness. As we go from lean to fatter sires, and it's old work, 12 years old, and we've gone past plus one by quite a way now, there are benefits in terms of early maturity and reproductive performance in ewe lambs joined at seven months if they're coming from a genetically fatter background.

I also look at this a slightly different way, and that is that having that extra fat in the system gives you the option of joining at a lighter weight. If you look across this way, you might be able to join ewe lambs closer to 40 kilos if they're from a genetically fatter background, compared to 45 if they're from a genetically leaner background. So you can cut the pie different ways.

Enough of that. That's the sort of stuff we've known for a little while. There are clear benefits of extra genetic fat to resilience, mainly the liveweight change in summer, the capacity to be more consistent with reproductive performance across years, and clear links with reproduction. So we need to consider the value of fatness in genetic evaluation systems, and as I said in my introduction, we currently don't. There is no value put on fat in the standard indexes, and/or no consideration given to potential antagonistic relationships between, say, fleece weight and fatness. Fleece weight carries a very high economic value, but maybe it's not quite so high if it's having a negative impact on fatness, and on some of the things I'll talk about later.

Most genetic evaluation systems value animals on a per head basis, but you make money on a per hectare basis. Those that have tried to predict the effects on stocking rate and production per hectare generally just assume a bigger animal needs to be run at a lower stocking rate. They use metabolic liveweight, liveweight to the power of 0.75, which is closely related to the maintenance requirement of an animal. So they say a bigger animal has a higher maintenance requirement, therefore you need to run fewer of them, and generally bigger animals don't stack up too well in many genetic evaluations.

We know that's a pretty limited system. For one, the maintenance requirement of the ewe is probably less than 50% of the total feed consumed on your farm over the course of a year. We certainly know now that there is variation in maintenance requirement. I talked before about the effects of fatness and lean on maintenance requirement. We also know there are differences in feed efficiency, the efficiency with which animals retain energy, and differences in the way animals partition energy. So just using liveweight or metabolic liveweight is a pretty blunt instrument for predicting the potential impacts of different genetics on stocking rate and profit.

Animals partition energy differently

Back there again is some work by my colleague Sarah a little while ago that really got us started down the line we've been working on for the last five years. Sarah compared a wool-type genotype and a carcase-type genotype. She put them in an animal house, measured them by CT scan, measured their fatness, and in this example we've converted that to energy content. The carcase-type genotype, the red bar, was essentially fatter to start with. I think it was 33% body fat, and the wool genotype was 31.

Then she fed them flat out for 42 days and measured their fatness again. They ate the same and they grew the same. So the old way of thinking would say there were no differences in efficiency. But what happened was the carcase-type genotype partitioned energy differently. A whole lot of its growth was dedicated to fat, and it went from 33% body fat to 39%. The wool genotype went from 31 to 32 or something like that.

In other words, liveweight and energy retention are different things. They did the same thing in terms of liveweight gain, but there were massive differences in how they partitioned the energy into lean or protein, fat, and wool in that case. I'm conscious of time, but this then led us to say: hang on, our gut feel that animals are not all the same was right, and in fact there are massive differences in how animals use energy. We needed to get a feel for just how important that could be.

So our good colleague John Young did some modelling based on the best information we had at the time, some years ago now, looking at the effects of genetic differences in whole body energy across the bottom. This used work by Norm Adams, who supervised Mark's PhD. He had taken some animals selected for high and low fleece weight, and the high fleece weight animals had about 2 kilos, or 15%, less fat. So they produced lots of wool, but they had about 2 kilos or 15% less fat. That's what set up the range on the bottom.

What does a 15% difference in whole body energy, whole body fatness, mean? John modelled that in a whole-farm context, taking into consideration stocking rate, pasture utilisation, supplementary feed and everything else. Don't get too carried away by the numbers, but genetically fatter animals were suggesting an optimum system, which by definition optimises profit and runs the optimum stocking rate, where stocking rate increased by 30-odd per cent, supplementary feeding went down by a heap and profit went up by a heap.

This is probably over the top, because correlations between traits weren't considered. It was simply that if you're a fatter animal, you can run a whole lot more of them, they don't need as much supplement, and it's going to be more profitable. The point was the magnitude of the numbers. We're talking about something significant here, especially for producers who want to pursue stocking rate, and we need to know more about it. Thanks to AWI, who chipped in the money back in about 2018, we've been chipping away at that over the last five or six years.

