Articles, Blog

Grains Research Updates 2015 | Goondiwindi | Irrigated wheat variety performance 2014 – A. Peake


I guess before I go any further, I just acknowledge
my co-authors who’ve been helping me out greatly. So Matthew Gardner from AMPS, who has been
doing a lot of work down on the Liverpool Plains with the project; Kerry Bell, a statistician
with DAFF, who does a fantastic job of putting up with me and working with unusual trial
designs and difficult analyses; Nick Poole from FAR in Victoria, who’s experienced
in high yielding grain production, and obviously acknowledge GRDC and the Federal Government
and growers, of course, for putting in the money that actually funds the work. So a little tiny bit more background: this
project is running between 2012 and 2017. This coming year will be our second-last year
of experimentation. And we’re aiming to try and produce better yields and less lodging
under irrigation through agronomic means. Lodging is a big constraint to yield in irrigated
systems. We’re working in conjunction with a better irrigated wheat germplasm project
that’s doing some pre-breeding to try and improve the genetics. And last year we ran
some trials in Emerald, Narrabri, Spring Ridge and Gatton. I’m not going to talk about
the Gatton trials today. I normally wouldn’t have come this year to talk. I try and talk
every second year and I like to get two years of results before I update people on what’s
going on, but we got some quite interesting variety results last year that I thought people
might want to know about. Because last year was different to previous years, in terms
of its weather patterns, and that caused some different lodging in varieties. So we trialled 18 varieties across three of
those locations: Spring Ridge, Emerald and Narrabri. The varieties were all sown on two
sowing dates and two N regimes. The trials are fully irrigated but as with any irrigated
farms sometimes pumps break down and people get sick and so they probably weren’t perfectly
irrigated, but it was a pretty good irrigation regime last year. And we use a comprehensive
fungicide and insecticide program to make sure that biotic stresses don’t limit yield.
So we don’t want any fungus or insect stopping those varieties from achieving their potential.
So I guess it’s important to understand that some of these varieties have disease
susceptibility issues that would need to be addressed with a chemical program on a commercial
farm. And our target plant population is 100 plants per metre squared, calculated on the
bed area, so we don’t include the unsown furrow in that calculation. So the aim of our N regime is to get a very
large amount of nitrogen on to make sure that nitrogen is not limiting to yield. Now, we
generally do that in two ways: with an in-crop or delayed N regime where we only really need
about 70 kilos of N per hectare at sowing – that’s combined soil plus fertiliser N.
In the sowing N regime we try and put most of that nitrogen up front, because we’ve shown
in the past that the in-crop N regime does significantly reduce lodging and increase
yield. And we’ve got a sowing date comparison that happened as well. And that comparison
was early sown versus late sown: mid May versus the end of May at Emerald and Narrabri. And
a little tiny bit later for those two sowing dates at Spring Ridge. So I’d like to throw some cautionary notes
at people: varieties and agronomy perform differently between environments and years
and farming practices. So I always encourage people, if they’ve seen something new and
they want to try it on their farm, to try it on a small scale first. I find it a bit
scary when I talk to a grower about in-crop nitrogen application and he has never tried
it before and he wants to go and plant his whole farm to in-crop nitrogen. I say, “well,
just try it on one paddock first and get familiar with it, so you understand the ins and outs
of it.” As I mentioned before, they’re disease-free
trials. Keep that in mind. Some of the varieties may be susceptible to diseases. And check
on marketing options for these varieties too. There’s two new varieties that we trialled
last year from the southern markets: Cobra and Trojan. They did extremely well, but they’re
not being marketed in the north particularly. Ben might want to say something about that
later, but understand how you’re going to market them because their quality is still
provisional. And, so yes, we got some different results to previous seasons, so we’ll start
talking about that shortly. We’ve started to and characterise our irrigation
experiments and seasons using something that I’m calling, for want of a better term,
radiation degree days. So how much radiation, which is what the plants grow on, do you get
per unit of temperature? And this is an important measurement because temperature makes the
crops – higher temperatures make the crops flower faster, roughly speaking. The cooler
the temperatures you get for the same amount of sunny weather, the better your yield potential.
