APSIM NextGen Canola model – new flexibility and functionality

Dr Julianne Lilley1, Dr Jeremy Whish2, Dr Neil Huth3, Dean Holzworth3, Neville Herrmann1 

1CSIRO, Canberra, Australia,

2CSIRO, St Lucia, Australia,

3CSIRO, Toowoomba, Australia


The Agricultural Production Systems sIMulator (APSIM) consists of a suite of models developed to simulate the soil, crop and management interactions of farming systems. The canola model was developed in the late 1990s and has been used to investigate different aspects of canola production across Australia ever since. More recently the APSIM Initiative has been building the next generation of APSIM to create a simulation framework that can run larger and more complex simulations at multiple temporal and spatial scales. The Plant Modelling Framework for APSIM NextGen is more versatile and the efficiency of capturing our understanding of crop processes within the model is improved. For example, new routines describing genetic parameters, or the physiological response to new traits or events such as abiotic or biotic stress are readily incorporated, modified and tested.

The new APSIM Next Gen Canola model has been validated against an extensive set of detailed phenology observations (n>400) and yield observations (n>1000) from field experiments that include a wide range of cultivars, locations, seasons and agronomic treatments. A feature of the new model includes improved phenology prediction by better accounting for response to vernalisation and thermal time with parameters to describe new cultivar phenotypes. This allows the evaluation of a novel phenology type in the farming system and improved prediction of canola development beyond the traditional growing regions.


Dr Julianne Lilley is Team Leader of the Integrated Agric Modelling & Decisions team within Agriculture and Food which develops and uses soil, crop, pasture and livestock models in a farming systems context for application in agricultural research across Australia and internationally. Julianne is a crop scientist with over 25 years research experience in crop physiology. Her research has included climate change impacts, crop root system function, crop water-use efficiency, dual-purpose cropping, whole-farm productivity and resource protection. She applies crop physiological understanding embedded within crop simulation models to greatly expand the outcomes of experiments and draw robust conclusions about the impacts of agricultural management decisions.

She is currently leads research investigating farm management practices which optimise profitability of canola, understanding environmental triggers and the underlying genetic control of flowering in canola, and developing the Yield Gap Australia website, a tool developed at CSIRO to inform policy makers and grain producers of the gap between potential and actual yields from local to national scale.

Slow early versus fast late – which works best and where?

Rohan Brill1, John Kirkegaard2, Danielle Malcolm3, Andrew Ware4

1Brill Ag, Ganmain, Australia,

2CSIRO, Canberra

3NSW DPI, Wagga Wagga

4EP Ag Research, Port Lincoln


Optimum start of flowering dates are established for canola across much of Australia but it is possible to flower at the optimum time by sowing slow varieties early or fast varieties later. We conducted 14 experiments across eastern Australia in 2017 and 2018 to determine the optimum sowing strategy (sowing date and phenology type) across a range of yield scenarios (0.4 to 5.7 t/ha). We found that early sowing of slow developing varieties was most successful at sites that had received high (>200 mm) fallow rainfall. At these sites there was also a consistent benefit of selecting a high vigour hybrid variety compared with a low vigour open-pollinated triazine tolerant variety. Later sowing of fast developing varieties was advantageous at low yielding sites and surprisingly at very high yielding sites. Canola growers can adjust their canola sowing strategy (sowing date and phenology type) based firstly on fallow rainfall and secondly on expected in-crop rainfall.


Rohan is an agronomist and farmer with Brill Ag (since 2020) at Ganmain in southern NSW with current research, development and extension projects in canola and pulses.  Previous to this, Rohan was an agronomist with NSW DPI at Wagga Wagga, focussing on tactical canola agronomy.

