Sowing strategies to improve the productivity of crops in low rainfall sandy soils

Introduction

Sandy soils are a dominant feature of the low rainfall southern cropping zone.  Sandy paddocks are subject to low productivity, and a key factor that affects crop emergence and early vigour is the variable soil conditions, with water repellence, low water holding capacity and low fertility being problems which farmers are trying to manage on sandy soils in this region (Unkovich et al. 2015). 

Water repellence is highly variable within paddocks, and there are 1.4 million hectares of sands affected in South Australia (SA) (e.g. an estimated 36 per cent of paddock area in the Murray Mallee display moderate to severe water repellence, Unkovich et al. 2015).  It is caused by sand grains being coated with hydrophobic waxy compounds which originate from the breakdown of crop residues. Water repellence stops soil from wetting up evenly, and even following successful wetting can again return to a repellent state upon drying. The primary consequences of water repellence are poorly established crops and consequently increased risks of soil erosion and low crop yields.  At the seed row level non-wetting sands present challenges for crop establishment due to their gradual and localised wetting patterns, leading to slow and patchy seedling emergence staggered over time and sometimes continuing up to three to four months post-seeding, depending on rainfall and seed placement.

Research on low rainfall sands has highlighted the benefits of strong early crop establishment and nutrition (Unkovich et al. 2015). Experiments in Western Australia have shown that on-row and near-row sowing (i.e. sowing as close as practical to the previous crop row without disturbing it), can increase crop establishment and biomass from sowing into zones of lower water repellence values, as well as benefiting from the intact root system of the previous crop providing a pathway for water infiltration (Roper et al. 2015, Ward et al. 2015).

How does the seeding operation influence production on sandy soils?

Optimising the seeder set-up is critical for establishing an optimum and uniform plant density regardless of the crop and season on sandy soils. Of all soil types, sandy soils present the highest risk of quickly drying out and reducing seed germination. Also, the lack of soil cohesion means furrows are typically unstable with walls slumping/collapsing and filling in under windy conditions hence reducing crop establishment, while emerging seedlings are also at risk of sand-blasting.

Seeding system technologies display wide variation in the levels of soil disturbance (e.g. disc/tine opener, row spacing, operating depth and speed), the seed and fertiliser placement configurations (single, paired or ribbon seeding, combined or split banding, with or without subseed disturbance), and the furrow closing device (seed pressing vs furrow pressing).  

Key requirements for sandy-soils include seed placement into ‘lasting’ moisture, which requires conservation of furrow moisture in the seeding operation. Specific technologies such as liquid banding can also positively influence nutrition efficacy and root disease management, while the ability to accurately track and allow precision row sowing is increasingly relevant to improving crop productivity.

Here we report on several experiments that test a range of seeding systems in sandy soils of the low rainfall Mallee region and the resulting recommendations for seeding systems.

On-row/near-row sowing

At a Moorlands (SA) non-wetting site in 2015, on-row sowing using a triple disc seeding system significantly improved CL Grenade wheat crop establishment (by up to 29 per cent) relative to inter-row sowing.  

This beneficial effect has been observed in many other trials and from paddock experiences. In this case, losses to residue hairpinning and the lack of deeper moisture delving under the disc seeding system are likely to have limited the full benefits. At this site, seed row establishment was also found best when coinciding with previous traffic lanes (whether across or along seed rows), acting as a depressed surface layer benefiting from water harvesting and possibly promoting a greater uniformity of wetting/moisture retention from the associated compaction. No relevant yield data could be acquired due to widespread frost damage suffered at the site in early spring that caused an 80 per cent loss in grain yield.

Table 1: 2015 Karoonda site: wheat establishment, biomass at first node (GS31) and anthesis (GS65) and grain yield, averaged across row position or sowing date treatments. Sowing treatment results annotated with a different letter are significantly different from another (P=0.05).

