Is there a preferred wheat or barley variety to
grow in a P deficient soil?
Sean Mason1, Glenn McDonald1, Bill Bovill2, Willie Shoobridge3, Rob Wheeler3
1 School of Agriculture, Food and Wine University of Adelaide; 2 CSIRO Land and Water, Canberra; 3
SARDI New Varieties Agronomy
Why was the trial undertaken?
The imperative for efficient use of P in broad acre agriculture is an increasing issue
due to the likelihood of increased fertiliser prices contributing to greater
production costs in the future. Wheat and barley varieties may vary in their responsiveness to P either
by having root traits that increase access to soil P or by more efficient use of the P that is taken up. In
combination with different yield potentials, external P requirements and phosphorus use efficiency
(PUE) could vary. Previous experiments conducted at Minnipa and Mallala in 2012 and 2013 revealed
small significant responses to P applications among various wheat and barley varieties however no
significant differences could be obtained for PUE, potentially due to the relatively small yield response
obtained. Trials were repeated in 2014 at Condowie and Sherwood where very low P levels were
measured, in an attempt to generate greater yield responses to P and identify if there are any
significant differences in PUE between varieties.
How was the trial done?
Two replicated field trials using wheat and barley were established at Sherwood in the SE Mallee and
Condowie in the mid-North of S.A. Both sites were at similar low levels of available P as measured by
either DGT P or Colwell P compared to their respective critical values. The two sites did have
contrasting phosphorus buffering index (PBI) values with the heavier soil type at Condowie (neutral
pH) approximately double the PBI value of the sandy loam acidic soil type at Sherwood. On 23 May
(Sherwood) and 3 June 2014 (Condowie), 6 varieties each of wheat and barley were sown at 5 rates
of P: 0, 5, 10, 25 and 40 kg P/ha as triple superphosphate. The varieties sown were selected from a
range of current commercial varieties and some old varieties that have been reported to show
differences in P responses. The P use efficiency (PUE) is defined as the yield at 0 P relative to the
maximum yield. The P requirement was assessed by fitting a curve through the yield response data
and the required P rate was estimated as the rate that gave 90% of the overall yield response.
Key Messages
• Yield responses to P were associated with promotion of early crop biomass in both wheat and
barley. Compared to N, there appears to be less risk of high P rates adversely affecting yields.
• Compared to differences in yield among varieties, differences in responses to P have been
small. At this stage variety selection should be based on potential yield rather than any
differences in PUE to achieve the greatest return in investment from P.
• Phosphorus nutrition levels should be continually monitored especially those on replacement
P programs and soil types with moderate to high PBI levels. Required P rates for both trials
were at least double the calculated replacement P rates.
• More efficient replacement P rates could be obtained if they are adjusted in accordance with
PBI levels if they vary significantly within a paddock.
Page | 1
Background
Aim: To investigate responses to phosphorus (P) fertiliser of common wheat and barley varieties on P
deficient soils.
Maximising yields on the basis of providing adequate P nutrition can be achieved by applying sufficient
amounts of P fertiliser to soils where P is limited. Fertiliser applied to the crop contributes only 5-30%
to the crops total P uptake and therefore the rest of the crop’s P requirements needs to be supplied
from existing soil P reserves. Identifying varieties that have greater PUE in a P deficient soil is of great
interest to farmers in the MSF region especially those on soils that have an ability to fix applied P as
indicated by higher PBI values. As mentioned in the one page summary previous experiments
conducted at MAC and Mallala in 2012 and 2013 revealed small to moderate significant responses to
P applications among various wheat and barley varieties however no significant differences could be
obtained for PUE potentially due to the relatively small yield response obtained (EPFS 2013 pg 129131). Optimal P rates on these soil types were also difficult to obtain due to linear responses to P
applications up to 25 kg p/ha especially at Mallala where PBI values were close to 100. Trials were
repeated in 2014 at Condowie and Sherwood with the aim of obtaining greater overall responses of
wheat and barley to applications of P and to improve the determination of optimal P levels by
increasing P rates. The hope was that this would provide the best chance of identifying any significant
differences in PUE between common wheat and barley varieties.
About the trial
The P use efficiency (PUE) is defined as the grain yield at 0 P relative to the maximum grain yield
obtained, which is calculated at the plateau of the response curve which is fitted through the yield
response data. The P requirement was estimated as the rate that gave 90% of the overall yield
response to P application. The economics of returns from obtained yield vs cost of applied P was
calculated based on prices of $280/t for APW wheat and $270/t for Malt barley, and a fertilizer price
of $750 (DAP) (PIRSA Gross margin guide 2015).
Soil test results:
Table 1. Mean and spatial variation in available P values for each trial location. Ten cores were taken
in 10 plots of each trial and measured separately. DGT P presented as
-g P/ L, critical v
56, 95% CI), Colwell P and critical Colwell P in mg P/kg. Data are shown as mean ± standard error of
the mean.