Manipulating canola canopies through agronomy & genetics in the high rainfall zone

Dr Jens Berger1, Dr Andrew Fletcher1, Mr Sam Flottmann1, Mr Adam Brown1, Dr Heping Zhang1, Jeremy Curry2, Mark Seymour2 

1CSIRO, Wembley, Australia,

2DPIRD, South Perth, Australia


Canola production is becoming increasingly important in the high rainfall zone, where yield is determined primarily by biomass, trading off harvest index (HI).  However, input management for high biomass production carries greater financial risk, particularly if  rainfall is below average. Moreover, high biomass production can have negative consequences for growers, including harvesting difficulties associated with tall crops, high stubble loads and in-season water use, and increased Sclerotinia risk.

To understand grower capacity to influence the canola yield/harvest index trade-off we used a factorial range of agronomic and genetic levers to manipulate canopy size and yield potential in on-farm HRZ trials:

  • Cultivar: high vigour (RR) versus low vigour (TT)
  • Plant density
  • Input levels (N x S)
  • Early season grazing

These treatments had a huge impact on plant height (119-176 cm), biomass (8.2-15.7 t/ha) and yield (1.9-6.2 t/ha).  Yield differences were driven by HI, while HI was dominated by genetics and its interaction with agronomy.  HI was lower in RR than TT canola but there were important differences/interactions within both groups.  High HI cultivars had stable HI that was not modified by agronomy (TT: Invigor 450 (35%), Hytech Trophy (34%); RR: GT 53 (32%), P45Y28 (29%)).  Low HI cultivars (TT: P45T03 (29%); RR: H540XC (23%)) further reduced their HI in treatments that increased biomass.  Experiments are currently underway to unpick this contrasting interaction.

By understanding biomass partitioning in canola we hope to de-risk production for growers giving them the tools to manage the canopy they require.


Jens has almost 30 years experience studying adaptation of dryland crops using a combination of physiology, genetics and agronomy.  The CSIRO canola team has been using these approaches to focus on the performance of canola across Western Australian rainfall gradients.

Capturing phenological diversity in canola using fewer key environments

Dr Jeremy Whish1, Dr Julianne Lilley2, Dr Susan Sprague2, Mr Brett Cocks3, Ms Brook Anderson3, Ms Nell Evens7, Mr Barrett Sinclair5, Mr Andrew Ware6, Mr  Matthew Nelson4, Dr Bill Bovill2, Dr Bangyou Zheng1, Mr Alexandre Boyer2, Dr Shannon Dillon2, Dr Chris Helliwell2 

1CSIRO Agriculture and Food, St Lucia , Australia,

2CSIRO Agriculture and Food, Canberra, Australia,

3CSIRO Agriculture and Food, Toowoomba, Australia,

4CSIRO Agriculture and Food, Wembley, Australia,

5Kalyx Australia , Perth , Australia,

6EPAG Research , Port Lincoln, Australia,

7Kalyx Australia , Young , Australia


Phenological development in canola is driven by temperature (thermal time, vernal time) and daylength which vary significantly across Australian canola-growing region.  Minimising the number of sites, seasons and sowing dates required to understand these responses in a large diversity panel of canola germplasm is desirable.  We examined the accumulation of thermal and vernal time across Australia to identify a minimum set of environments that could adequately separate the phenological drivers among the varieties.  Using the cardinal temperatures for thermal and vernal time accumulation previously established for canola, a series of long-term simulations using historic weather data were conducted across relevant environments and sowing dates in Australia.  The ratio of vernal to thermal time accumulation was used to remove duplicate environments until eight overlapping environments, comprising four sites (Boorowa NSW, Beverly WA, Gatton QLD and Cummins SA) with mid-April and mid-May sowing were identified. (Table 1).  Concentrating efforts at four sites with two sowing dates has facilitated an intensive and cost-effective sampling strategy to identify flowering response in a large diversity panel.