Sowing Treatment Establishment (plants/m2) GS31 biomass (t/ha) GS65 biomass (t/ha) Grain yield (t/ha)
Inter-row 19b 0.21b 1.01b 0.68
Near-row 69a 0.51a 2.73a 1.00a
April 60a 0.58a 1.95a 0.93a
May 29b 0.15b 1.80a 0.76a
At Karoonda (SA) on a water repellent sand in 2015, near-row sowing using a narrow point seeding system more than tripled Mace wheat establishment (69 vs 19 plants/m2, LSD=24 at P=0.05). Near-row sowing went to produce more than two times greater biomass at GS 31 (first node) and GS65 (anthesis) compared with inter-row sowing (Table 1). By maturity this difference amounted to a 0.32t/ha (or 30 per cent) greater mean yield, which was not statistically significant however. The lack of significant difference reflects the variability induced by the uneven effects of non-wetting soil on crop establishment and the impacts of the very dry spring conditions. 

A contrasting site at Loxton, on a sand that did not express water repellence, had a slightly higher plant density on the inter-row compared with near-row (158 vs. 146 plants/m2, LSD=11 at P=0.05) but with no significant impact on crop biomass or grain yield.

Measurements of surface (top 10cm) soil water showed near-row sowing was into a higher level of surface moisture than inter-row sowing, and secondly that the near-row position accumulated more water than the inter-row position between the two sowing dates on which the technique was tested (Figure 1). In a season when water repellence expression was rather marked this difference proved quite important. 

The near-row position had a significantly higher level of soil microbial activity, and lower root disease scoring of infection levels (data not shown; Gupta Vadakattu CSIRO) indicate that this would have had a significant effect on seedling health. As a result of improved plant numbers and seedling health, wheat biomass was significantly higher with near-row sowing and this has provided a competitive effect with brome grass, registering 28 weeds/m2 (=2022 seeds/m2) compared with 105 weeds/m2 (=7332 seeds/m2) on inter-row sown plots (LSD=48 for weeds/m2 and 2575 for seeds/m2 at P=0.05).

Figure 1: Surface soil water (mm/10cm depth) at sowing on April 27 and May 21 for plots sown near-row (near) and inter-row (inter). A column annotated with a different letter has a significantly different amount of surface soil water (P=0.05).

Figure 1: Surface soil water (mm/10cm depth) at sowing on April 27 and May 21 for plots sown near-row (near) and inter-row (inter). A column annotated with a different letter has a significantly different amount of surface soil water (P=0.05).

Seeding system responses in Mallee environments

Seedbed utilisation (SBU) is the proportion of the row spacing occupied by the crop, and can vary in practice between 10-15 per cent to 70-100 per cent depending on seeding system technology and row spacing. Increased SBU has many benefits, but is more difficult to achieve when operating under wider row spacing (30-35cm). Incentives to keep wider row spacing include reduced power requirements, lower seeder weight and cost, improved residue handling and inter-row sowing capabilities, and the ability to use higher speed without compromising the quality of seeding.  

The interest in paired-row (or split-row) systems and spreader seed boots as tools to increase SBU at existing row spacing has increased in recent years. Many paired-row system designs integrate common narrow openers such as knife blades with a split row delivery attachment to achieve two adjacent rows five to eight cm apart, in effect doubling existing SBU rating. In comparison, spreader boots aim to create a band/ribbon sowing configuration. Paired-row and spreader boot systems exist in single and double shoot options, and should be associated with a matching press-wheel width for best results in sandier soils. 

Eight double shoot seeding systems were evaluated in a replicated trial at Murrayville (Victoria) during 2014 across three soil types (stony, mid-slope and sandy rise) in a swale-dune Mallee setting. In the context of very favourable soil moisture at seeding and follow-up rainfall 80 per cent wheat crop emergence (140plants/m2) was achieved on average across soil types, with the emergence varying from 75-83 per cent range due to seeding systems and from 65-96 per cent due to soil type. Emergence peaked in the mid-slope, and was significantly lower in both the stony soil and sand hill. Emergence loss compared with the mid-slope was highest (15 per cent) on the sand hill using the district technology (Figure 2).