So we’re doing some analyses where we’re looking at that trend of that ratio between
radiation and thermal time over the season. So that’s time progressing through the season,
and that’s the ratio of radiation to thermal time. And up the top – lines that get up the top
are the cooler years where you’re getting good sunny days and it’s fairly cool weather.
The lines that are down the bottom of the trends are – of the graph are the hotter years.
So we’ve done this for 25 years’ worth of data just to give us something to compare
with. Now it was quite interesting: 2013, for those
of you who remember it, was a really hot, dry year. The season just went like that.
Everything moved so quickly. Now, this data is just for Goondiwindi, but this trend is
– it’s not exactly the same but it’s pretty well similar over, you know, Narrabri, Gunnedah,
Emerald. You know, the shapes changes a little bit but generally speaking the trends are
the same. So 2013 was a really fast season, and we actually struggled to get much lodging
in 2013. The lodging that we got was very late, often after physiological maturity before
harvest. So we got data, but how good was that data? We’re sort of starting – the
more trials we run the better picture we’ll get of that. 2014, however, was quite different, and it
was quite cool at the beginning of the season. And these conditions are what we know produce
greater lodging risk. We’ve shown relationships between the amount of what we call vegetated
biomass – so the lush thick growth that you get at the start of the season; lots of tillers;
big fat leaves – this is what creates lodging risk because it weakens the stems of the plant,
and it weakens surface root systems. So those conditions are part of what contributes to
high lodging risk. It’s not the whole thing, but it’s a significant part of it. So it
was quite a different year to 2013 early in the season when that lodging risk is developing. And for comparison, there’s 2008, which
was the big lodging year when a lot of people grew irrigated wheat that lodged. And that
was a year that really kicked off this sequence of projects. And it’s interesting that it
wasn’t a massively high lodging-risk year early, but it did, through the middle of the
season, and this is another period when the crop’s height is developing, and height is
another thing that’s related to lodging. The taller the plant, the taller the lever, the
easier it is to go over. So – and the other thing that was in 2008: most growers had an
awful lot of nitrogen in their soil. They were cotton growers who had a failed cotton
crop or no cotton crop. They had massive amounts of nitrogen. So we’ll keep running our experiments,
and we’ll keep trying to relate the trends in our experiments to the different years
and just see what we can find out in terms of which years are creating the biggest lodging
risk. Unfortunately, our N regimes didn’t work
last year. We had low N at sowing at all of the sites to 90 centimetres. But at Narrabri
and Spring Ridge they were ultimately high fertility paddocks. And we had faba bean stubble
in Narrabri. We had high organic carbon at Spring Ridge. And we didn’t see a nitrogen
response particularly at those sites. So a small inconsistent response at Spring Ridge,
and no response at all at Narrabri. And so even though we started with 70 units of N,
which is our normal target, it was a good learning experience for us, because we don’t
normally work on such high fertility paddocks. We’re normally working on cotton paddocks,
where the cotton has pulled everything out of the system. And so we’ll have to develop
some new N regimes for that kind of paddock. At Emerald – so, yes, these sort of organic
N sources are breaking down rapidly under irrigation, and the crops just aren’t really
responding to that early N differential. At Emerald we also had some units of N shallow,
because we had grown a forage sorghum crop to try and pull all the N out, but there was
actually fairly substantial soil N below 1.2 metres, and so we think the crop was accessing
that pretty quickly and again we didn’t get much of an N response. But, anyway, enough
on the N response. We didn’t achieve our best practice N regime, and we didn’t get
a significant effect at Emerald and Narrabri. So hooking straight into some results: so
what we’ve got here is a range of varieties. I’m feeling a bit intimidated with Alison
Kelly in the room, a statistician from DAFF. I forgot to put my –
I forgot to put my LSDs on the graph. The
least significant difference on the yield graph is about half a tonne to the hectare.
Lodging is really difficult to do a least significant difference, because we often have
to do what we call a square root transformation. So the LSD actually changes as the data gets
bigger, so I’ve just thrown all this in there so Alison knows that I’m thinking
about these things. Okay. But anyway, so let’s look at the variety
results, because that’s what this talk is really about: which varieties were doing different
things last year. And what we saw was this new thing from Southern Australia, Cobra,
did very well at Narrabri on both sowing dates. Crusader, which we know is a low lodging type,
did well. Livingstone and Merinda, which we also know are good MR lodging varieties. Mitch,
Sentinel, Suntop, Trojan and Wallup all did consistently well across both sowing dates.
