Mark Ferguson: Welcome back to Head Shepherd podcast. This week, Tom Gubbins from Te Mania Angus Australia joins us. Welcome, Tom.
Tom Gubbins: G’day, Mark. How are you?
Good, mate. Thanks for your time. If you could just start us off with a bit of the history of how the herd started and how your role has evolved over that time.
Originally Te Mania started in New Zealand in 1928, I think, with the closing of the Angus Herd book. My grandfather registered animals in New Zealand, and Mum and Dad purchased 23 cows and a few bulls from Frank Wilding and brought them to Australia. They started Te Mania Angus in partnership with Frank and then moved off by themselves and became Te Mania Australia.
They started pretty heavily into quantitative breeding straight away. I can remember as a small kid having Russell Sully from the Ag Department out with the slide rule, doing all the weight ratios well before computers. I'd love to know how one works. They were certainly doing quite a lot of sums with it pretty quickly...working out weight ratios and making a hundred the mean and then showing what was lower or greater than a hundred...sort of the distribution as a phenotypic:weight ratio system.
Then we moved to the in-herd EBVs, which ripped out all the environmental variation and just gave us the genotypic variation. Later on, Te Mania moved into BreedPlan. We were involved at the start, putting all the Angus together into BreedPlan and creating the linkages so that we could compare between breeds and genotypically, which was pretty exciting. We've followed that way of doing it and have watched over a long time now the targets and objectives being reached.
Cool. Obviously, small beginnings at 23 cows, what are we running there at Te Mania today?
We've got about 900 cows joined and 690 heifers. Then we've got a 500-cow embryo programme up at Mansfield that James McCormack runs.
It keeps us busy, and apart from recipients, all of them are registered. All the recipients are downgraded cows that are a good life support system for carrying a calf, but they're not things that we want to breed from anymore. So we're able to get the top 10% of the cows producing 500 calves a year in that program, which is pretty amazing. A bit of genetic gain there.
We've been ET'ing since the early eighties. You've got to do all the other artificial breeding first before you start ET'ing, I think. It's a diminishing rate of returns, spending money on breeding.
You've got to do the macro things first and then start doing the micro things. And ET is definitely one of the last things that you do to try and get genetic gain.
So, James took all the recipients away. He was a manager of where I am now, and he went home to his family farm. We did a contract with him to run all our recipients for us, and he's been doing that for 10 years.
So they're all frozen embryos that head up to him, or are they fresh?
No, he collects them as well, actually. So we send him recipients and cows, and he just sends us back weaners.
It works well, and there's a lot of autonomy for him in the contract too, so we don't have to bother him too much. He gets paid on key performance indicators of embryos and all sorts of things. So it's good, it works really well.
Excellent. Obviously EBVs have been a massive part of genetic gain. I'm sure it's a big part of who gets chosen to be a life support system versus those who get to be flushed.
One thing I would be keen to chat about is how we know that EBVs work for production traits, but I think they have just as big a role to play in things like foot structure, temperament, and other things that might be considered 'stockman' traits. But really, the data speaks volumes in there as well.
That's right, it does. We've been involved in structural EBVs now for quite a while, since the early 2000s. We've been collecting structural data since the validation project in the eighties. We are currently using a different EBV model to Australian Angus.
They were using the version we use, but they've dumped it for the American version. We've stuck with the old one, which is a threshold EBV. It measures the number of animals that have got suitable foot structure and represents that as a number of animals that have got acceptable foot structure. A bit similar to a calving ease EBV, which does a similar thing.
Whereas the American Angus one has said that a '5' foot structure is ideal and anything above 5 is less than ideal. So the goal is to get to five, but we actually don't biologically know whether five is ideal. We just think it is. And I think the biologically ideal is probably more where the population mean sits.
If the foot structure on claw set, for instance, is 6.2. Well if we aim for '5' we may be breeding past the ideal, back towards the not ideal. So I'm a bit sceptical about that.