Table 1.  Minimum environment set with differing ratios of vernal and thermal accumulation

  1. Fast-Vernal : Slow-Thermal (Boorowa NSW, May-15)
  2. Fast-Vernal : Moderate-Thermal (Boorowa NSW, April-15)
  3. Moderate-Vernal : Decreasing-Thermal (Beverley WA, May-15)
  4. Moderate-Vernal : Slow-Thermal (Beverley WA, April-15)
  5. Moderate-Vernal : Moderate-Thermal (Gatton Qld, May-15)
  6. Slow-Vernal: Fast-Thermal (Gatton Qld, April-15)
  7. Very-slow-Vernal : Fast-Thermal (Cummins SA, May-15)
  8. No-Vernal : Fast-Thermal (Cummins SA, April-15)


Dr Jeremy Whish is a Principal Research Scientist. He uses modelling to understand the complexity of farming systems and identify management strategies to reduce production risk. He has published broadly on many aspects of the farming system including biotic and abiotic constraints, crop physiology, phenology, and risk within the management of crops and rotational sequences. 

Hyperyielding Canola Year 1 – Management for 5 t/ha canola

Rohan Brill1, Darcy Warren, Kat Fuhrmann, Nick Poole 

1Brill Ag, GANMAIN, Australia, 2FAR Australia, , , 3FAR Australia, , , 4FAR Australia, ,

The Hyperyielding Crops project was rolled out across mainland Australia in 2020 with canola sites at Wallendbeen, NSW; Gnawarre, Victoria; and Millicent, South Australia. The aim of the project for canola is to determine the management strategies that achieve 5t/ha of canola in these environments. The primary management strategies investigated in 2020 included, cultivar choice, nutrient management, disease management and plant density. Yield of the best treatment in 2020 was >5t/ha at Wallendbeen and just below 5t/ha at Gnawarre and Millicent. The highest yielding treatment was generally at least double the yield of the lowest yielding treatment at each site. Cultivar choice was the single most important influence on yield variability at all three sites, on average accounting for 2 t/ha difference in yield. Nitrogen management was the next most important influence on yield, accounting for an average 1.2 t/ha difference in yield. Fungicide response was small at Wallendbeen (0.3 t/ha) and Gnawarre (0.2 t/ha) and not significant at Millicent. Where there was a yield response to fungicide recorded it was difficult to attribute the yield loss to any one disease. Plant density had no impact on yield at Wallendbeen and Gnawarre but there was a small penalty from low (<15 plants/m²) at Millicent. Similar trials incorporating an early sowing of winter canola will be included at the three sites in 2021 as well as at Frankland River in Western Australia.


Rohan is an agronomist and farmer with Brill Ag (since 2020) at Ganmain in southern NSW with current research, development and extension projects in canola and pulses.  Previous to this, Rohan was an agronomist with NSW DPI at Wagga Wagga, focussing on tactical canola agronomy.



Thermal time and vernal time response of commercial and pre-release Canola cultivars

Mrs Danielle Malcolm1, Rohan Brill2, Hongtao Xing3,4

1NSW DPI , Wagga Wagga, Australia

2Brill Ag, Ganmain, Australia,

3NSW Department of Primary Industries

4Wagga Wagga Agricultural Institute


Trials were conducted at Wagga Wagga in 2018, 2019 and 2020 with commercial and pre-release canola cultivars sown in late March and late April, to investigate the phenology differences between varieties.

Contrasting seasons were experienced across the experimental period, with differences in varietal response to thermal and vernal time observed  Vernal time was accumulated quicker in 2020 (relative to thermal time) than the previous two years but thermal time was accumulated at a slower rate. This meant that certain varieties did not need as much thermal time to reach flowering.

In 2020 both fast and slow developing cultivars responded to the high proportion of vernalising temperatures by releasing the ‘handbrake’ on their vernal response, reducing the amount of thermal time required to switch from the vegetative growth stage to the reproductive growth stage. From the early sowing date, varieties flowered earlier than in previous years. The fast-mid cultivar 44Y90 (CL) flowered a month earlier in 2020 than 2019. The slow developing cultivars 45Y91 (CL) and ATR Wahoo flowered 17 and 13 days earlier. Nuseed Diamond, considered a spring cultivar with little to no vernalisation response, flowered 21 days earlier in 2020 than 2019 despite 2020 being a cooler season.