Now, the theory on lodging is that the earlier you sow, the more lodging risk you create,
because you give the vegetative phase longer to produce more biomass. So Narrabri matched
up with that theory in that the early sowing date, this black line, tended to lodge more.
So when I analysed the Narrabri data I thought, beauty, everything’s working as we expect. But when we went to Emerald, however – so,
just pointing out again that the lodging susceptible varieties are the ones that yield less. We
did get pretty good correlations between yield and lodging, and I’ll talk about that a
bit more later. But Emerald really surprised me. We actually
got the opposite trend at Emerald, where the early sown trial yielded more and lodged less
than the late sown trial. And that – I had to really get my head around that, and I go
into a bit of detail in the paper in the book if you really want to, I guess, hear all of
the possibilities that may have caused that trend. I’ll talk about that again in a second,
but just to go over the varieties again: similar varieties were doing well. So Cobra did well
on the early sowing date. Caparoi did a bit better. Kennedy seems to really perform at
Emerald and at Gatton. It really likes those hotter environments. We didn’t get the same
sort of yields at Emerald. It just wasn’t quite as good an environment for us. We had
a few establishment problems in the paddock. Mitch, Sentinel, Suntop, Trojan, Wallup did
reasonably well on the early sowing date. Trojan not so good on the late sowing date.
So similar kinds of trends in varieties. But what we saw at Emerald was if we looked
at the lodging progression after flowering time – and lodging is most important to measure
during grain filling when the crop’s still green, because that’s when lodging really
upsets light interception and it affects yield the most. And so these are the two sowing
dates, and we’ve sort of overlapped the data and a little bit to look at the relationship
between lodging and flowering date on each sowing date. And you can see that the late sowing date
really got hammered by the lodging event, but that’s the same lodging event on the
early sowing date, if you know what I mean. So the same – we’ve overlapped the data
a little bit to relate it to flowering, so that’s actually the same lodging event.
So what happens is the late sowing date – if you got a massive lodging event, the late
sowing date is naturally going to cop it earlier in its life cycle. So if – it’s an important
thing for us to understand if we’re going to sow late and the crops are going to grow
further into our storm season, there’s potentially a greater likelihood that those late sown
crops are going to catch a big storm and lodge at an earlier growth stage. So that’s really
important for us to think about, and we may have to go and do probably a climate analysis
to look at how many big rainfall events; what’s the frequency of big rainfall events through
spring, and is that probability changing through September and October, or is it the same?
So it would be interesting to know. So at Spring Ridge we got some interactions
between variety and N regime and sowing date. They were inconsistent – on the early sowing
date we saw the opposite trend to what we often see, where the sowing N for some reason
did a little tiny bit better. But on the late sowing date we saw what we usually see, and
the delayed N strategy really helped us avoid lodging a little bit, and yield better. And
bear in mind that most of the varieties we’re trialling were actually better suited to that
late sowing date. So they were mainly the quicker varieties. And so this was the trend that was probably
more applicable to varietal maturity. And if I had the time, I would really go through
and pick out those results, and I probably should compare the long season varieties sown
on the early sowing date, to the late season variety sown in the late sowing date. That
would be the most logical comparison to make. Don’t get too hung up on this graph. I like
to do the statistically correct thing and show you the individual results, because there
are interactions going on, but there’s still a fairly consistent picture going on with
certain varieties performing poorly, and certain varieties like Trojan and Cobra performing
very well. Some of these – there’s some big interactions going on here with this thing,
Impala, which really seems to respond sometimes to this delayed nitrogen regime. So the sowing N on the late sowing date did
a lot worse than the delayed N on the sowing date. But we didn’t get the same response
in the early sowing date. So what we really want to do in the next couple of years is
cut down the set of varieties to the ones that are the really outstanding ones, and
start looking at them in more detail, so we have a bit of picture of exactly why we’re
seeing some of these responses. This is the lodging for those varieties. This
is lodging across the first four weeks of grain filling. So we actually cut out the
last two weeks of data for this particular graph, because the late lodging event came
in and really it was very indiscriminate, and the error went up substantially if we
included the late lodging data. So this was actually a better picture of what was – this
data was better correlated to yield, and I should have put a different graph up to show
that. So we saw again Trojan and Cobra did well.