It does intrigue me and it's a discussion I've had a few times, what is ideal? Because the ideal is set by humans guessing. Things like the pastern angle, when you look at a natural population of Red Deer running around New Zealand - their hocks are just about touching and their pasterns are pretty springy. Yet if that was a cow, we would cull it.
I guess the first job is to characterise it and then work out the ideal after that.
The only way to really quantify it is to do a huge research project where you measure all the foot scores and leg scores, etc., and then get longevity in the cows, or the dropout rate for structural reasons out of finishing steers, to see whether there's actually a lameness correlation to your structural scores. Because that's why you're doing it, isn't it? And no one's actually really done that. So in the absence of that information, I think that we just should go for the mean biological score, which animals seem to cope pretty well in the environment with.
That's a good point. Temperament is obviously something that has been carefully watched. So, you're generating the breeding value on that these days?
So we look at the temperament score. We've been doing a crush score. That's 1-5, one being really, really docile. Animals that you can touch and they almost don't react, they just walk into the crush, get a treatment and walk out. All the way through to five, which is an animal that would actually attack you if you went into the yard with it.
And then the other three scores are in between those two, so they're fairly easy to subjectively differentiate between. Those five scores then go into BreedPlan and hey presto, there's a pretty damn good heritability. Which, if there's a good heritability across herds and across environmental groups, that means that's the audit for me to say that it's working pretty well.
Have you got many buyers chasing docility? Do people actually look at that breeding value these days?
Look, probably a little bit too much, to be honest. But if that's what they want, then the information's there, and they can choose it.
I just wonder whether they're putting a little bit too much selection pressure on it and losing some selection pressure in some more profitable or some other traits that might give them a much bigger bang for their bucks. But that's all right. We present all the information and offer the advice, and then people can make their own decision.
But they are definitely selecting on temperament, and the temperament EBV is working. It's working well.
Just go back to the structural one on foot for a second. We've found that, particularly for BreedPlan and EBV sceptics, or should I say someone who's at an early stage of understanding, the structural EBVs have been really good at explaining to people the genotypic vs phenotypic effect of structure.
Because you can see an animal standing there in front of you that has perfectly good feet. But it doesn't have perfectly good EBVs. If the animal's going off to King Island or Southern Gippsland, where it's wet, and the feet are prone to growing out badly… the animal is phenotypically standing there with good feet, but the genotypic information from the EBV is saying that the progeny aren't going to be nearly as good as the father. And so the people are actually making decisions on that when they might be sceptical about the 600-day weight EBV. Which really fascinates me, because I think that's what people have got to understand - the difference between phenotype and genotype is quite significant.
I think you're right. Structural traits are a great example of things that, if it's in a really nice dry environment, you're going to get a different outcome when you chuck it under rainfall. EBVs allow you to see that. Whereas you can't see that unless that animal's been exposed. You basically get to predict the future, that's what EBVs are; you become a fortune teller.
The more difficult thing is that we've got to try and predict the future, actually. So that we can work out what those EBVs need to be in 10 years. That's the tricky bit. We're always doing that here, trying to work out where we need to be in 10 years. Because that's about how long it takes to get it organised.
Even with full-on ET and all the rest of it, cattle are still slow. They still have a nine-month gestation and one calf at a time.
Not only that. You've also got to think, well, we've got something that we need to address. Now we've got to start collecting the phenotypes for it and then get a research project up to actually work out whether there's a heritability.
So all that takes time, too, before you even start breeding them.
When you and Lucy are discussing that 10-year plan, are you taking a bit of a lead from bull-buying activity, or is it all from consumers or science? Or is it all a bit of a combination?
A bit of a combination. At the moment, we're taking a bit of advice from social opinion about animal welfare and climate change. Whether you believe in climate change or not, I think that if you don't think that it's going to change the economics of agriculture in the next 10 to 20 years, then you're missing the whole social understanding of what people want.
People are feeling guilty about the environment, and they want to have that guilt removed. So they want to be buying products that remove that guilt. And we can do that genotypically for them if they're going to pay for it.
