Breeding your perfect sheep with the neXtgen Index

Five steps towards the perfect sheep

Awesome, thanks Josh, and thanks to Mumblebone for hosting the day. Awesome to see such a good turnout. Josh put a question to the speakers at the start of the day: if there was one thing that everyone would take away from what you're trying to say today, what would it be? Summed up, I'm probably going to say question everything. I know that my dad and my boss are sitting in the front row going, "Of course you'd say that." But as Howard pointed out before, as farmers we're often sitting in that bottom left corner, in the conservative section, where a lot of our values are. So how can we question what we're doing, and how can we improve?

Today I'm going to be talking about breeding the perfect sheep. I don't know how many people there are in this room, but if we take 75 as a round number, I think there's going to be 75 different ideas of what the perfect sheep looks like. So if you want to have an argument, or a discussion I should probably say, about that, you can come and see me or my boss afterwards, because that's basically what we do for a job.

So what is the perfect sheep, and how do we breed it? Ferg and I had a pretty solid drive to get here last night, if anyone follows Ferg on Twitter. Surprisingly, we've actually spent a fair bit of time in the car together over the last two years, and we've had a lot of discussions and still can't quite work out what this perfect sheep looks like. But if we're going to try and get towards this perfect sheep, we've put a few steps here of how we can get to it.

The first thing will be to define that sheep. It's going to be different for everyone here in the room, but by actually putting down, usually on paper, your breeding objective, we can define what that sheep looks like. Then establish your baseline: work out where you currently sit, what genetics you currently have on your farm, and how they're actually performing in your environment. Then make a plan. We now know where we want to get to and where we are, so make a plan of how to get there.

Allocate resources: right, I know where I am and where I want to get to, so what resources do I need to get to my goal? And then reflect and refine. At the moment we're sitting here going, these are probably the best tools possible, and at neXtgen we speak a lot about utilising data and ASBVs to try and make more genetic gain. But we're not going to sit here and say that we have all the answers right now. So reflect: what's working, what's not working? And then try to refine, because there are all these new things that are going to come in the future, such as footrot and hopefully structural breeding values, and there'll probably be more things like carbon efficiency, but we won't go down that rabbit hole.

Defining your perfect sheep

So in defining your perfect sheep, some of us can look across at our neighbours, who, God forbid, might have footrot. But what works at your place might not work at their place. What works here in Wellington might actually be a very different sheep to the sheep best suited to a system at Hay in the Riverina, or to a sheep in Gippsland, where we were yesterday, under higher rainfall. In Australia there's a massive range of environments, and we see a range of different sheep types that are actually suited to those environments. So really work out what's the best sheep for your production system, what's going to work on your farm, not what's working for someone else.

I think in the past, if we look at where a lot of the studs are in Australia, they're often located in some pretty sweet country in the sheep–wheat belt. So everyone's probably traditionally gone, right, where can we find the best-looking sheep? And they've gone, this is going to have to work on my place. So really question that: what sheep do you actually want on your place, and what traits are most important?

So which traits do you want to look at for your enterprise? I think as an industry we've probably always looked very closely at those traits that make you money. But there are also these other traits that are really important: those that save you money and those that save you time, especially now in this modern environment where labour is quite hard to get. And then there are those traits that delight a customer. What's this future consumer going to look like? As Howard spoke about earlier, the consumer actually doesn't hate us, well, not all of us. So how can we keep producing a product that's going to delight that consumer?

Once again, the most profitable sheep might not actually be right for you if you're running a solo operation, or if you want to go on holidays. So try to put some of these other traits into your business. Don't just think about profitability; think about the whole picture of your farming enterprise.

So if we split up these traits and start to try and form what my breeding objective is actually going to look like, first are the traits that make you money. Work out what the key profit drivers are in your business. We've got this funny little conundrum where we're trying to look at a ram or look at a ewe and not only work out what its genetic potential is, because all breeding values are per head, but also what's actually driving profit on your farm. Is that kilograms of lamb per hectare, kilograms of wool per hectare, and surplus stock sales? So we've got this per-head animal, and all those traits that are about this animal per head, and we're trying to optimise the system.

It's been really cool travelling around for the last few years. There are systems further south where they have very seasonal rainfall patterns, and by putting more growth into some sheep down there, they've been able to finish lambs earlier, by seven months of age, shear the wool off them, and run more ewes per hectare. As a whole system, that system is actually cutting more wool without necessarily selecting on wool cut per head.

So really work out what the profit drivers are for your system. How can I make a sheep that wants to run on my property, optimising stocking rate, so it really works for your system? There's a lot of people in this room, some from South Australia, Victoria and probably Queensland, so everyone here is probably going to have a different sheep.

