Key messages

• Selecting the ‘best’ suite of rotations to apply on a farm is complex. This analysis considers ~3500 different rotation phases.
• Tactically adjusting rotation choice in response to unfolding weather conditions increases farm profit by 6% for the most profitable type of weather-year and 23% for the least profitable type of weather-year.
• The most important tactical adjustment is to canola area, varying by up to 63% in response to unfolding weather conditions; but there are scores of other tactical decisions made, that in combination, also boost farm profit.
• Identify optimal rotation selections for a typical mixed enterprise farm business in the Great Southern region of Western Australia.
• Quantify the financial importance of making tactical rotation decisions as the year unfolds.
• Demonstrate the use of the Australian Farm Optimisation model (AFO).

Aims

• Identify optimal rotation selections for a typical mixed enterprise farm business in the Great Southern region of Western Australia.
• Quantify the financial importance of making tactical rotation decisions as the year unfolds.
• Demonstrate the use of the Australian Farm Optimisation model (AFO).

Introduction

Rotation decisions are a complex aspect of farm planning. There are many different land use options and sub-options that have varying effects on soil condition, pest and disease prevalence and weed seed banks, and which affect crop and animal production costs and revenues. For example, triazene-tolerant canola needs to be considered separately to a Roundup Ready canola and a spray topped pasture needs to be considered separately to a naturally senesced pasture. A further complication is that current paddock condition can be affected by the land use choice and management that was chosen up to five years previously. For example, the weed and pasture seed bank can be affected by land use choice four to five years prior (Monjardino et al 2004). Soil nutrient level and disease prevalence is affected by paddock land use over the preceding two or three years or longer (Dixon and Tilston 2010, Brooks et al 2018). Factoring in all these components quickly makes rotation planning a large and complicated task, especially with the overlay of changing input and commodity prices.

A further complication of rotation choice is Australia’s highly variable climate, which results in significant variation in production and profit (Feng et al 2022, Laurie et al 2018, Trompf et al 2014). Kingwell et al (1992) showed that weather and price variations have significant effects on optimal farm management and profitability. To handle the volatile nature of farming, farmers can alter their ‘big-picture’ strategic management to set up a more flexible business and implement short-term tactical adjustments in response to unfolding conditions (Anderson et al 2020).

Most previous farm management research has assumed that every year is the same (Bathgate et al 2009, Kopke et al 2008, Young et al 2010, Young et al 2020). When year-to-year variation has been included, management has not been optimised and frequently the tactical management options considered have been over simplified (Godfrey et al 2019, McGrath et al 2016).

In this paper we overcome previous limitations by applying an improved optimisation model – the Australian Farm Optimisation (AFO) model – which allows for a detailed representation of year-to-year variation and an extensive array of tactical management options, to identify and quantify optimal rotation management.

Method

Australian Farm Optimisation (AFO) is a whole farm linear programming model that supersedes the historically popular MIDAS model (Bathgate et al 2009, Kingwell 2011, Kingwell and Pannell 1987, Kopke et al 2008, Pannell 1996, Thamo et al 2013, Young et al 2011, Young et al 2020). The model represents the economic and biological details of a farming system, and includes modules for rotations, crops, pastures, sheep, crop residue, supplementary feeding, machinery, labour and finance. Furthermore, it includes land heterogeneity by considering enterprise rotations on any number of soil classes or land management units (LMU). AFO is designed to evaluate rotation choices. Firstly, AFO has a detailed and flexible rotation module that includes rotation phases up to 5 years in duration with 3827 cumulative rotation phase options. Secondly, AFO includes year-to-year climate variation and a large range of tactical management options including altering land use choice, altering the timing of operations, reseeding, hiring contractors, converting crops to standing fodder or hay, altering stock nutrition profile, undertaking early season crop grazing and adjusting stock sale timing. Thirdly, AFO includes powerful stock and feed budgeting modules which are an important aspect of rotation decision making in a mixed farming system. Finally, AFO leverages powerful solving algorithms that efficiently identify optimal management for a given farm system.

For more description of AFO see the model documentation: https://australian-farm-optimising-model.readthedocs.io/en/latest/index.html.

The farm system modelled by AFO is a typical farm in the Great Southern region of Western Australia comprising a mix of crop and livestock enterprises, a six-month growing season and receiving between 400 and 550mm of annual rainfall, mostly in the growing season. The farm is 2130 hectares and includes three land management units (LMU) (Table 1). The calibration of crop and pasture inputs was completed through a combination of simulation modelling and consultation with regional experts. The cropping enterprise represented is a high input/high output system typical of current practices.