So, at the moment, we're looking at social things. We obviously ask the bull clients, are there things in the herd that we think that we're addressing, but perhaps we're not? To make sure that what we're trying to do is happening.
We're looking more at the clients’ herds to see whether the decisions of the past have turned out to be making the change in those herds that they wanted and what we envisaged would happen.
The trends in the beef industry really need to come from the consumer. Marbling came from the consumer. It didn't come from the bull buyers telling us they all wanted marbling cattle because if we had done that we still wouldn't have acted.
It's like that Henry Ford comment, “If I'd asked my clients what they wanted, they would've said faster horses.”. I think that applies in this situation. We've got to be a little bit further out in front.
One thing that you started a long time ago was Team Te Mania. Which, in my understanding, is a powerful part of your business in making sure you are keeping track of things more than what you can see in your own herd. Do you want to tell us a little bit about what Team Te Mania is?
Team Te Mania came out of the Beef CRC 1, which was a big beef research project here in Australia to work out heritability of carcass traits and other animal characteristics. We needed a lot of money to fund the research, so it couldn't come out of commercial businesses.
So they set up progeny test systems in commercial herds and then linkages. They collected carcass data and processed and worked out heritabilities of all sorts of things. When that project finished, we went to the collaborators and our clients that had been involved in it, and we asked them whether they'd like to come into a partnership with us to continue that on. That became Team Te Mania. And so it meant that we supplied bulls and semen into those businesses.
The semen became linkage between the commercial herds, and the bulls became the animals that we tested. We continued to collect phenotypic information that had already been proven under the Beef CRC to be heritable.
We kept on doing that, and then the indexes got built, and value got assigned to those traits, so the index started to reflect the extra value of fat or negative fat or marbling and yield. That slowly started to change the direction of the business.
We at Te Mania Angus had live carcasses coming in, so we can scan for marbling using an ultrasound machine. But the correlation between the ultrasound machine and the actual carcass score is about 60%, so it's low. But you can get slow genetic gain over a long period of time from that, and we did. Once the carcass information starts coming in, the data difference when you get a real score compared to something that's correlated to it is quite significant.
Team Te Mania evolved out of that, and as it's gone on, it's become a marketing group, which we didn't really foresee. There were a few things we didn't foresee. It's also become a really good sounding board. We didn't realise at the time, but actually having independent businesses who are responsible and accountable for their profit and loss statement, questioning you over the genetic direction... It's quite different from you questioning yourself over your genetic direction. It's actually a little more cut and thrust, as you can imagine. It's been very good for us because it means that we've had to listen, and we have a good discussion about it. And it's gotten stronger and stronger. So the relationship and that point of, possibly conflict, almost, has been very, very beneficial.
Because your genotypes are predicting IMF, obviously you need to keep feeding those phenotypes in to keep the genotypes working well?
That's part of the issue with single-step. If you lose your phenotypes, the relationship matrix is related to nothing. So it's got to be there, it's got to keep on pouring in. This is quite disruptive, because in the past, the whole EBV or BreedPlan pipeline has been that if you put some data in, you get some benefit. Now with genotyping, if you put some data in, everyone gets a benefit. So that's changed the economics of the way the pipeline works.
As a consequence, if you put some data in, everyone gets a benefit. Perhaps you might think that you may not have put the data in at all... It's much more philanthropic now to actually put data into the system because with genotyping, as long as there's a good pool of data, you can get what you want anyway, perhaps with a little bit less accuracy. We've got to do some work in that area to work that out so that the data is rewarded properly and that we don't start losing that nucleus of information.
I think that's a two or three-hour podcast just in that alone and trying to work that out. But there is structural change. It's going to have to be changed because there has to be a reward. Nothing happens forever unless the economics line up with the activity, generally. It'll last for a while but eventually you have to line it up.
That's right. We could just keep on producing funded projects in Australia. We're doing these beef information nuclei, but we've got to keep funding it.