So the main traits that we often look at and think about for making money are probably the main ones that are incorporated in industry indexes. Your growth traits, so your post-weaning and your yearling weights; fertility traits, which are now your weaning rate, split into three traits; and then fleece weight. I think as an industry we've probably been really good at looking at those traits.

Traits that save money and time, and traits that delight a customer

The next ones are the things that save you money and time. I've pinched this little barrel diagram here. It's used a lot in cropping with fertilisers, trying to work out what you can put on your system. As an industry we've probably been very good at trying to push more production, trying to put more water in that barrel, trying to improve profitability. But as Ferg spoke about before, if you've got something on your farm, whether it's worms or footrot, that's actually limiting your production, all that production is then leaking out.

If footrot is a problem in your flock, it probably doesn't matter how much growth, how much fertility or how much wool cut those animals have got on them, because they actually can't walk around. We see this with a lot of these welfare traits. If worms are the main burden on your farm and you can't drench them, how much production are you losing, both subclinically and on your farm?

So as an industry, have we really selected on these traits, and what traits do I really need to look at to save us money and time? Not just footrot, not just worms. There are the flystrike traits: breech wrinkle, breech cover, dag, and hopefully urine stain in the future. And then obviously the body strike traits, which have been pretty solid over the last year, so looking at fleece rot and wool colour.

The last one is the things that delight a customer. As I was saying earlier, as farmers, our product is all going to an end consumer, so we might as well try and produce something that they really enjoy. These things often influence the price we receive on farm. If we think about the things consumers like at the moment, it's a fine micron. You're obviously getting a premium for that, probably not so much in the last couple of months.

And then what are some future consumer trends? Looking at things like IMF, keeping an eye on that. At the moment there are not massive premiums, but Gundagai Meat Processors are starting to do a bit in that space, trying to reward producers that are actually thinking about that end consumer and trying to produce a product that they like. Shear force as well, tied into that eating quality.

And then non-mulesed. We'll have The New Zealand Merino Company speaking here in a minute about the premiums, and we've also got Icebreaker here, who are looking to source that non-mulesed wool. So really look at those welfare traits if you are trying to delight that end consumer. What's that future consumer going to look like in 20 years? Are our consumers still going to be buying mulesed wool?

Establishing your baseline

Once you've run through that criteria, the really key thing to do is to establish your baseline. We now know where we want to get to and what breeding values we want to look at, so now we've got to work out where we're at currently. There are a couple of ways we can do this.

One is a flock profile test. At the moment this is only available in Merinos, which I imagine is most people in this room. A flock profile test is where you take a genomic sample out of 20 ewes from a certain class of ewes. So if you've just classed your maiden ewes, you can take a genomic test of 20 of those animals, and it'll come back with these results here, which display where your sheep sit genetically and compare them directly to industry. That's for about 15 ASBVs.

So a flock profile is a great way for you to benchmark where your sheep currently sit genetically. Going around, we see there are often a lot of people that go, "My sheep perform really well." But how do we actually work out what's genetics and what's your management? So we're really trying to benchmark where the genetics sit, because some people are really good managers and don't have very good genetics, and vice versa. Some people's management might be slipping, but they have really good genetics. So work out what the weak link in the chain is in your sheep enterprise.

The other way you can do it is through a sire team average. If you go back and get the tag numbers of your last five or your last ten years of ram purchases, you can then look up the ASBVs for those and take an average of them. Once you've got that average, that should give you a pretty good idea of where your flock will sit genetically.

It can also be a great way to filter ewes you want to purchase onto your farm. If you're looking to buy in genetics, especially now while prices are down, and swap old genetics out, and you actually ask a few questions of that vendor, you can potentially get their ram details and then do an ASBV average of those to work out what's actually underneath these ewes that I'm going to buy in.

Making a plan

So, make a plan. This is the third part. We now know what this perfect sheep looks like and where we currently are, so we need to make a plan for how we're going to get there.

One thing here is flock structure. Sorry Howard, there's no reference on this, so I've been caught out: I actually did make this up. But this is an example of 10,000 ewes. If you cull 10 per cent of those ewes in each age group every year, for being dry, busted feet or busted udders, and then two per cent die, this is what your flock structure will look like if you're keeping those animals until six years old.

So try to work out the best flock structure for your enterprise. How can we maximise the most profitable animals on your farm, which will often be those two to four-year-old ewes? What's the depreciation in my ewes from, say, a four to a six-year-old? How can we manipulate that structure, not just to make genetic gain but also to maximise profitability? Obviously this structure will probably look different in studs. If they're looking to make genetic gain, it might be skewed, so they're keeping more ewe lambs and more younger sheep that are genetically superior, and maybe dropping these animals at the end out and cull-classing a bit harder on them.