You have to stock them up to get the value

Just a point here. Genetic fat is potentially valuable, but sorry about jumping around: I've taken that profit line, the top line there, and put it on this graph. What it's saying is that if I have genetically fat sheep, the benefit, which is the big number, is only derived if I make them work. I capitalise on having those fatter sheep by running more of them. That's how you make the money.

The bottom line more closely resembles the old wether trials we had, and probably even the MLP project recently, where we just run them all together. The genetically fat ones get fatter, but you don't get the value from running more of them. So what am I trying to say? If you have genetically fat sheep, to really reap the rewards you need to lift the stocking rate. That's how you extract the most value from them. If you don't want to do that there are still benefits. They'll need less supplementary feeding, they'll give you a few more lambs, and they'll give you consistency between years and so forth. But the big whack comes from putting the extra sheep on.

That led to the project I'll talk about now, and I'll skip through it. I can't remember where the local MLP site has been around here, but there has been one. We had one site at Pingelly in Western Australia, in 2016–17, where 29 different sires were compared. Many of you would be familiar with the Merino Lifetime Productivity project. There were five sites around the country, with let's say 30 to 50 sires at each site, and the daughters were measured for everything for their whole lives. So we've got really good numbers on their wool production, their worm resistance, their reproductive performance and so on, but they were all managed together.

What we did was take the wethers, the brothers, because we wanted to better understand a whole lot of key questions. How big are these genetic differences in whole body fatness? What's making animals fat, what's the biology behind it? Do they eat more? Do they have a lower maintenance requirement? Are they more efficient at extracting energy from feed? Are they more efficient in how they partition that retained energy? Then we wanted to look at the genetic correlations, the relationships between all these traits and whole body fat. The industry currently uses yearling fat, or ultrasound fat, as a predictor of whole body fat. And finally we wanted to run the economics and work out what it all means, to better understand the value of fat in breeding objectives, which brings in all the other traits as well.

Big differences between sire groups

I'll hit you with a few graphs now, but the next three are simply there to show the diversity. The numbers 1 to 29 mean absolutely nothing; they're just random. These animals had run together for three years as a single mob, and then they came into this experiment as three-year-old adult wethers. At that point we've essentially got teams of wethers that range in condition score from 2.5 to 3.4. Put your hat on going into summer: think about having a mob of sheep in 2.5 versus 3.4, and what that might mean for how you might be placed for the next six months until it rains again.

That's a massive range, which is great; that's what we wanted. We also put them through a DEXA scan, which enabled us to get a really good handle on their body composition rather than just their condition score. Again, 1 to 29 means nothing, but we're seeing a range in whole body fatness of anything from 16 to 24%.

Why is that significant? The modelling was done on a 15% difference. Here we've got a 50% difference. From 16 to 24 is a 50% range in whole body fat, and yet the modelling with a 15% difference was saying that fat might be important. So it's a wonderful resource to work with to try to understand what's going on.

We also see big differences in feed intake. This intake is adjusted for liveweight. Normally we'd think it should all be the same once you've taken liveweight out of the equation, but it's actually a 30 or 40% range in potential intake. All I'm saying is pretty obvious: animals are really very different in a whole lot of traits, and how can we use that to better value traits, or appropriately value them?

This one is a little bit hard to get your head around. It's one of our measures of efficiency. I'm looking at energy intake, megajoules per day, across the bottom, and energy retained up the side, which is the change in fat and muscle converted to an energy value. Don't worry about the numbers. All I'm trying to show once again is that there's a lot of range there. Where each line stops is as much as they ate, so there are lots of differences in what they ate, as I showed you in the previous one. Some animals eat more than others. But the steepness of the line also reflects their efficiency. Some animals use the energy much more efficiently than others.

Fatter animals eat less and retain more

There are a couple of reasons for that, and these are some interesting ones. Hopefully I'm not blowing your mind with too many graphs. Here I'm taking some of the differences we've just seen and asking how they relate to how fat the animal was, its whole body energy or whole body fat. What we find is that fatter animals eat less. Their intake capacity was reduced. We've actually taken this out and measured it on ewes as best we can, using methane production as a proxy for intake, and we found exactly the same thing. So I'm pretty confident on this one that animals with a high proportion of whole body fat eat less.

Animals that are fatter have a lower maintenance requirement, which is what we would have hypothesised. They have a greater proportion of their bodies made up of fat, which just sits there. The last one is a different term, but it's just another measure of efficiency, and it relates to the previous one. If animals are fatter, they're more efficient. They eat less, but despite eating less they can retain more energy, because they have a low running cost, and then they partition more into getting even fatter.