Lancer did reasonably well, but it lodged at Emerald. I’ll just go back to Emerald,
and just point out some of these things. So we have Lancer, Mitch and Wallup were three
varieties that came out really well out of the first couple of years of testing, but
we can see Lancer lodging at Emerald; Wallup lodging and Mitch lodging. So we saw very
very little lodging in those varieties in the first two years of the project. And we have seen lodging in these varieties
last year, to a lesser extent in Narrabri: Mitch, Wallup, Lancer. The other thing that surprised us last year
was that Suntop, we didn’t get much of a look at Suntop in 2013. And there was one
trial where we got late lodging, and Suntop was one of the intermediate to worst lodged
varieties in that particular trial. But it was a late lodging event, and Suntop did yield
well. So we had a hard time figuring out what preliminary rating to give Suntop. Last year
we trialled Suntop and it’s done – it has actually been one of the least lodged varieties
in all of our trials, and so that’s good news I guess, as long as that’s going to
be a consistent response. I’m just going to flip past that. So in summary,
this is the table – it’s in the book – of the percentage difference to the trial mean.
So Cobra at Emerald on the early sowing date was 16% above the trial mean yield. So the white boxes are where they’re more
than 10% above, and the grey boxes are where they’re more than 5% above – or between 5%
and 10%. So you can see that Cobra and Trojan did really
well consistently across locations. One little blip there for Trojan, but in general they
were very very consistently high performing for yield across locations. So it’s just our
first year that we’ve looked at them. Bear that in mind. Cobra’s default quality classification
in the north is APW, and just be aware that it is MS to S for stripe rust – both stripe
rust pathotypes. So if someone was going to have a go at it, you’ve really got to keep
an eye on the disease management. Trojan is only an ASW default in the north,
and it’s MS to S for yellow spot. Yellow spot can be pretty bad in a lodged paddock,
because lodging brings the crop closer to the ground, and yellow spot spreads through
splash dispersion of spores I understand. So we saw some bad yellow spot in one of our
trials last year – in one of the Gatton trials, which I haven’t shown here. On the off chance anyone wants to rush out
and put in a paddock of Trojan and Cobra, you might want to know the maturity of them,
because there’s probably not a lot of data out there. And so at Emerald and Gatton they
were both falling as a mid-season type. They were sort of halfway between Kennedy and Gregory
with their flowering response. At Spring Ridge they tended to be quick maturing. They were
similar to Merinda and Crusader. But at Narrabri, they were inconsistent, and I don’t have
time to go into the specific responses. If someone wanted to know about how they went
at Narrabri I could get you that data later. So just be aware varieties will interact with
a year in terms of their flowering response. So this is just the first year we’ve seen
these things. We build up a picture over time. We’ll get an idea of how quickly they’re
going to flower. So the next three most consistent varieties
were Mitch and Sentinel and Suntop. So they also performed quite well for yield, but not
quite at the same level as Trojan and Cobra. But out of those three, Suntop is the only
APH variety. But bear in mind, you know, Suntop is the – we’ve only really had one really
good year’s worth of data on Suntop. But given that 2013 was a really low lodging year, we’ve
got to wonder how good our data was in 2013 where we didn’t get much lodging and the
lodging that we got was really late. And I’ve started to do some analyses which
show that that late lodging is not nearly as well correlated to yield as the lodging
that happens earlier in grain filling. So we’re not sure how good that data from 2013
is actually going to be in the end. So this is just a summary of the rating that
the project has given all of those varieties based on the three years’ worth of data. It
will be a little bit different, particularly to the New South Wales variety guide, because
Peter Matthews has got data from the MIA, which is a bit different to ours, because
these varieties have different flowering responses and that’s probably related to different
lodging responses. So our aim over the next – yes, so I guess
I just go back and say again, what we call G by E and by M: genotype/environment interaction
and also management interaction, can always through a surprise. So we would love our varieties
to perform the same every year, but you just can’t guarantee what’s going to happen
if you get a completely different year. If we get a year that’s right up – you know,
a super cool year all season, what’s going to happen? Which variety is actually going
to stand up the best? Now that we’re collecting a bit of data, we’re going to start doing
some multiyear analyses and see just how bad that G by E by M is. I still feel that our
varieties are going to perform relatively consistently for lodging, I hope, but we’ll
wait until the analyses come out before we make that call. So, yes, over the next couple of years we’re
going to try and narrow down our varieties. So we’re going to really start looking at
these MR varieties and see if we can tease apart the differences between them for lodging
and yield potential. So by the end of the project maybe some of these we might be able
to move into the RMR category, but some of them might have to move this way as well.