It'd be better if we tweak the model so that information just naturally came out of commercial transactions, rather than having to go and find levy payer money to build research projects to find the data. I think with industry data in the future, with the density of chips and also the AI technology being able to scan through masses of data to see which ones could be added to the genetic database without anyone even knowing possibly, those sorts of things could happen in the future.
I think you're right. I think the combination of the technology phase that we're in at the moment means that there will be change, it's just a matter of time.
If we stick with genomics for a little bit, I'm interested in your opinion whether it's been a game changer for your business or whether it's been an incremental improvement? I think it's fair to say that across the wider industry it's been a game changer, but for individual businesses that are already at the front of the game, I wonder whether it had that bigger impact?
It has had a big impact. It has had a bigger impact because we make fewer mistakes, and that's what accuracy in EBVs is all about. EBVs are an estimation, and obviously, the critics fly in here…. But they also come with an estimation and a standard error, which is the accuracy. And the accuracy has improved. Every time you get an improvement in accuracy, the decisions you make are more potent and they stick more. So, you can pull younger animals out and speed the genetic interval more, and with less risk, because you're not going to have as many failures.
Even though you may look at it statistically and see the accuracies only come up by 6 or 7%, when you're in the high fifties or the low sixties, to come up 7% is pretty significant. And it has made a huge difference.
We in the team need to be thinking in the future how we can do more and more phenotyping, really. To start to look at harvesting data out of the non-progeny test groups and things like that. There are potential advantages in grabbing bits of data. But all very hard to quantify though.
Definitely, but I think what you say is true. We see very small differences in the accuracy of the breeding value quoted, but in terms of the error made by using a ram lamb or a yearling bull, in your case, is way less. It helps you work out which grandparent has got more of or whatever and therefore, which traits that that young animal is going to be superior in. And so, that certainly has reduced the error.
One thing that's happened too is that the relationship matrix is more accurate. Before genotyping came along, we didn't really know…. All we had was the stockman's skill at working out which calf belonged to who. Particularly, which heifer, I mean, God, have you ever been out in the heifer paddock and seen them swapping calves? It's just unbelievable. And what happens to a relationship matrix in EBVs when the calf actually doesn't belong to its mother or its father. Now it's got 50,000 parents, which is pretty wonderful, isn't it?
Yeah, massive sorting that out.
Possibly a bit of a rabbit warren, but we'll go down there anyway. I'm interested in what Te Mania is up to in the maternal/feed efficiency/methane space and where you think the important things are going to be and what you're doing about it?
I know in NZ, but also here, there's a wonderful cow with the doability and she's a ripper. She's just always fat and easy to run. Is she eating more than all the others, therefore taking more of her fair share of the pasture that you're putting in front of her? What is she doing? Is she not putting the energy into the calf? Is she just smaller and less productive? Has she taken longer to get there?
There are all sorts of questions, and I think we argue about it in cattle because we actually don't know the answers. So it's the best place to go if you want to have an argument. Go to a place where there are no facts, and you can scream at each other for as long as you like.
We need more research into it, and there is a Trans-Tasman cow productivity trial with MLA and Beef and Lamb New Zealand at the moment to try and collect eye muscle areas.
We don't talk about the amount of lean that is used in seasonal transition. Cows are not just using fat. They're using their muscles as well. Is it more important to feed a cow every day and not have them putting on and off weight? Or is it better to have them putting on and off weight and not feed them any? Or does that also vary from environment to environment?
There are masses of questions in here that need answering.
That's my favourite area of discussion, probably because you're right, there is no real answer. So we talk about and have a theory that can't be challenged, but I think we will get to the bottom of it eventually.
It's obviously a real struggle because the tools have been sparse in terms of working out what animals are doing when grazing pasture. We're getting closer with some sensors to predict feed intake. But we still don't know exactly what they're selecting out of that sward, even if we knew what their total selection was.
So, there's a lot of work to go, and we're at an interesting stage in history, and we'll get some technology to help answer those questions. But really, there's lots to work out. There are so many variables that we're trying to play around with that there's no wonder we haven't really got the answers.