Conversely, two paired row systems (designed to place seeds on undisturbed side ledges) achieved the best establishment at the site and were least affected by soil type (Figure 2). Under the seasonal conditions, the paired row treatments also yielded best (five per cent above control, overall), in line with previous SBU research. These results are consistent with recent WA research findings in non-wetting sands with narrow points and paired row winged boots (Blackwell et al. 2014).

Figure 2: Crop establishment pattern for contrasting seeding systems along 12 subplots spanning over a 205m long stretch covering 3 soil types at Murrayville, 2014 (NB: District control technology = chisel tine with knife point, rubber seed boot and narrow press wheel ; Paired row systems were fitted to the same tines with suitable press-wheels).

Figure 2: Crop establishment pattern for contrasting seeding systems along 12 subplots spanning over a 205m long stretch covering 3 soil types at Murrayville, 2014 (N.B. District control technology = chisel tine with knife point, rubber seed boot and narrow press wheel; paired row systems were fitted to the same tines with suitable press-wheels).

From the commercially available designs on the market, paired row and spreader boot systems can be classified into two broad categories: 
i) Systems placing seeds on undisturbed soil, with or without a deep-till centre furrow. These systems comprise compact designs integrated behind the furrow opener, which may create extra soil disturbance but provide improved access to sub-layers of soil moisture. 
ii) Systems placing seeds in loose furrow backfill. These typically lower cost systems aim to maximise seed spread within the available furrow shape and size, and often operate further behind the opener. They are less able to establish crops in marginal moisture, more prone to residue catching issues and can have difficulties in achieving sufficient seeding depth.

At equal SBU, all paired-row/spreader boot systems should have similar potential to reduce fertiliser toxicity risk. At equal crop establishment and fertiliser use efficiency they should also have similar potential to enhance grain yield and improve crop competition against weeds. Potential disadvantages of paired row and spreader boot systems include reduced outlet size (less suited to larger or awned seeds); higher risk of blockage in sticky conditions; higher soil disturbance with potential risks of crop damage from pre-emergent herbicides; and more variable depth of seed cover across the paddock. Some of these can be mitigated by contour following technology, a reduced speed of operation, and well timed operations.

Wheat crop establishment at the Moorlands non-wetting site also showed interesting responses to seeding system technology last season; a season marked by low moisture and drying soil conditions at sowing, a rainfall deficit post-sowing (10mm rainfall in six events over four weeks), and the first significant rainfall post-sowing of 20mm being 52 days after sowing (DAS). This resulted in seedling emergence still occurring at 70-75 DAS. The ‘Mallee standard’ double shoot knife point furrow opener (followed here by twin disc sowing system) established reasonably well on a non-wetting sand at Moorlands in 2015, recording 77 per cent (149p/m2) wheat emergence (Figure 3). Wheat crop establishment significantly worsened under single shoot systems (due to fertiliser toxicity), especially when combined with lower soil disturbance. For example: 

  • single shoot, single disc system (34 per cent emergence) 
  • single shoot, shallow till seeding point (58 per cent with 13N+9P+12S fertiliser, and 82 per cent with no fertiliser)
  • double shoot, triple disc system, six versus 11kph speed (65-77 per cent emergence, best at the higher speed).
The best treatment (90 per cent emergence) was obtained by adding a shallow operating scooping share ahead of the triple disc to clear away the top three to four centimetres soil layer onto the inter-row zone and assist with placing seeds into moist soil. This scoop required low operating speed (five km/h) at 25cm row spacing. Paddock-ready ‘scoop’ solutions would need testing, but could include concepts based on modified front coulters or knife points to emphasize an effective surface soil clearing at common operating speeds. WA research findings (Blackwell et al. 2014) also showed improved crop establishment and grain yield in non-wetting sands obtained with winged narrow points, explained by an improved clearing of surface soil and control over furrow backfill. 

