Take home message

• Soil testing in the correct depth increments is important in order to separate the surface (0-10cm) and sub-surface (10-30cm) nutrient status in order to identify the magnitude of nutrient stratification. This will determine yield response to deep applied nutrients.
• Yield responses to deep applied phosphorus (P) in Central Queensland (CQ) have ranged between an extra 14-100% of the treatment with no deep P
• Yield responses to deep applied potassium (K) in Central Queensland (CQ) have ranged from an extra 8-29% of the treatment with no deep K
• Responses to deep banding P and K in the sub-surface layer can be influenced by; (1) nutrient status of surface layers, (2) in-crop rainfall, (3) crop species, (4) nitrogen status

Introduction

Over the past five years the UQ00063 project has been monitoring a series of nutrition-based trials sites across Central Queensland (CQ). These trial sites were chosen based on soil test evidence showing varying degrees of nutrient depletion in the surface (0-10cm) and sub-surface (10-30cm) layers.

This is particularly evident for the non-mobile nutrients of phosphorus (P) and potassium (K). In some established zero tillage production systems there is a marked difference between the nutrient concentration in the top 10 cm of the soil profile and the deeper layers (10-30 cm and 30-60 cm) that cannot be explained by natural stratification. It would seem that this pattern of soil analysis is becoming more evident across CQ, particularly in the Brigalow scrub and open downs soil types.

This project aims to find whether a one-off application of P and/or K, placed in the sub-surface layer (10-30cm) can provide a grain yield benefit and whether that benefit can be maintained over several years.

Experimental outline

Eight trial sites were initially established with deep P trials, and six of those sites also had deep K trials co-located in the same field (i.e. both P and K were marginal to low). All trials followed a similar protocol with eight common treatments. The four main treatments were based around increasing rates of P or K. For the P trials; rates of 0, 10, 20 and 40 kg of P/ha were used. The K trials included 0, 25, 50 and 100 kg of K/ha. In all treatments, background fertiliser was applied to negate any other potentially limiting nutrients. This background fertiliser included; 80 kg of nitrogen (N), 50 kg of K (if it was a P trial), 20 kg of P (if it was a K trial), 20 kg of sulfur (S) and 1 kg of zinc (Zn).

These main rate treatments were accompanied by two other treatments that included the highest and lowest rate of nutrient without the background P or K (i.e. P trials had no background K or S, and vice versa), although background N and Zn were retained. These allowed us to assess the interacting effects of P or K if the other nutrient was also limiting (e.g. would we see a K response if we did not first fix low P). A final two treatments included farmer reference (FR) plots, and an extra zero plot to give two controls for each replicate. The FR treatments had nothing additional applied during the initial application of treatments and were not deep ripped, thus representing normal commercial practice of the current farming management (Table 1 and Table 2).

Treatments were applied using a fixed tine implement that delivered the P and K bands at a depth of 20 cm and the Nitrogen (N) and Sulphur (S) bands at a depth of 10–15 cm. The bands of fertiliser were placed 50 cm apart in plots that were six to eight meters wide and 28-32 m long. The bands of fertiliser were placed in the same direction as the old stubble rows. Most trials had six replicates. Commercially available fertiliser products were used in all treatments (Table 3).

On some sites the effects of an additional starter fertiliser treatment applied at planting were also evaluated, with the main plots split into ‘with’ and ‘without’ starter. Starter fertilizer was applied at a general rate of ~30 kg/ha, either as MAP or Starter Z.

Table 1. Summary of nutrient application rates for P trials

Table 2. Summary of nutrient application rates for K trials.

Table 3. List of commercial granular products used in nutrient treatments

Results

From the eight trial sites established with this methodology, six proved to be responsive to deep P application. Of the six deep K trials that were co-located with the P trials, four proved to be responsive. The size of the responses to both deep P and K have been variable across seasons and crop species.

