Investigating western farming systems

Take home messages

• Winter cereals showed improved water use efficiency (WUE) compared to legumes and summer grown crops at both Narrabri and Mungindi
• Generally, dollars per mm of crop water used was greater for winter crops ($2.20) than summer grown crops ($1.30)
• Although summer crops had lower WUE than winter crops there is a benefit in growing summer crops to manage root lesion nematodes
• Crop choice influenced the fallow efficiency (FE). The median fallow efficiency after winter cereals equalled 0.26, whereas fallow efficiency after chickpeas equalled 0.14
• Modelling suggests a high intensity cropping rotation can be the most profitable, if crops are planted on 100mm or more plant available water (PAW).

Introduction

While advances in agronomy and the performance of individual crops have helped grain growers to maintain their profitability, current farming systems are underperforming; with only 30% of the crop sequences in the northern grains region achieving 75% of their water limited yield potential.

Growers face challenges from declining soil fertility, increasing herbicide resistance, and increasing soil-borne pathogens in their farming systems. Change is needed to meet these challenges and to maintain farming system productivity and profitability. Consequently, Queensland Department of Agriculture and Fisheries (DAF), New South Wales Department of Primary Industries (NSW DPI) and CSIRO are collaborating to conduct an extensive field-based farming systems research program, focused on better use of the available rainfall to increase productivity and profitability, with the question;

“Can systems performance be improved by modifying farming systems in the northern region?”

In 2014 research began in consultation with local growers and agronomists to identify the key limitations, consequences and economic drivers of farming systems in the northern region; to assess farming systems and crop sequences that can meet the emerging challenges; and to develop the systems with the most potential for use across the northern region.

Experiments were established at seven locations; with a large factorial experiment managed by CSIRO at Pampas near Toowoomba, and locally relevant systems being studied at six regional centres by DAF and the DPI NSW (Emerald, Billa Billa, Mungindi, Spring Ridge, Narrabri and Trangie (red & grey soils)).

Table 1.Systems implemented at each of the locations. Trangie has systems applied on a red and grey soils. Pampas includes summer dominant, winter only and mixed opportunity cropping systems. Pampas also includes combinations (i.e. higher legume + diversity) not listed here.

Farming system description

1. Baseline is typical of local zero tillage farming systems with approximately 1 crop per year grown using moderate planting moisture triggers of 60% plant available water capacity (PAWC). Crops grown in this system are limited to wheat/barley, chickpea and sorghum. These crops have nitrogen fertiliser applied to achieve 50th yield percentile as determined by the PAW prior to planting and based on APSIM yield simulations for each site.
2. Lower crop intensity reflects a conservative rotation to accumulate greater PAW for the next crop (80%). The same nutrient management as the baseline system is applied. Crops grown are also similar to the baseline, but may also include cotton as a high value crop at some sites.
3. Higher crop diversity allows a greater suite of crops to be grown to better manage disease, root lesion nematodes and herbicide resistance. Planting triggers and nutrition are the same as the baseline system. The unique rules for this system focus on managing root lesion nematodes, with 50% of the selected crops to be resistant to Pratylenchus thornei, and 1 in 4 crops resistant to Pratylenchus neglectus. To manageherbicide resistance, two crops utilising the same herbicide mode-of-action cannot follow each other. Crops grown in this system include wheat/barley, chickpea, faba bean, field pea, canola/mustard, sorghum, mungbean, maize, millet and sunflower.
4. Higher legume aims to minimise the use of nitrogen fertiliser by growing every second crop as a pulse (legume), with a preference for those that produce greater biomass and greater carry-over nitrogen benefits. Crops grown in this system are similar to the baseline (wheat/barley, chickpea, sorghum) with additional pulse options (faba bean, field pea, & mungbean). Crops will be fertilised (N) to achieve average yield potential for the PAW, with nitrogen only applied to the cereal crops.
5. Higher crop intensity aims to minimise the fallow periods within the system and potentially grow 3 crops every 2 years. Crops will be planted on lower PAW (30%) and have a greater reliance on in-crop rainfall. Crop choice is the same as the baseline system, but with mungbean added as a short double-crop option. These crops are fertilised (N) to achieve average seasonal yield potential for the PAW prior to planting.
6. Higher nutrient supply will have N and P fertiliser applied to match the fertiliser requirements of a 90^th yield percentile crop; with the risk that crops will be over fertilised in some years. This system will be planted to the same crop as the baseline each year, so that the only difference is the amount of nutrients applied.