It's something really useful to do. We've probably been seeing it a lot recently: in 2018 and 2019, when it was really dry, people don't have a lot of ewes coming through in those age groups, and then they have a power of ewes coming through in their 2021 and 2022 drops. So their flock structures at the moment are looking a bit all over the place, and it's really hard to balance in a variable rainfall environment. But try to work out how many ewe lambs you actually need to keep each year, and how many you can actually class out.

So have that flock structure set out, and have a measurement and management calendar, so you know when things are going to get done in your enterprise. This is probably really important when you've got multiple staff in your business. Whether you're a stud or breeding your own rams, work out what the key things are that you really want to select for in your sheep enterprise, and when you measure them each year, so you're getting consistent measurement. And for your commercial producers, work out the key times you need to do things. When do I need to condition score? What are the easiest times to do things that are going to give me the biggest reward for my effort?

Choosing rams and ewes to match your objective

The other thing is your genetic selection strategy. You've probably already begun to work out there are so many moving parts in trying to pick that perfect animal. If you ever get to sale day and your catalogue looks something like this, with scribbles all over it, a few ticks, a few crosses, a three out of five or four out of five for structure, for wool, how can we make this process simpler?

If we're going through and trying to pick keeper rams here, we have no pressure, no one else trying to compete with us for a ram. We're sitting here all day trying to pick the best 20 rams or the best 10 rams in this whole drop. But when you're taking 30 data points, how do we actually work out which rams genetically are the best? And then on sale day, if you're sitting up there looking through 400 rams and trying to compete with everyone else that's there, what's actually going to be the simplest way to pick the best rams for you without getting confused?

So how do we select and pick the best rams and ewes to match your breeding objective? There are multiple things out there. If you're looking at rams at a sale, you can use ASBVs, but also match that with your visual appraisal, so things like structure, wool type, whatever other things are important to you. And then also using raw data. That's not more important, but if you're a commercial producer classing ewes, you might only have raw data, so how can we incorporate that to try and pick the best ewes?

Often in a classing box it's all about one person's opinion. What are those individual cut-offs for each animal? Right, that animal doesn't quite make it on the wool, so that animal gets scratched. But did that animal have other attributes that actually moved it more towards that perfect sheep? Often when we go through rams or ewes, we're always making these cut-offs in our heads: we can't have that, we can't have that issue. We end up selecting these animals when we're actually trying to get over here. So where do we make this compromise? You probably don't want to get Ferg and I in a classing box together, because it's just a lot of arguing and working out where the actual cut-off is and what's the best animal to get us there.

A customised index for Mumblebone

This is probably going to confuse it, but often we're looking at over 30 traits when we're trying to pick rams or ewes. We're trying to think about all these traits that go into your breeding objective to breed that perfect animal. So we've actually developed a customised index that we use here, for other studs and for commercial producers, and we can weight it so it changes for every individual breeding objective. This here is an example of a customised index we've made for Mumblebone.

I'll just explain how this works first. If I put this index on and index every animal in the 2021 drop in Sheep Genetics, this is how much better the top 10 per cent of animals in that drop are for each trait, relatively speaking. You can see that those animals are pushing up for weaning weight and post-weaning weight, because we've pushed increased growth on them, but they're actually not coming up as strongly for adult weight. We've put a negative emphasis on adult weight to try and bend that curve, so we have these ewes that grow early and then stop. Without using breeding values, it's very hard to pick those animals that have got that good growth curve over time.

We're still trying to work out what the perfect ewe weight over time is. We'll probably bring this adult weight back until we're at a point that we're happy with, whether it's a 65-kilo ewe, and then just keep trying to push weaning weight and post-weaning weight up from there.

We're putting weight on these carcase traits, yearling fat and yearling eye muscle, which are correlated with condition score. So we're trying to improve genetic condition score, and Ferg's going to talk about some of this stuff later, and the correlations with whole-body energy in sheep. So we're trying to increase the durability of those animals.

Fleece weight: we're putting a bit of an emphasis on yearling clean fleece weight and adult clean fleece weight. At the moment the industry average is about 17 for fleece weight. Here we've gone, right, we really want to move all these other traits around that fleece weight, but we're actually quite happy sitting potentially between 17 and, say, 25. If you're further west from here and running at lower stocking rates, maybe you can handle more fleece weight, whereas if you're further east from here and you've got higher stocking rates, maybe selecting on condition score would be more useful for you.

And then fibre diameter. For the sheep here, at the moment we've got a slight emphasis on decreasing fibre diameter, but in reality we're quite happy to keep that at 19 micron. Yearling FDCV, the coefficient of variation of fibre diameter, is an interesting one. We actually don't have a large weighting on that in the index at the moment, but it's coming along for the ride and keeps turning up in these animals that are good for everything else we've put in this index. Ferg will actually speak about some of the correlations and why that is later on today.

We're pushing up yearling staple length a bit, and staple strength is also coming up. There are some good correlations between staple strength and FDCV, so there's not much of a weighting on that, but once again it's coming along for the ride. Worm egg count: trying to reduce that, so breeding animals that are more resistant to worms.