Research work at the University of South Australia has shown that the design and operation of narrow points has a significant impact on the movement of surface soil out of the furrow and of deeper soil into the seed zone (e.g:  Solhjou et al. 2012). 

Figure 3: Wheat crop emergence rate at 28 DAS for a contrasting set of seeding systems  (2015 Moorlands non-wetting site, error bars = ±1 standard  error of the mean).

Figure 3: Wheat crop emergence rate at 28 DAS for a contrasting set of seeding systems 
(2015 Moorlands non-wetting site, error bars = ±1 standard  error of the mean).

Table 2 summarises guidelines for seeding system selection and operation in sandy soils, based on research and field experiences to date (Further information is in related factsheet at Mallee Sustainable Farming website). 

Liquid fungicide banding options to better control Rhizoctonia

Rhizoctonia root rot disease is one of the major biological constraint to cereal crops in sandy soils in the Mallee and Eyre Peninsula,  especially in soils with non-wetting problems. Recent field research work (2010-15) conducted in sandy soils of the SA Mallee, Yorke, and Eyre peninsulas have highlighted the potential for improved controlled of Rhizoctonia using liquid fungicides. In this context, options for in-furrow banding of liquid fungicides at seeding were evaluated with disc and tine seeding systems, including 100 per cent rate applied three to four centimetres  below the seed zone, 100 per cent rate furrow surface applied behind the press-wheel, and a 50:50 split application targeting both primary and secondary root systems.

Table 2: Preferred seeder set-up and operation for particular agronomic challenges relevant to sandy-soils. 

Keys: NN not recommended at all; N avoid if possible;  Y possible under conditions; YY recommended; - -: no direct issue either way. 

Seeder\
Factors
Paddock
Limitations
Stony soils Sandy soils Rhizoctonia pressure Marginal moisture High residue Pre-em herbicide (IBS) Notes
Seeder Operation

Deep-till sowing NN Y If backfill is not diluted YY Y If backfill is not diluted - - - - For sub-seed disturbance, moisture seeking Note: tillage depth to suit seed row spacing 
High speed sowing (TINES) NN Y Y Y Y Y High speed tine sowing possible with controlled soil throw not affecting adjacent seed rows
 High speed sowing (DISCS) NN Y If clearing of top soil improves Y If disc penetration maintained Y If seed placed in moisture Y If hairpinning not an issue YY To improve incorporation Maximum speed to suit paddock conditions and seeder technology

RTK required with good seeder tracking capability and straight stubble rows
Inter-row sowing - -  N In non-wetting sand YY Y YY YY
On the row or near row sowing Y YY In non-wetting sand NN Y NN N As high risk for crop damage
Furrow Opener Narrow point N Sep. for paddock roughness YY YY Y If seeking moisture delving Y If good residue management YY Tine seeders remain the mainstream technology in the Mallee
Best practice paddock management improves the potential of disc seeders
Single disc YY Y If high soil disturbance N Esp. with low soil disturbance Y If moisture seeking capable YY N - only with best practice
Triple disc Y If reducing coulter depth YY Y With best management Y YY Y With higher disturbance
Seed Placement

Centre row banding YY YY Esp. with seeding at tillage depth YY Esp. with deep till furrows Y If seeds are placed into moisture YY Esp. if inter-row sowing YY Easiest to match with low soil disturbance and water harvesting furrows
Paired row banding or band seeding Y If compact and extra disturbance is minimised Y Esp. with seeding at tillage depth Y If coupled with best rhizoctonia management YY if seeds placed on undisturbed soil Y If good residue management YY Improved soil/seed contact on undisturbed soil base 
Side banding Y If tillage depth is minimised Y If seeds are placed into moisture Y If coupled with best rhizoctonia management YY Y If good residue management YY Improved soil/seed contract on undisturbed side ledge
N,P Fertiliser Placement