Discussion

Crop sequence

Six systems have been implemented at the Narrabri and Mungindi farming system site (Table 1). Due to the implementation of the system rules, different cropping sequences have evolved across both sites (Figure 1). The first two years of this experiment experienced wetter than average winters at both sites. The frequent rainfall allowed for all systems to meet planting triggers, even the low intensity systems, which require 80% PAW for crop sowing. The 2016/17 summer was exceptionally hot and dry, when both sites had cotton growing. Unfortunately, Mungindi received below average rain in 2017, missing the winter planting triggers. Narrabri had average rainfall in 2017 resulting in five of the six systems meeting PAW triggers. Both sites received low rainfall during 2018, but moisture accumulation in Mungindi did allow all six systems to meet planting triggers, while Narrabri missed the winter cropping period. Scattered rain in the 2018 spring allowed the planting trigger for sorghum in the high intensity system at Narrabri and a cover crop in Lower intensity at Mungindi. Further rain over the summer enabled three more systems to reach planting triggers at Narrabri.

Figure 1. Crop sequences planted at Narrabri and Mungindi as a result of implementing
the system rules.

System productivity (2015 to 2018)

To date, four systems have produced similar accumulated grain production at Narrabri – baseline, higher nutrients, higher legume and higher intensity. While at Mungindi, baseline, lower intensity (mixed), higher legume and high nutrients accumulated similar grain yields. The results show that modified farming system rules have not improved grain production in the western environment. Interestingly the higher diversity systems at both sites, had reduced grain production compared to the baseline system (Figure 2).

Figure 2. Cumulative grain yield at the Narrabri and Mungindi farming system experiments.

System economic analysis

Over the 4 years of experiments for each system, data has been collected on the crop yields, inputs costs including fertilisers, seed, herbicides, pesticides and machinery operations. This allows the calculation of the accumulated income and gross margins for each of the cropping systems deployed at each location. Consistent prices for each commodity (10-year average adjusted for inflation) were used to avoid introducing discrepancies in the data. All grain yields were corrected to 12% moisture to account for variable harvest moistures.

Table 2. Grain pricing used in calculations based on median prices over the past ten years, less $40/t cartage costs,
for selected crops

The baseline system at both the Mungindi and Narrabri sites resulted in the highest gross margin ($1008/ha and $2480/ha respectively)(Table 2). Both sites had similar cropping sequences in the baseline system; Narrabri: wheat – chickpea – wheat, and Mungindi: wheat – chickpea – cover crop – wheat.

Table 3. System gross margin comparison for Mungindi & Narrabri (2015-2018) showing total income, costs, gross margin, return on variable costs (ROVC), system water use efficiency and the maximum cash outlay experienced between profitable crops (Variable costs before the next positive cash flow)

Determining the right crop for improving WUE

Between 2015 and 2018 there have been 10 different crops grown between the Narrabri and Mungindi sites. The crops were grown across various seasons allowing the collection of data from high rainfall seasons (winter 2016) to below average seasons (2018). Winter crops have been consistently higher yielding than spring/summer crops. Although the sites had good soil moisture at sowing, low in-crop summer rainfall and excessive temperatures have impacted yields during the project life (1.8 t/ha to 0.7 t/ha). Of the winter crops, faba beans have the highest mean yield (3.5 t/ha).  This is more than 1.4 t/ha higher than field pea and 1.5 t/ha greater than wheat.

The high productivity has meant that faba bean used the most water (589 mm) compared to 205 mm for wheat. The high conversion of moisture to grain for wheat is highlighted by the water use efficiency (WUE) of 9.98 kg/mm, almost 4 kg/mm higher than faba bean and over 5 kg/mm higher than chickpea (Table 4).

For growers looking to improve their return on the conversion of rainfall to grain productivity ($/mm return), we evaluated the gross margin of the crops and applied the gross margin to crop water use. Again for the winter grown crops, faba bean had the highest gross margin per mm of rainfall ($2.45/mm). Winter cereals ranged from $2.27/mm to $2.00/mm. Field pea’s had the lowest return for winter crops with $1.17/mm.

Of the three summer crops grown, sorghum was clearly the most efficient at grain production (4.18 kg/mm), while sunflowers and cotton had similar efficiency (2.42 kg/mm & 1.8 kg/mm). When we evaluated gross margin per mm, we found that sunflowers were the best crop for gross returns per water use ($1.37/mm), almost $1/mm greater than sorghum, which had the lowest return ($0.38/mm).