There have been 13 crops grown across the six responsive deep P sites and nine of these have shown significant differences at the P<0.001 level (Table 4). For the responsive K trials there have been nine crops grown of which seven have shown significant differences.

Table 4. Summary of significant site responses to deep P and K across crops and seasons.
Maximum yield response achieved (as a % of the ‘0P’ or ‘0K’ treatments)

Responses to deep banded P have in general been, consistent with soil test values (Table 5), with large grain yield responses of between 14% and 100% recorded (Table 4).

The K responses have been smaller, with yield responses ranging from 8% to 29% and these responses have not always followed the predicted response based on soil test values (Table 5). For example, the Comet River site has a low exchangeable K reading in the sub-surface layer (10-30cm) but so far has not responded to deep K in chickpeas or wheat, while the Clermont site had a higher exchangeable K status in the same layer but did respond to deep K in sorghum.

At least part of the reason for the uncertainty around K responsiveness relates to seasonal conditions and the amount of K that can be utilised out of the topsoil when it is wet enough. It also reflects that the soils are not as depleted in K as they are in P at this stage of the crop cycle. There may also be slow release K minerals present at some sites which the exchangeable K test does not measure.

Table 5. Soil test values for P and K at all responsive trial sites.

When assessing soil test results to identify possible nutrient constraints, the following tables (Table 6 and Table 7) are a useful guide. These figures are based on extrapolation of trial data that has been accumulated across all Queensland nutrition sites over the last five years. It is also worth noting that the critical values for K may increase somewhat in soils with a high cation exchange capacity (i.e. >40 cmol/kg)

Table 6. Current recommendations for assessing soil P status from soil test analysis

Table 7. Current recommendations for assessing soil K status from soil test analysis

Interactions with deep P and K responses

Consistency in response to deep P and K is dependent on a number of factors. Some of these factors relate to the concept of nutrient stratification, which is driven by the lack of mobility of both these nutrients in our clay soils. Stratification occurs when there is a build-up of P and K in the surface layers from either starter fertiliser application (P) or the constant breakdown and release of nutrients including P and K out of stubble. While enrichment of the surface layers by these two processes is occurring, there is a depletion of these non-mobile nutrients out of the sub-surface layers from normal plant uptake, particularly during dry seasons. When roots are living on stored soil moisture, they are also accessing subsoil nutrients to be able to grow successfully.

A number of the P responsive sites in CQ have adequate levels of P in the surface layer (0-10cm) but that P status drops off significantly in the sub-surface layer (Dululu, Springsure Creek and Comet River). This characteristic can affect the seasonal response to deep placement of nutrients. When adequate in-crop rainfall is received then the surface layers hold moisture and allow the plant to access those surface nutrients. This reduces the reliance of the plant on being able to forage for water and nutrients deeper in the profile.

When the crop has little in-crop rainfall, which is quite common in CQ winter crops, the surface layer dries out and the plant is unable to access those non-mobile nutrients trapped in the dry surface soil. This is why quite often the strongest response to deep banded P and K is in dry seasonal conditions.

Timing of rainfall in relation to crop stage can also have an influence on response to deep P and K. Different crop species set components of yield at different stages in the crop cycle. For example, grain number in wheat is set during the first six weeks after planting, and the number of tillers (and hence grain heads/ha) are determined shortly thereafter, so this is a critical period for crop P availability. Conversely, a crop like chickpeas sets its yield from the start of flowering (60 days after sowing) and the next 30 days following – considerably later in the season. If surface soils are dry during these critical periods, the plant has to access nutrients in the sub-surface layers, and it is in those seasons when deep placement of P and K have the largest impact.

This concept is best exemplified by data collected at the Dululu site, where there was no significant response from the deep placed nutrient treatments in wheat in 2016 but there was a significant response to both P and K in the chickpea crop in 2017 (Table 4). Rainfall data (Figure 1) would suggest the chickpea crop experienced a very dry surface soil profile at the start of flowering and over the following 40 days, meaning reliance on deeper soil layers for P and K was likely to be high. In contrast, the previous wheat crop had a number of significant in-crop rainfall events that would have ensured that the crop had access to surface soil nutrients in moist soil for extended periods during the season.