But just because a cow's fat and another one's thin doesn't mean to say that the fat one's efficient and the thin one's inefficient. In fact, I think that that is so far from the truth. I'm not blaming one or the other either. It's just way too blunt to say that. We see it in people saying, "Fat increases fertility." Now we know that if we want to get all of our cows in calf, we have them on a rising plane of nutrition and in a good condition score, and they'll get in calf easier than if they don't.
But we also know that when we preg test some years, we look at the dry pen that's full of skinny ones, and sometimes we look and there are quite a few fat ones in there. It's really weird, but I think the phenotypic correlation between fatness and fertility is the fatter animals getting in calf. But genotypically, that's not the case. Genotypically in cattle, that correlation isn't significantly strong enough to put emphasis on fatness in cows to improve fertility.
Other than in heifers? The science on heifers was true, wasn't it?
Yes. There is a small statistically effective effect on heifers, but when they grow up to cows, it's not there. What was that? Was it perhaps maturity? Was it something else like the early maturing heifers got fat?
If we'd had maturity in that research program, you probably would've gone, "Wow, all the early maturing ones are in calf!" Which we actually know now to be the case, that maturity has a very high correlation to fertility.
And that's probably what happened, and that's probably what fat is really. Again, this is an area that we can go around in circles for a long time. But I think maturity patterns are at play and whether that's because there is more being laid down or is it just being laid down earlier or what? But anyway, as you say, there's lots to learn there.
Hopefully, we might have a bit on that soon because we've put heat time collars on all our heifers. We had 670 heat time collars on our heifers last June. There'll be nearly 900 going on this year. That will tell us the first heat they have in their life and then whether they have heats all winter. So we've got two years of data so far, and it's going to be crunched in the next six months to give us some idea.
What we're trying to get is the first heat that the heifer ever has. And we're getting heifers that are getting in calf to a fixed-time AI programme at 14 months that haven't had a heat at all in their life. I'm just not sure whether we should be letting them loose. They should probably slip off the editing table, those ones.
So their first ovulation was to the fixed-time AI?
Yeah. I don't know, would they have been picked up by a bull? Possibly not.
So what's the heat time sensor measuring? Is it measuring a combination of temperature and activity, or is it just activity?
It's rumination and activity. As soon as they're in season, they stop ruminating and they start walking. When you look at the graph data, every 21 days there's a little spike of activity and a drop in rumination. It's clear as day, every 21 days. Then you look at where you think that first one is, and you look back 21 days before that and there's a tiny little blip in the data even back there.
So is the cow that got pregnant to AI, is she one that just is a bit different? She keeps ruminating and doesn't start walking? Correlation and causation are interesting, aren’t they?
Absolutely. So that's a possibility. The data coming out of the dairies on how accurate this is is pretty outstanding.
Where are you at with measuring methane?
Well, for net feed intake, we're putting machines in next month, and we'll start with 132 animals at a time. That's the size of our contemporary group. We designed it around that. Our biggest contemporary groups are 125, so they'll be able to handle one contemporary group at a time.
They're going to go in there for 70 days and come out with how much they put on and how many kilograms of feed that actually took and relate that back to their body size. And it can tell us about net feed intake. So that'll be interesting.
I feel we probably should have just skipped that and gone straight to methane. The methane machines are pretty damn good too. Perhaps we need to put both in. So that's a decision we need to make here in the next little while.
There's a very big variation in animals' methane output, and so for a genetic program like ours, big variation is exciting because it means we can make significant changes to the amount of methane that the animals are emitting, genotypically, pretty quickly.
But some of my concerns are how are we going to account for it? How do we say, "Oh, well, our cows don't produce methane. See that one over there? That one doesn't and that one does"?
I think it's actually an opportunity for your expertise to start to move to the front academically and say, we know that that animal genetically produces less methane. It's actually probably an opportunity for geneticists to say to the industry, "We know that that is the case," even though you don't have to test every individual. We can create a pathway of animals that have come from this particular bloodline, and they'll produce quantifiably this much less methane.