Table 4. Grain yield and water use efficiency of crops sown at the Mungindi and Narrabri farming systems sites (2015 – 2018)

Sowing moisture’s role for water use efficiency

Interestingly, planting moisture appears to have played an important role in the individual crop’s water use efficiencies (Figure 3). Increased plant available water (PAW) at sowing of wheat, chickpea and sorghum increased water use efficiency (WUE). Crops planted in the higher intensity systems (0-80mm PAW) had more crop failures and reduced grain potential than crops sown when PAW was greater than 150 mm. The exception to this rule has been the winter legumes of faba bean and field pea. Both crops had lower water use efficiencies when planted at higher PAW, most likely due to waterlogging and their sensitivity to saturated soils.

Figure 3. Sowing moisture influence on crop water use efficiency 
(all farming system sites 2015 – 2018)

Selecting the right crop for nematode management

While winter crops produced both higher grain production per mm of crop water and dollar returns per mm, summer crops played an important role in the farming systems for managing nematode numbers. This was evident at both Narrabri and Mungindi sites where a rotation of wheat – chickpea increased root lesion nematode (P.thornei) populations to levels that will impact grain productivity of future susceptible crops (Figure 4). At Narrabri the sequence increased numbers from 1.8 nem./g soil to greater than 8.5 nem./g soil, while at Mungindi the system started with 10 nematode/g soil and increased to over 19 nem./g soil. Both sites have since decreased numbers due to extended fallow periods. Where summer crops were included in the sequence (during the same period), nematode numbers stayed below 2 nem./g soil at both sites. This allows for greater crop/variety choice for future rotations, as susceptible crops won’t be as affected by the lower P.thornei numbers. This was evident at Mungindi, as P.thornei impacted the wheat yields in both the baseline and higher nutrients systems. The baseline and higher nutrient systems had wheat after a long fallow from chickpeas in 2016. Establishment was variable in these treatments, with the wheat yielding a mean of 0.8t/ha (8.5 kg/mm), whereas wheat in the low intensity system following cotton 2016/17 resulted in more even establishment and yielded 1.3 t/ha (11.4 kg/mm).

Figure 4. Pratylenchus thornei populations at (a) Narrabri and (b) Mungindi as impacted by different farming systems
(2015 to 2018)

Crop-by-crop effects on fallow efficiency

At each farming systems site, fallow water accumulation was monitored (four years of data) and used to compare how different crop types impact subsequent fallow efficiencies (FE) (Figure 5). This data shows the high variability in fallow efficiency that occurs from year to year. However, there were also some clear crop sequence effects on subsequent fallow efficiencies.

Figure 5. Summary of observed fallow efficiencies following different crops and different fallow lengths (SF – short fallows 4-8 months, LF – long fallows 9-18 months) across all farming systems sites and treatments between 2015 and 2018; winter cereals include wheat, durum and barley. Boxes indicate 50% of all observations with the line the median, and the bars indicate the 10^th and 90^th percentile of all observations. Italicised numbers indicate the number of fallows included for each crop.

Fallow efficiencies were higher following winter cereals than chickpeas. The median fallow efficiency (LF and SF) following winter cereals was 0.26, while following chickpea the median fallow efficiency was 0.14. Median fallow efficiencies following sorghum were similar to wheat (0.26), but short-winter-fallows after sorghum were more efficient than long fallows. This difference between short and long fallows was less obvious following winter cereals. This is likely due to winter fallows being more efficient than summer fallows, due to lower evaporation losses, and possibly lower soil water content at the start of the fallow. Short winter fallows for sorghum production are more efficient, while long-fallows spanning into summer are less efficient. This also explains the similar fallow efficiency of short (summer) and long fallows (summer + winter) after winter cereals.

This means that the impacts of each crop on the accumulation of soil water in the following fallow should be considered in the cropping sequence. For example, a fallow receiving 400 mm of rain after a winter cereal would accumulate 108 mm on average, while the same fallow after a grain legume would have only accumulated 56 mm. This difference is likely to have a significant impact on the opportunity to sow a crop and/or the yield and gross margin of the following crop in the farming system.

How risky is your rotation – modelled outcomes

Rotation trials are traditionally phased, so that each crop in the rotation is grown every year. This project has based planting decisions on PAW triggers, so doesn’t have fixed rotations that can be phased. Therefore, to build a greater understanding of crop sequence interactions in different environments, a series of simulations were run using the APSIM model.