Table 1. Key features of the modelled farm

¹Reported yield is on LMU 4 (best-performing areas of the farm) in a canola–cereal or pulse-cereal rotation weighted across all weather-years.

Pasture growth rates and crop yields in each rotation in each weather-year (Table 2) were generated using AusFarm simulation modelling (Moore et al 2007) and information provided by a local agronomist. Climate data (1970 to 2020) was sourced from the Kojonup weather station. Soil data for each LMU was sourced from existing data in the APSOIL database (Dalgliesh et al 2012). Other key features of the modelled farm are shown in Table 1.

Table 2. AFO Kojonup weather-years for the current climate

¹Early break (i.e., start of the growing season): before the 5th May; Medium break: between the 5th May and 25th May; Late break: after the 25th May.

²False break: pasture feed on offer reaches 500kg/ha followed by 3 weeks of no growth.

³Good spring: above the median (86mm) rainfall for September and October; Poor spring: below the median rainfall.

⁴Yield scalar is the relationship between yield in the given weather-year and the average yield. This was calculated using the output of APSIM modelling using Kojonup climate and soil data from 1970–2019.

Results

Tactically adjusting rotation choice in response to unfolding weather conditions increases farm profit by 6% for the most profitable weather-year and 23% for the least profitable weather-year (Table 3). In early break years it is optimal to increase canola area by 63% and in late break years it is optimal to decrease canola area by 56% (Table 4). All of the tactical adjustments occur on the productive soils (LMU 3 and LMU 4). Sandy soils (LMU 2) are never tactically adjusted, always remaining in continuous pasture. On LMU 3 (sandy gravels), where barley grew in the previous year, it is optimal to establish canola in years with an early break and follow up rains (i.e., weather-years z0 and z1 – see Table 2). In early break years with no follow up rains (weather-years z2 and z3) it is optimal to follow 85% of the barley with canola and the remaining 15% with barley. In medium and late break years (z4 to z7) it is optimal to follow barley with wheat. The difference in rotation selection based on the presence or absence of follow up rains in early breaks shows that in years with an early break it is optimal to delay the rotation decision on a proportion of the area until follow-up rains are received. On LMU 4 (sandy loam), when the season breaks early and has follow up rains it is optimal to establish canola on all areas that grew spray-topped pasture in the preceding year, and on 46% of the area that had a non-manipulated annual pasture in the previous year. In all other years it is optimal to remain in annual pasture. Similar to above, the difference in rotation between early break years with and without follow up rains shows that it is optimal to delay the tactical decision to increase canola area until follow up rains occur.

For succinctness we report here on the land use tactic of seasonally adjusting the area of canola. However, in combination with choice of canola area are many other complementary tactical decisions not mentioned due to the need to be parsimonious. One illustration of complementary tactics is that, in the examined scenario, it is optimal to reseed a proportion of pasture after false breaks and dry seed wheat and canola in late break years.

AFO can undertake a wide range of sensitivity analyses, for example, changing the probability of the weather-years to represent a drier climate. Such AFO analyses show that in a drier climate it is optimal to increase the area of fodder crops in cropping programmes.

Table 3. Key descriptors of the optimal farm plans with and without tactical rotation changes for a typical Kojonup farm

^a‘Expected’ is the weighted average of all weather-years., ^b‘Max’ is the maximum across the weather-years. ^c‘Min’ is the minimum across the weather-years.

Table 4. Optimal land use area (hectares) in each weather-year

Conclusion

AFO is an advanced whole farm optimisation model that identifies the optimal suite of rotations and tactical rotational adjustments on a typical Great Southern farm. Modelling results indicate that tactically adjusting rotation choice in response to unfolding weather conditions increases farm profit by 6% for the most profitable type of weather-year and 23% for the least profitable type of weather-year. The major rotation tactics involve increasing canola area by up to 63% on productive soils in rotations following barley or spray-topped pasture, in early break seasons to capitalise on the longer growing season. Not reported are many additional complementary tactics.

AFO is a powerful tool for many different farm analysis topics, not just rotation selection and rotational tactics within weather-years.

References

ANDERSON, W., BRENNAN, R., JAYASENA, K., MICIC, S., MOORE, J. & NORDBLOM, T. 2020. Tactical crop management for improved productivity in winter-dominant rainfall regions: a review. Crop and Pasture Science, 71, 621-644.

BATHGATE, A., REVELL, C. & KINGWELL, R. 2009. Identifying the value of pasture improvement using wholefarm modelling. Agricultural Systems, 102, 48-57.