The results observed for the mid and slow developing cultivars were all able to be explained by the temperature differences across season, however the fast development of the spring cultivar Nuseed Diamond in 2020 is more difficult to explain and may indicate that there are other mechanisms at play.


Danielle has been working in Canola Agronomy team at DPI since 2016 with involvement in various experiments. Danielle has managed the phenology trials at Wagga with commercial and pre-released cultivars. Danielle has also been involved in the Optimized Canola Profitability project and High Yielding Canola project and more recently is involved with the Frost tolerance project in Wagga and a project looking at improving production of grower retained Open Pollinated canola seed.



Prediction of canola phenology through integration of single nucleotide polymorphisms (SNPs) with a crop model

Dr Bangyou  Zheng1, Dr Jeremy Whish1, Dr Julianne Lilley2, Mr Alexandre Boyer2, Dr Alec Zwart3, Dr Chris  Helliwell2, Dr Shannon Dillon2 

1CSIRO Agriculture and Food, Queensland Biosciences Precinct 306 Carmody Road, St Lucia, Australia,

2CSIRO Agriculture and Food, GPO Box 1700, Canberra, Australia,

3CSIRO Data61, GPO Box 1700, Canberra, Australia


Canola phenology is a major determinant of the adaptation of canola to different environments; the productivity of canola can be maximised by targeting phenology to the optimal flowering window. Although many crop models have been parameterised to predict canola phenology, these models are reliant on estimation of model parameter values from phenology scores for a given variety obtained in multiple environments. Canola breeding and agronomy research would benefit from robust models that predict phenology based on known genes or single nucleotide polymorphisms (SNPs). Here, we present a framework that blends genomic prediction and crop process modelling to predict phenology for a given variety across a range of environments based on its genome. Genome wide SNP data from more than 300 canola cultivars were integrated into a canola model in APSIM Next Generation (APSIM NG) to predict phenology traits (e.g. green bud and first flower). Datasets collected from controlled environments and field experiments with several time of sowing and locations were used to train and validate the canola model in APSIM NG (APSIM-Release) and the new integrated genomic model (APSIM-GP). In APSIM-GP the model parameter values are estimated using global optimisations directly from genotypic information (i.e. SNP data) to phenotypic values (i.e. observations).  Both APSIM-Release and APSIM-GP could predict canola phenology of all cultivars with an acceptable level of accuracy. APSIM-GP can be used by breeders and farmers to predict canola phenology for new cultivars with only known SNP information to accelerate breeding programs and close the yield gap.


Dr Bangyou Zheng is a data and experimental scientist at CSIRO in Brisbane, Australia. He received his PhD degree in agriculture science at China Agricultural University. His research focuses on the crop physiology, crop genotype to phenotype prediction, crop modelling, climate adaptation, high throughput phenotyping, big data management, processing and visualization.



A simple potential yield calculator for canola in the high rainfall zone

Dr John Kirkegaard1 

1Csiro Agriculture And Food, ,

Benchmarking farmer yield against a physiologically defensible yield potential has been an effective tool for crop yield improvement in Australia.  The approaches used to assess yield potential in water-limited environments have sensibly focussed on water-limited yield potential.  These vary from simple and accessible seasonal calculators such as that developed by French and Schultz (1984) up to sophisticated daily timestep models such as APSIM which underpins yield predictions in YieldProphet.  In the high rainfall zone, there will often be cases where water supply is not the major limit to yield, but rather the ratio of light and temperature (photothermal quotient – PTQ) in the critical period of yield formation.  Lower light and warmer temperatures in this period speeds development and reduces photosynthesis during the period when grain number is determined.  The relationship between PTQ and yield potential in wheat has been well described and has been demonstrated to be an accurate predictor of potential yield when no other factors are limiting (e.g. water, nutrients, pest, diseases, temperature extremes).  We used published physiological relationships between wheat and canola to propose a simple relationship between potential yield and PTQ in the critical period for canola and tested it against high yielding canola crops in Australia’s high rainfall zone including a record 7.2 t/ha commercial crop in southern NSW.  A simple ready reckoner was developed that used a cascade of PTQ, water supply and N supply to consider the yield potential of field-grown canola in the high rainfall zone.