I think it is time. Back in 2009, we started feed efficiency stuff in sheep. I ran the argument that it doesn't matter because no government will ever be able to legislate on how much methane that farm's producing based on the genotype of the animals. I think we're seeing now that it won't be about governments, it will be about brands, and brands will reward people that are making change and have got some science behind the changes they're making.
I think there'll be a combination of legislation plus brand leadership. And I think that what we've seen in other welfare or other maybe social-driven things that have ended up in our breeding objective, that it's not ever by legislation. It's more about who's willing to pay. I think that will be the same in carbon.
Completely contrary to the left-wingers that go on about modifying people's behaviour and demonising cattle, the capitalist structures in our society are actually now enforcing it.
We're finding that the corporations in ag in Australia, to get money to invest in ag, need to show a very, very clear social conscience on the environment and animal welfare. So they're doing whatever they can, and so I think the drive's going to come from them.
Yeah, I should have said big business rather than brands.
And also coming from the consumer demand. But consumer demand is nowhere near as powerful at the moment as investor demand.
Really good points. Well, that's all going to play out, and we're going to see that, certainly in the next decade, come at us pretty strongly.
In your selection strategy, you are sitting there writing your 10-year plan. You've got things like methane, feed efficiency, maturity pattern, all those things that aren't currently in any Angus index. How do you weigh all those traits to work out which are your top heifers that go into your ET program?
Are you trying to bring in those new traits into that thinking? Or do you have another program that does the blue-sky stuff?
Until they're fitted into the index, you've got to add them in subjectively, which is a mess.
If I can sort of drill into that a little bit, the weighting on a trait in an index is mathematically calculated for that trait's economic importance over scarcity and value.
Each of the 20 odd traits in the index are all balanced according to their value in the economic pathway and the environment that those animals are running in. It's pretty intense mathematics and economics and genetics all pooling together there.
When we don't have a trait included in the index, we've got to add it. And it's dangerous ground because we have to subjectively say, "Oh, we want a bit more of that," and have no knowledge of the damage.
But it's not as bad as not having a genotype at all. It's not as bad as saying, "We want a bit more spring of rib in those." So, I'm going to put a 50 percent selection pressure on spring of rib before I even look at EBVs. Then you are seriously going to start going backward because spring of rib doesn't, according to every single bit of information I've ever found, have any economic correlation apart from all the people that are potential buyers of those animals seeing them and liking their spring of rib.
Do we know how heritable spring of rib is? Are we even looking at something under genetic control?
We don't know that.
What are the opportunities in the future for beef producers, for bull breeders? If you look from the sidelines as I do, bull prices have been scooting up over the last few years, and that can mean structural changes are coming. Are we going to see any shift in how beef genetics are traded in the next 20 to 50 years? Or are we auctions forever? What do you think is going to happen to the genetics industry and beef?
We do have an internal wholesale bull calculator. What it does is it has a stab at working out the mean bull price this year, according to the Eastern Young Cattle Index which gives us an idea of the commodity price of beef. And then we multiply that by a factor, which gives us the mean bull price. And that mean bull price is our average genotypic index score.
So we find the 50th percentile band within our herd, and then we increase the value of the bull according to their economic contribution in the index. If the mean is $12,000, then as they go up the index scale, they get incrementally more money according to the extra contribution they're going to make to the herd that they're going into. And down the other side, on the other end. Because an index is an indication of its progeny difference, so the index is what that bull's contribution will directly be to the calf.
What we do is we look at the auction price over time and then compare it to work out what the factor is. But it's pretty good. A long time ago, it used to be more indicated on the wool price actually. So if someone got a good wool check, they used to spend a bit more money on their bulls, which was a bit crazy.
I think we see the same in grain prices affecting ram values. If the croppers have had a good year, the ram price goes up I reckon in Australia at least.
Anyway, it's been a great chat, Tom. Thanks very much for your time.