With seeds Y Subject to SBU threshold for toxicity N Unless wetting is sufficient - - NN - - - - Note: Increased fertiliser toxicity risks in low moisture environment
Deep or split banded Y If tillage depth is minimised YY YY YY - - - - Deep banding often required deep-tilled furrows
Furrow Closing

Press wheels Y If hard rubber tyres YY Esp. wide V tyre + wetting agent - - YY YY - - -> Soil to seed contact and water harvesting furrow benefits
Rotary harrows N NN - - Y Esp. following press wheels Y For spreading clumps N Unless safe to do so -> Equalising soil throw and residue clumps
-> Insulation of furrow moisture under loose cover
-> This option preferred for compaction sensitive, weak textured soils
Seed pressing + loose cover Y Subject to durability Y - - YY - - Y If control over row contamination of herbicide
Combined options with seed treatments were also evaluated. An overview of crop responses (example shown in Table 3) demonstrates the following:

  • Liquid application of fungicide Uniform® at 400mL/ha gave a large root system response, with up to 0.32-0.46t/ha potential grain yield benefit in wheat and barley respectively. The effect on the crop is well above the seed treatment and/or fungicide coated fertiliser.
  • Some practical issues (liquid banding technology, water rates, fungicide cost/ha) are limiting the on-farm adoption of project recommendations, centred around the best performing split-banding 50:50 surface and below seeds, representing a more effective strategic protection across a number of season-dependent scenarios.
  • Current data show a significant yield response to banded fungicide is most likely when conditions combine to support higher yielding crops, such as early sowing and cooler spring finish. Overall, barley and wheat treatment responses were highly correlated (r=0.95), whereby wheat averaged 65 per cent of barley yield responses.

Table 3: Fungicide management of Rhizoctonia:  Summary wheat/barley yield responses for seed treatments with Vibrance® (Vb) at mL/100 kg seed, and Uniform®  (Uf) fungicide banded in-furrow (IF) and on furrow surface (FS) at mL/ha rates. §Yield increase significantly greater than untreated based on individual site analyses, *Net yield increases significant (p < 0.05) based on META analysis of all site data.

Crop Treatment Years Trials with significant yield response p>0.05§
Trials with +ve yield responses Grain Yield (t/ha) 
Untreated Treated Net
Barley Vb seed 360 3 1 of 10 6 of 10 2.34 2.37 0.02

Vb seed 360 +Uf IF 200 3 5 of 10 9 of 10 2.34 2.53 0.18*

Uf IF 300 3 5 of 10 8 of 10 2.34 2.55 0.21*

Uf IF 400 2 5 of 6 5 of 6 2.65 2.95 0.30*

Uni IF 150 + Uf FS 150 2 3 of 6 5 of 6 2.65 2.93 0.28*

Uni IF 200 + Uf FS 200 1 2 of 3 3 of 3 3.03 3.48 0.46*
Wheat Vib seed 360 3 3 of 11 10 of 11 2.18 2.25 0.07

Vb seed 360 + Uf IF 200 3 6 of 11 9 of 11 2.18 2.31 0.13*

Uf IF 300 3 8 of 11 9 of 11 2.18 2.33 0.15*

Uf IF 400 2 4 of 6 6 of 6 2.22 2.39 0.17*

Uf IF 150 + Uf FS 150 3 5 of 6 6 of 6 2.07 2.30 0.22*

Uf IF 200 + Uf FS 200 1 3 of 3 3 of 3 2.79 3.11 0.32*
Importantly, various interactions contribute to unpredictable variability in crop yield responses in paddocks classed as high risk from season to season.  A key question remains: How can we improve the predictability of crop responses to fungicide in order to boost adoption and maximise industry impact? Further research is required to better understand the unclear interactions with season, herbicide history, nutrition and other root diseases. In particular, there are opportunities to better integrate Rhizoctonia and nutrition management for improving economic outcomes.