DGT P
Wheat
CV(%)
Barley
CV(%)
Page | 2
14 ±
2.0
40
18±
2.0
31
Sherwood
Colwell
Critical Colwell
PBI
P
P
54±
15± 1.0
22
8.0
27
43
42±
19± 2.0
20
5.0
37
33
DGT P
17±
2.0
37
17±
1.0
39
Condowie
Colwell
Critical Colwell
PBI
P
P
97±
22± 2.0
28
3.0
24
8
85±
17± 1.0
26
1.0
22
10
Results
Yield results:
Significant responses to P applications and significant differences among varieties were obtained for
grain yield in both wheat and barley at both sites (Table 2). Despite overall larger responses to P
compared to the 2012 and 2013 seasons there was no significant Variety x P interaction in either
wheat or barley at either site. In other words, for both wheat and barley the yield differences among
the 6 varieties were too small to pick up significant differences in their responsiveness to P. Barley
varieties tended to yield higher than wheat especially at Condowie which in part can be attributed to
the occurrence of yellow leaf spot at early development for susceptible wheat varieties (Scout, Correll)
as the trial was sown into wheat stubble.
Table 2. Mean yields across all P rates for each variety at each field site
Sherwood
Condowie
Sherwood
Variety
GY (kg/ha)
GY (kg/ha)
Variety
GY (kg/ha)
Barley
Wheat
Barque73
3255
2962
Correll
2705
Commander
3151
2962
Gladius
2647
Fleet
3471
2939
Mace
3019
Galleon
2545
2816
RAC875
2756
Hindmarsh
3312
2853
Scout
2583
Yarra
2587
2617
Wyalkatchem
2827
LSD (P=0.05)
CV%
451
8
254
5
LSD (P=0.05)
CV%
380
8
Condowie
GY (kg/ha)
2386
2294
2341
2344
1802
2296
359
9
Comparison of PUE % and optimal P rates between wheat varieties reveals that higher efficiency (>
PUE %) didn’t necessarily result in lower external P rates (Table 3). Between sites there was some
correlation between varieties requiring lower P rates at Sherwood and lower optimal P rates actually
being able to be calculated at Condowie. The linear nature of response at Condowie for all barley
varieties doesn’t allow for comparison between sites.
Page | 3
Table 3. Response of each variety to applications of P expressed as PUE% and the corresponding
required P rate to reach 90% of maximum yield. *Maximum yield not obtained
Sherwood
Condowie
Crop
Variety
PUE %
Required P (kg/ha)
PUE %
Required P (kg/ha)
Wheat
Correll
66
18
74*
> 48
Gladius
58
14
82
39
Mace
57
13
73
44
RAC875
60
19
80*
> 48
Scout
49
26
70*
> 48
Wyalkatchem
42
12
82
43
Overall
62
20
71*
> 48
Barley
Barque73
74
23
82*
> 48
Commander
57
16
88*
> 48
Fleet
59*
> 48
92*
> 48
Galleon
50*
> 48
79*
> 48
Hindmarsh
57
21
81*
> 48
Yarra
49
29
79*
> 48
Overall
59
29
83*
> 48
P/ha) or greater, the relatively flat linear response meant that the yields obtained in 2014 at these
higher P rates (> 25 kg P/ha) were not necessarily the most economical with current grain and fertiliser
prices (Table 4). At Sherwood economic benefits were obtained above typical replacement rates due
to the higher relative efficiency of P applications.
Table 4. Economic analysis performed for the two P response trials based purely on fertiliser cost and
yields obtained. Prices used can be found in the text. Economic optimal P rates for each category are
highlighted in bold.
P treatment (kg/ha)
Condowie
0
5
10
25
40
Sherwood
0
5
10
25
40
Page | 4
Fertiliser cost
$/ha
Returns from yield ($/ha)
wheat
barley
Net return ($/ha)
wheat
barley
0
19
38
94
150
542
607
597
681
715
720
735
747
793
863
542
588
560
587
565
720
717
709
699
713
0
19
38
94
150
561
717
794
894
907
596
755
817
951
1003
561
698
756
801
757
596
736
779
857
853
Implications for commercial practice
Yield responses to P were associated with promotion of early crop biomass in both wheat and barley
(data not shown). Compared to N, there appears to be less risk of high P rates adversely affecting
yields in a growing season typical of 2014.
Compared to differences in yield among varieties, differences in responses to P have been small. At
this stage variety selection should be based on potential yield rather than any differences in PUE to
achieve the greatest return in investment from P.
Phosphorus nutrition levels should be continually monitored especially those on replacement P
programs and soil types with moderate to high PBI levels. Unless the relative inefficiency of P
applications and the capacity of some soils to fix P have been considered, replacement P inputs on
these soil types could be driving down P levels.
At similar starting P levels for both sites much higher fertiliser efficiency was evident at Sherwood
which could be explained by the much lower PBI value (reduced fixation). Therefore more efficient
replacement P rates could be obtained if they are adjusted in accordance with PBI levels if they vary
significantly within a paddock. We encourage the continued use of farmer strip trials (leave a strip of
nil P fertiliser) in combination of with Colwell P and DGT results for on farm validation of the soil tests.
For paddocks with moderate to high PBI levels significant information could be obtained by
incorporation of a P rich strip (e.g. 30 kg P/ha) next to the standard rate (10 kg P/ha) to ensure P
deficiency is not masked by relative low fertiliser efficiency.
Acknowledgements
The experiments were run with the financial support of SAGIT (project code – UA1201). The trial was
managed by Rob Wheeler and the expertise of his team
is acknowledged.
Links and references
Eyre Peninsula Farming Systems Summary 2013 Efficiency of wheat and barley varieties in a P deficient
soil. Pages 129-131.
Page | 5