Nearly there. So what does it all mean? The first thing is: how do we increase whole body fat? The good news is that it's very strongly correlated with, in this case, sire breeding values at yearling age. The correlation between fat depth, essentially the yearling fat that's on the board over there, and whole body fat is 0.7. If you measure it a different way, and some of you might be familiar with genetic correlations, it's 0.85 or something. So a very strong correlation. Condition score is again an extremely strong correlation, both as I've expressed it here against sire breeding value and as a genetic correlation.

What does that mean? It means we don't need another breeding value. The breeders in the room would probably be pleased about that, because there's been an absolute flood of breeding values and it's getting more and more complex. If we want to increase whole body fat, we don't need to start running around with DEXA scanners and stuff like that.

Fat and the other traits

I've stuck up some other interesting ones. With coefficient of variation of fibre diameter, I don't know the mechanism and I'm not sure if Ferg would, but if we select for low CV of diameter we get animals that have more body fat. Worm egg count is another really interesting one: select for low worm egg count and you get fatter animals. So I think they're all going in the right direction.

Clean fleece weight is the interesting one. There's no question there is a negative relationship between fleece weight and fatness, and I don't know whether that's a good thing or a bad thing. That's what we're trying to work out: the appropriate values. A couple of things. One is that the correlation is 0.33, which is a bit like Ferg's graphs, maybe slightly stronger. A correlation of 0.33 means that only 10% of the variation in fleece weight is explained by fatness. In other words it's like a shotgun: there are dots everywhere, and you can break the correlation.

So I'm not worried one bit about 0.33 as being something to be alarmed about, because there's plenty of scope to select fatter sheep that still produce a lot of wool. If you're interested in the magnitude of the trade-off we got between fleece weight and fatness, I'm still working on it, and there's lots of scatter so I'm not even sure what the average line means, but it's probably in the order of 150 grams of wool per kilo of fat. I'll come back to that later on. Other work found 400 to 500 grams of wool per kilo of fat. We certainly couldn't find that in our work, where we've got much more diversity of sires.

What does it mean for profit? I have to say this is a work in progress. I was talking to our economist at 1 am this morning, and I stopped off at Albury and had another session with him. I was waiting for the updated graph to appear in my email and it didn't. To be honest this isn't super valuable anyway, because it's 29 sires that are hardly relevant to any of you. But the one point is that across those 29 sires, based on a preliminary analysis taking into consideration all the measured traits on all the daughters up to six years of age, plus our best estimates of this new biology, if you want to call it that, about energy traits and feed intake on the wethers, fatter animals were more profitable on a per hectare basis.

That will get sorted, and as I said it's of limited value anyway. What you really want, and what we've started on, is better estimates of the relative economic value of fatness and the relative economic value of fleece weight and other traits, which start taking into consideration a better understanding of differences in intake and efficiency and so on.

Wrapping up

Just to wrap up, Josh. Whole body fat is highly heritable. It depends a little bit at the moment on how we express it, but it's 0.6 to 0.7, so we can absolutely change it, like we can change yearling fat and carcase fat. It's strongly correlated with condition score and ultrasound yearling fat, therefore we don't need additional breeding values, which is significant. And there's no question that genetically fat sheep were more efficient at storing energy during the good times, when they were fed ad lib. That's what we want.

Despite not eating as much, which is intriguing, the other parts of the equation more than counteracted the fact that they ate less. They can still retain more energy as fat during, let's call it spring, which was ad lib feeding in a pen, and then use it at other times of the year. There are favourable correlations with other traits as well, whether that's staple strength, wrinkle, conception or litter size. And there's a moderate correlation, I think 0.4, between fatness and survival to four years of age. Again, it's all pointing in the right direction.

It's adversely correlated with fleece weight. I've touched on that; it's not super strong, and we can look at it different ways. Our preliminary estimates, taking into consideration impacts on stocking rate and everything else, are that an extra kilo of fat might be worth $4 to $8 per kilo. Why is that relevant? As I said before, we think you might be trading off on average 150 grams of wool to get that extra kilo of fat, but that extra fat has value. We will fine-tune those estimates. In fact we're back over this part of the country with the real economist, I think in October, so he's got to have his numbers right by then.

Just to finish off: there's a little bit of work still to be done, but hopefully you've got a better understanding of fat as a trait that you might have a closer look at come ram selling time. Thank you.