BROOKS, K. D., BENNETT, S. J., HODGSON, L. M. & ASHWORTH, M. B. 2018. Narrow windrow burning canola (Brassica napus L.) residue for Sclerotinia sclerotiorum (Lib.) de Bary sclerotia destruction. Pest management science, 74, 2594-2600.

DALGLIESH, N., COCKS, B. & HORAN, H. APSoil-providing soils information to consultants, farmers and researchers.  16th Australian Agronomy Conference, Armidale, NSW, 2012.

DIXON, G. R. & TILSTON, E. L. 2010. Soil-borne pathogens and their interactions with the soil environment. Soil microbiology and sustainable crop production. Springer.

FENG, P., WANG, B., MACADAM, I., TASCHETTO, A. S., ABRAM, N. J., LUO, J.-J., KING, A. D., CHEN, Y., LI, Y., LIU, D. L., YU, Q. & HU, K. 2022. Increasing dominance of Indian Ocean variability impacts Australian wheat yields. Nature Food.

GODFREY, S., NORDBLOM, T., IP, R., ROBERTSON, S. & HUTCHINGS, T. 2019. Using GrassGro® with@ RISK® Merino ewe/first-cross lamb enterprise risk profiles with variable prices and weather.

KINGWELL, R. 2011. Managing complexity in modern farming. Australian Journal of Agricultural and Resource Economics, 55, 12-34.

KINGWELL, R., MORRISON, D. A. & BATHGATE, A. D. 1992. The effect of climatic risk on dryland farm management. Agricultural Systems, 39, 153-175.

KINGWELL, R. & PANNELL, D. 1987. MIDAS, a Bioeconomic Model of a Dryland Farming System, PUDOC, Wageningen, the Netherlands.

KOPKE, E., YOUNG, J. & KINGWELL, R. 2008. The relative profitability and environmental impacts of different sheep systems in a Mediterranean environment. Agricultural Systems, 96, 85-94.

LAURIE, A., CURTIS, M., HEATH, R., MCROBERT, K. & DARRAGH, L. 2018. Australian agriculture: an increasingly risky business. Australian Farm Institute.

MCGRATH, S., VIRGONA, J. & ALLWORTH, B. The cost of hitting lifetime ewe targets in scanned and unscanned Merino flocks-a simulation using a dynamic model, AusFarm.  2016 Graham Centre Sheep Forum, 2016. 22-26.

MONJARDINO, M., PANNELL, D. & POWLES, S. 2004. The economic value of pasture phases in the integrated management of annual ryegrass and wild radish in a Western Australian farming system. Australian Journal of Experimental Agriculture, 44, 265-271.

MOORE, A. D., HOLZWORTH, D. P., HERRMANN, N. I., HUTH, N. I. & ROBERTSON, M. J. 2007. The Common Modelling Protocol: A hierarchical framework for simulation of agricultural and environmental systems. Agricultural Systems, 95, 37-48.

PANNELL, D. J. 1996. Lessons from a Decade of Whole-Farm Modeling in Western Australia. Applied Economic Perspectives and Policy, 18, 373-383.

THAMO, T., KINGWELL, R. S. & PANNELL, D. J. 2013. Measurement of greenhouse gas emissions from agriculture: economic implications for policy and agricultural producers. Australian Journal of Agricultural and Resource Economics, 57, 234-252.

TROMPF, J., YOUNG, J., FERGUSON, M., ANDERSON, G., TOCKER, J., VOGT, S., HERBERT, A., RITCHIE, A., HILL, J. & THOMPSON, A. 2014. More Lambs More Often - Bullet proofing your business against variable seasons.: Rural Industries Skill Training, Hamilton Victoria.

YOUNG, J., THOMPSON, A., CURNOW, M. & OLDHAM, C. 2011. Whole-farm profit and the optimum maternal liveweight profile of Merino ewe flocks lambing in winter and spring are influenced by the effects of ewe nutrition on the progeny’s survival and lifetime wool production. Animal Production Science, 51, 821-833.

YOUNG, J. M., THOMPSON, A. N. & KENNEDY, A. J. 2010. Bioeconomic modelling to identify the relative importance of a range of critical control points for prime lamb production systems in south-west Victoria Animal Production Science, 50, 748-756.

YOUNG, M., KINGWELL, R., YOUNG, J. & VERCOE, P. 2020. An economic analysis of sheep flock structures for mixed enterprise Australian farm businesses. Australian Journal of Agricultural and Resource Economics.

Contact details

Michael Young
Youngs Farm Analysis
Kojonup
Ph: 0428825054
Email: youngmr44@gmail.com