Farming Systems Agronomist



Management packages to increase canola yield and profit in low-rainfall southern cropping environments.

Dr Therese McBeath1, Dr Elizabeth Meier2, Mr Andrew Ware3, Dr John Kirkegaard4, Mr Michael Moodie5, Mr Bill Davoren1, Mr Ed Hunt6

1CSIRO, SA, Australia

2CSIRO, QLD, Australia

3EPAg Research, SA, Australia

4CSIRO Agriculture and Food, ACT, Australia

5Frontier Farming Systems

6Ed Hunt Ag Consutancy, Wharminda, SA, Australia


The successful inclusion of break crops such as canola remains a challenge for farming systems in low-rainfall cropping environments. There have been significant gains in canola productivity through early sowing, understanding of critical stress periods, hybrid cultivars and improved nitrogen (N) fertilisation in many canola producing regions but require careful adaptation for risky low-rainfall environments. A series of experiments were implemented over 4 growing seasons in the low-rainfall environments of southern Australia to tested combinations of sowing date, cultivar selection and N management strategies. This was combined with simple whole-farm profit-risk analysis to identify combinations of practices where the potential production and risk were understood. Earlier sowing (April) was  only beneficial where seasonal conditions allowed good crop establishment.  Analysis of hybrid cultivars revealed a yield advantage of >20 % over open-pollinated (OP) cultivars needed be sustained across the full range of season deciles to be profitable. There was relative insensitivity to the timing of N application, but an adequate dose of N was critical to improve canola productivity. Our combined analysis showed there are opportunities to make significant gains in the yield (up to 110% compared with current standard practice)  and profit-risk outcomes (in the order of 30% increased gross margins across all season types) for canola in low-rainfall environments.


Therese is a Research Team Leader in the Systems Program of CSIRO Agriculture and Food. She works across a range of systems and soils project to solve production and profit issues for Southern cropping and mixed cropping-livestock farms.

Assessing and quantifying frost damage in canola using purpose-built frost shelters

Dr Rajneet Uppal1, Danielle  Malcolm1 

1NSW DPI, Wagga Wagga, Australia

Frost has become a major threat to Australian canola industry due to increased frequency of extreme abiotic stress events and exposure of early sown canola to greater frost risk. Significant yield loss can occur when frost events coincide with reproductive development of canola. In 2017, frost reduced canola grain yield by approximately 0.3 t/ha in NSW, a total of 120,000 tonnes valued at $63 M. Previous research on frost tolerance focus on using multiple sowing dates to create a range of flowering dates to understand frost response however, the commercial relevance of the experimental results is doubted because true ‘frost exclusion” treatment is generally missing. A replicated field experiment with two canola varieties and two temperature treatments (frosted and non-frosted) was conducted at Wagga Wagga Agricultural Institute, Wagga Wagga, Australia in 2020 to develop a method for determination of frost damage in canola crop. Purpose- built frost shelters were used to insulate canola plots in “frost exclusion’ treatment (non-frosted) and were compared with ambient temperature in control plots (frosted). Frosted and non-frosted treatments were at par for grain yield, oil quality and harvest index due to fewer frost events, lack of severe (-4°C) frosts at mid-flowering to grain -filling and compensatory effect of good in-season rainfall in 2020.


To come


About conferences.com.au

conferences.com.au provides delegate registration, website and app solutions, and financial management for conferences, conventions and scientific meetings.

Terms & Conditions

All registrations and bookings are subject to our standard term and conditions.

Contact Us

Please contact the team at conferences.com.au with any questions regarding the conference.
© 2017 - 2020 Conference Design Pty Ltd. conferences.com.au is a division of Conference Design Pty Ltd.