And then, I'm going to say it again, we're looking at those non-mulesed traits. Breech wrinkle and breech cover actually don't have that much of a weighting, but a lot of these animals that are turning up are actually really good for breech wrinkle. If you look at the correlations, it's got good correlations with fertility and with growth. So it seems that these animals that are coming to the top are actually quite bare around the breech. And we're really trying to push on dag, so reducing dag in those animals. We're starting to work on urine stain, so hopefully we'll have that in there in the future, so we can have these animals that are really resistant to flystrike.

We've split up the reproduction traits into conception, litter size and ewe rearing ability. Instead of just pushing weaning rate, which is mainly driven by litter size because it's the most heritable of those traits, we're really trying to push conception and ewe rearing ability more than litter size.

IMF and shear force: we've got IMF going up. We've got a small weighting on that just to try and keep it ticking along, and it's also positively correlated with fat, so as we're driving that, we see IMF going up. Shear force we're trying to hold where it is. Obviously there aren't really many premiums in the market for that at the moment, but shear force is going to be negatively correlated with growth, so we don't want to breed this really high-growth animal and lose these eating quality traits, like has happened in some breeds. So we're just trying to keep that where it is.

And then these wool traits here. These are relatively new breeding values, but we're looking at fleece rot, really trying to reduce that, especially after a few wet years, plus fleece colour and wool character.

I think this probably makes it pretty obvious why it's very hard on sale day to try and find the perfect animal for you, and why in a classing box it takes us all day to work out what actually are the best 10 or 20 rams here to keep. It's actually really difficult trying to balance all these traits and work it out, which is why we've tried to develop an index that we can customise to each business or each stud to pick out the best animals for them.

Using an index on sale day

So utilising an index is really about trying to maximise genetic gain towards your breeding objective. It's really handy for ranking ram catalogues. On the day, when I go around, you'll often see me just looking at one number, which will be that ram's rank on the index, and then I'll have a couple of visual traits where I might give it a score out of 10 for feet and a score out of 10 for wool type if needed. So really try to simplify it on the day, so that when the auction's on and things are hot, you're looking at as few numbers as possible.

Searching semen sires: if you're looking outside of the industry, it can provide a direct comparison of where your sires currently rank against semen sires or against rams in other catalogues, so it's a great way to compare across catalogues as well. MateSel as well, which is a tool that we use for breeding, basically trying to maximise genetic gain: you put an index into that and it'll try and maximise genetic gain while minimising inbreeding.

We've also got an index that we use for ewe classing. We can put raw data into this index as well, which I'll talk about in a minute. By putting raw data in, we can try and keep an eye on some of those traits as well, like structure, and utilise that data instead of looking at 10 potential data points we've recorded.

Where to from here

So where to from here? As I said at the start, at the moment we've got an index with 24 traits there on the screen. But if there are things outside of that index that are in your breeding objective, we really want to make sure we capture them. At the moment we don't have structure and we don't have footrot in there, but we're hoping to make them breeding objectives, because we know as soon as we put data behind something, we can move it. We can make more genetic gain towards it. So we're looking to put in those structural and footrot traits.

Other unmeasured traits: working out where we sit with things like pigmentation, and then other coloured fibres. If we start scoring those, can we link to that data, and do we want to put that in the index? Trying to put an economic value on pigmentation is very difficult, though.

And then artificial intelligence. Angus was here before, talking about AI, and I've been trying to elbow him for a while, saying maybe we should move one of your engineers across and try to utilise AI in breeding programs. Now we've got all these traits and all this data, how can we use AI and computer models in the future to assist us in breeding a better sheep? Can we incorporate AI to breed sheep in the future? I think this space is pretty cool.

And yeah, that's about it from me. Hopefully I didn't go too far over time, Josh. So yeah, thank you.

Josh: Thank you for that insightful little look at Mumblebone and the data collection at Mumblebone, because they are collecting a lot. I appreciate your help here, because it is a lot. This book here, do you want to talk about that?

Henry Hickson: Yeah, so Josh got me up here to market, which is a big strength of mine. But basically we've got this ram buyers' guide. All the traits that I briefly touched on before are all in this book, and they're all explained. If you're crazy enough you can read it in one sitting, but it'll probably put you to sleep. It's designed to go back and look through.

We know that everyone on farms is busy, and often they just come to a ram sale once a year, and often they come there to talk about the footy, catch up, eat a sausage sandwich and have a coffee. So try to utilise this beforehand: look back at the breeding values and work out which ones you want to look at. If you want any more information about this, come and see us. There are copies of it down on the table there, and there's also a genetic plan guide, which runs through the ASBVs in a similar format to how I did at the start of this presentation.

Josh: Thanks, mate. Any questions?