Figure 1. Comparison of accumulated in-crop rainfall for the 2016 and 2017 seasons at the Dululu site.
Wheat grown at the site in 2016 and chickpeas grown in 2017.

Another plant characteristic that would have influenced the result at the Dululu site would be root structure. Cereal crops such as wheat have an adventitious root system, which has a high proportion of fine roots (high surface area for absorption) and generally a large root mass spread throughout the profile (like a spider web). This makes wheat a very good forager for nutrients, as it has more opportunities to exploit more of the soil containing non-mobile nutrients. Therefore, in soils with constrained P and K levels, wheat has an advantage with its root system in that it can minimise the impact of low nutrient status better than a chickpea crop can.

In contrast, grain legumes such as chickpeas tend to have a tap root system with fine lateral roots developing from the main tap root. This means generally a lower total root mass and a lower total surface area for the root system, so a root system that is potentially less able to exploit a given soil volume. Grain legumes tend to overcome this constraint by being very dependent on symbiotic relationships with organisms like mycorrhizae (VAM) to assist in acquiring nutrients.

There is no doubt that based on the trial data collected over the last 5 years, chickpeas have had a much stronger response to deep applied P and K than sorghum or wheat in the CQ region. While grain legumes such as chickpeas do have higher grain nutrient concentrations, cereals are also vulnerable to low soil nitrogen availability – something that does not typically impact on grain legumes. General soil test values across a number of CQ sites have suggested that low N status may be limiting cereal crop yields in a number of paddocks (Figure 2).

Figure 2. Summary of starting nitrates across several CQ sites.

This has led to low protein levels (7-8% in sorghum and wheat) in some trials where the responses to P and K have been limited. Chickpea crops grown after cereal crops have shown much larger responses to the deep P and K than in the previous cereal crop.  Part of this response is thought to be because there is no N limitation for a grain legume. Data collected from the Dysart site has shown good evidence of this issue (Figure 3).

Figure 3. Yield response to deep P over 5 consecutive crops in relation to zero P treatment
at Dysart trial site.

The difference between the 2015-16 sorghum and the 2017 chickpea crop (Figure 3) in terms of relative yields is dramatic (15% to 100%). It is possible that some of this yield increase may be due to chickpeas being not limited by nitrogen levels where as the sorghum crops can be.

The 2018 sorghum crop following on from the 2017 chickpea crop has achieved the highest relative yields of all the sorghum crops grown at this site since the initial application of deep P and K treatments. The main difference between the 2018 sorghum and the previous three sorghum crops has been a doubling of the urea application in the fallow prior to planting. The first three sorghum crops received 100 kg/ha of urea where as the 2018 crop received 200 kg/ha. While protein analysis is yet to be completed for the 2018 sorghum crop, the difference in yields alone from crop to crop are strong evidence of the importance of N status in realising the full yield benefit from P and K application in cereal crops.

Summary

In general, soil test values have predicted responses to deep placed P and K in 80% of sites tested in CQ using depth increments that ensure separation of the surface and sub-surface layers. While the scale of these responses has been variable, the P response have been reliable.

There are a number of factors that will influence the size of the response from deep placed P and/or K. These factors include:

• Concentration of nutrient in the top 10cm
• Amount and timing of in-crop rainfall
• Crop species
• Nitrogen status

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the author would like to thank them for their continued support.

Contact details

Doug Sands
Department of Agriculture and Fisheries
LMB 6, Emerald QLD 4720
Ph: 07 4991 0811
Email: douglas.sands@daf.qld.gov.au

® Registered trademark
™ Trademark

GRDC code: UQ00063 Regional soil testing guidelines