Recommendations

Recent research work in sandy soils demonstrates the potential for sowing strategies to significantly improve crop productivity, and suggests the following improved sowing strategies:

  • Seed placement onto undisturbed soil moisture (achievable via a number of single and double shoot seeding system technologies). Preference should be given to systems with higher seedbed utilisation via the use of paired row and spreader boot systems to reduce fertiliser toxicity as well as to increase access to moisture and maximise crop potential.
  • Targeted surface soil disturbance facilitating effective ‘furrow sowing’ in non-wetting sands to control backfill of dry surface sand onto the seedbed and secure seed placement into moisture.  Furrow sowing can also benefit from large ‘V’ press wheel profiles to create stable furrows, maximising water harvesting potential. 
  • The ability to conduct precision row sowing, either inter-row (to lower pathogen pressure when moisture is not limiting, improve herbicide safety and residue handling) or on/near-row (to improve moisture access, seedling vigour and weed competition, especially in non-wetting sands). 
  • The ability to synergise targeted nutrition, effective root disease management, and optimum application of wetting agent via the flexible use of liquid banding technology.
Further work is needed to validate whole of paddock approaches integrating the above early recommendations, and to demonstrate best-practice solutions including:

  1. Investigating differential benefits of ‘near-row’ sowing over ‘on-row’ sowing (expected from no disruption of existing stubble rows) and inter-row sowing, particularly in swale dune systems.
  2. Evaluating the impact of a lack of subseed disturbance (when placing seeds on undisturbed moisture) on root disease pressure and assess effective management solutions.
  3. Validating cost-effective integrations of liquid banding technology into reliable crop establishment systems.

Acknowledgements

Funding for this work was from the Stubble Retention initiative (GRDC project MSF00003), rhizoctonia management fungicide project (GRDC/SAGIT project DAS00125), Advanced seeding systems Mallee project (DAFF Caring for Our Country Community Landcare Grant LG-1205723-478), Agriculture, CSIRO, and University of SA. Thanks to the Loller family (Karoonda), Thomas family (Moorlands), Willersdorf family (Murrayville) and to the Bulla Burra Farm (Loxton) for their generous support in hosting the trials. Thanks to Robin Schaeffer, Matthew Whitney, Jeff Braun and Lou Flohr for discussions around trial design. The contributions from the extended research teams at Moodie Agronomy, SARDI, University of SA and CSIRO are also gratefully acknowledged.

Useful resources

Blackwell P, Hagan J, Davies S, Bakker D, Roper M, Ward P and Matthews A (2014). Smart no-till furrow sowing to optimise whole farm profit on non wetting soil. 2014 GRDC AgriBusiness Crop Update, 24-25 Sept 2014, Crown Perth, Perth WA. AgriBusiness Crop Updates 2014

Roper MM, Davies SL, Blackwell PS, Hall DJM, Bakker DM, Jongepier R, Ward PR (2015). Management options for water-repellent soils in Australian dryland agriculture. Soil Research 53: 786-806.

Solhjou, A., Fielke, J. and Desbiolles, J. (2012). Soil translocation by narrow openers with various rake angles. Biosystems Engineering 112: 65-73

Unkovich, M, McBeath T, Macdonald L, Gupta V, Llewellyn R, Hall J, Tonkin D and Baldock J (2015). Management of water repellent sands in the Southern cropping region, GRDC project CSP00195 report, CSIRO, Australia. 

Ward P, Roper MD , Jongepier R, Micin S and Davies S (2015). On-row seeding as a tool for management of water repellent sands. Procs 17th Australian Agronomy Conference, 20-24 Sept 2015, Wrest Point Conv. Centre, Hobart, TAS.

On-row seeding as a tool for management of water repellent sands 

Contact details

Jack Desbiolles
Univerity of South Australia/AMRDC, Mawson Lakes
08 8302 3946

Therese McBeath
Agriculture, CSIRO, Waite Campus
08 8303 8455

GRDC Project Code: MSF00003, DAS00125,