So there’s a good solid base of varieties that are around about that MR mark that we’ve
got to work with that are certainly a lot better than the varieties we were working
with a few years ago that are quite lodging susceptible, although that might be a bit
harsh on Gregory, given its performance last season. So we’ll see how we go. But I just remind people, varieties aren’t
a silver bullet when it comes to irrigated wheat production and reducing lodging risk.
An MR variety is a good start, but that’s all it is. You still want to use a sensible
seed rate: 100 plants per metre squared on the bed area. And I keep talking about bed
area, because if you’re getting a contractor in to sow a furrow irrigated paddock, and
you say “sow 50 kilos of seed to the hectare to give me that 100 plants per metre squared”,
well, if he sows 50 kilos to the hectare and he’s not sowing anything in the furrow,
what’s he putting on top of the bed? Which is 50% of the area. He’s putting 100 kilos
of seed on the bed or he’s doubling the plant population. So it’s just something to
keep in mind if you’re furrow irrigated farmer. Soil condition, unfortunately, plays a big
role in terms of germination percentage and establishment. So you’ve still got to know
your soil and know what your plant is going to do, but you want to end up with around,
even between 50 and 100 plants per metre squared on the bed is pretty reasonable if you’ve
got a little bit of nitrogen in your soil. Test for soil N. Understand what your soil
N is before you go into an irrigated wheat crop. Are you dealing with a really high N
soil? It’s going to be a big lodging risk. And there are ways to manage that, but you
want to know. If you can, ideally you just want to start with between 50 and 100 units
of N at sowing, and dribble the rest of the N on through the season. Plant growth regulators
can help, particularly in a higher N paddock. So if you’ve got more than 100 units of N
at sowing, and you’re trying to get more than six tonnes of wheat yield, then I would
be recommending that you use plant growth regulators. So next steps: I’ve already talked a bit
about this as I’ve gone through the presentation, but we’re going to look harder at those
MR varieties. We’re going to start looking at the best management for some of those selected
varieties. We really want to do an analysis for the agronomy conference later this year
on just what is the relationship between grain fill lodging and yield, and compare that to
final lodging score and yield. Because that final lodging score, I suspect, is not – certainly
the preliminary analyses is if you just go out and look at lodging at harvest time, that
data is not going to be very well correlated with yield. It’s what happened through grain
filling. How early did that crop start to lodge. That’s the big issue we’re finding.
And we want to keep looking at the effects of seasonal trends on lodging risk. Throw a curveball into the equation: we’ve
talked about lodging; we’ve talked about high yield; which varieties are best. But
for someone who water is limiting and you’ve got lots of land, do you really want to be
trying to get the maximum yield per unit of land? And I’ve been doing some economic analysis
and the answer is probably no. You probably don’t. If you’ve got lots of land and a
bit of water, you’re better off spreading that water over a slightly wider area and
aiming to try and get five or six tonne to the hectare on that area. And I’ve got some
pretty extensive analysis on that that people can come and get off me if they want. And
it’s a fair bit involved. So, again, we had a different year in 2014.
We got some different results. Mitch and Wallup and Lancer lodged more than we had seen in
previous trials. Just be aware that they can lodge; they’re not a silver bullet. They’re
still fairly lodging resistant. Suntop did better than we expected so it might be a good
option in the long run in the north. And Cobra and Trojan had the highest yields and relatively
low lodging. So a very promising start for those two under irrigated conditions in the
north. Again I acknowledge that the research was
done with the support of GRDC and growers. And just I always like to throw in a few pictures
at the end. You know, this is why we don’t grow Sunvale, and Baxter and Burke. There’s some very lodging susceptible varieties
out there. So varietal choice: picking an MR to start with is a good start when you’re
growing an irrigated wheat paddock.

Leave a Comment

Your email address will not be published. Required fields are marked *