Simulations

The APSIM systems framework was used to simulate crop rotations from historic climate records (1957-2017), with environmental signals used to trigger appropriate management decisions. However, these simulations only considered the dynamics of water and nutrients. Losses due to waterlogging, heat or frost shock events, disease, pests, weeds or crop nutrition other than nitrogen were not considered.

The simulations of all crop sequences were phased, so that each year of the rotation was exposed to each year of the climate record (1956-2016). All rotations were run at each of 6 sites (Table 1) to highlight the importance of matching crop choice and intensity to the environmental conditions. The selected sites represent an east-west rainfall gradient at both a northern (Pampas – Billa Billa – Mungindi) and southern latitude (Breeza – Gilgandra – Nyngan). There were only small differences between the two western sites, so this paper will focus on Mungindi (grey vertosol, APSoil No. 157, wheat PAWC – 186 mm, annual rainfall – 505 mm).

Rotations

This analysis looked at increasing crop intensity using both a fixed pattern and an opportunistic crop inclusion. A set of three base crop sequences were simulated, each with a low and a high crop intensity, with varying lengths of fallows and time in crop (Table 5). In these base rotations (high and low intensity) the crops in the sequence were sown every year (must sow crops) in a fixed pattern within their sowing window. If the sowing rule had not been met by the end of the sowing window, then the crop was sown at this time regardless.  In contrast, an opportunistic sequence was then simulated; where the opportunity crop was either sown or remained in fallow based on the volume of soil water. Crops were only sown when the volume of soil water exceeded the critical threshold during the sowing window. Simulations were also conducted with two different soil water thresholds to trigger a planting event (Base – 150 mm PAW at sowing, and Aggressive – 100 mm PAW at sowing). A failed crop is one that returns a negative gross margin (including fallow costs).

Table 5. Description of low and high intensity rotations where all crops are sown every year. An opportunistic crop rotation is where some crops are only grown when soil water exceeds a minimum threshold (shown in grey with an underline).

S = Sorghum, W = Wheat, Ch = Chickpea, Mg = Mungbean, x = 6 month fallow. Opportunity crops are underlined.

At Mungindi, the most conservative rotation of xW|xx|xCh|xx, had the least crop failures at 15% (Table 6), but also had the lowest gross margin ($152 /ha/yr) (Figure 6A, Table 6). The annual gross margin is improved by planting winter crop every year (xW|xW|xCh|xW), but increased the proportion of failed crops. The opportunity approach to increasing cropping intensity in this rotation has provided the same higher gross margin as the higher intensity annual cropping rotation, but proportion of failed crops decreased for 32% to 22% as the opportunity crop is only planted in 40% of years (Table 6).

Adding sorghum into a low intensity system (Sx|xW|xx) increased cropping intensity slightly compared to the low intensity winter rotation (0.5 vs 0.66), but increased crop failures (35%) and returned a lower gross margin ($130 /ha/yr) (Figure 6c, Table 7). This is due largely to the high failure rate of the sorghum (45%) in this rotation. Within this balanced winter/summer cropping system, cropping intensity can be increased to 4 crops in 3 years (SCh|xW|Mgx, 1.3 crops/yr) for a similar gross margin to the low intensity system, however crop failure rates increase to 45%. However, by taking an opportunistic approach to increasing intensity (SCh|xW|Mgx), crop failures are reduced to 30%. This approach returned the highest gross margin of any approach modelled here ($260/ha/yr, Table 7).

The planting triggers of 100 mm or 150 mm demonstrated that in most cases the lower planting trigger (100 mm) provided the highest gross margin with minimal increase in risk of crop failure and led to the planting of twice as many opportunity crops. This is likely due to the crops being planted earlier in the planting window, and therefore maximising seasonal yield potential. The exception to this is sorghum. Sorghum is by far the highest risk crop in this environment, so it benefited from the higher PAW (150 mm) at planting, decreasing crop failure and increasing gross margin.

Table 6. Individual crop performance for the conservative rotation at the low rainfall site of Mungindi. All crops are sown each year in the low and high intensity rotations. However, in the opportunistic rotation, wheat crops in the long fallow are only sown if the 100mm and 150mm planting triggers are reached

^†refers to the number of crops that failed relative to the percentage of crops sown

Table 7. Individual crop behaviour for different levels of cropping intensity conducted at the low rainfall site at Mungindi

^†refers to the number of crops that failed relative to the percentage of crops sown