Previous PageTable Of ContentsNext Page

The Australian Grains Free Air Carbon dioxide Enrichment (AGFACE) experiment – specifications and scope

Rob Norton1, Mahabubur Mollah2, Glenn Fitzgerald3 and David McNeil4

1 The University of Melbourne, Private Bag 260, Horsham, Vic, Email
Department of Primary Industries, 110 Natimuk Rd., Private Bag 260, Horsham, Vic,
Department of Primary Industries, 110 Natimuk Rd, Private Bag 260, Horsham, Vic,
Tasmanian Institute of Agricultural Research, Private Bag 98, Hobart, Tas,


Australia is particularly challenged by impacts of rising atmospheric carbon dioxide and the consequent perturbations in climate. The Australian Grains Free Air Carbon dioxide Enrichment (AGFACE) project in Horsham, Victoria was designed to simulate predicted atmospheric carbon dioxide levels in the year 2050. The experiment measures the interacting effects of carbon dioxide (ambient CO2 380 ppm; elevated CO2 550 ppm), irrigation (rainfed, irrigated), nitrogen (0, +), and variety (Yitpi, Janz) on wheat growth and production. Carbon dioxide was injected over the crop in open-air 12 m rings from emergence (July) until maturity (December). The most technically difficult aspect of estimating the impact of climate change is to identify the interactions between CO2 and temperature, and to do this, the experiment uses two sowing times at both the main site (Horsham), and a second site at Walpeup in a warmer and drier environment. These conditions should produce 6C range of temperatures during anthesis, and the data on growth and yield can be used to calibrate existing crop simulation models.

Monitoring within the rings at Horsham during 2007 showed that all 8 elevated CO2 rings maintained levels at very close to the 550 ppm target for more than 90% of the time. These data provide confidence in the design of the control and delivery systems developed for this experiment.


Climate change


FACE stands for Free Air Carbon Dioxide Enrichment which is a field based technique to raise the level of CO2 in the air around growing plants. The FACE technique has been used internationally at more than 30 sites (Brookhaven National Laboratories 2008). Trees, crops or pastures are grown within rings that have a regulated supply of CO2 fed into the experimental area, controlled by sensors that regulate the amount and position of release to maintain levels at the required concentration. FACE systems have evolved to enable plants to be grown under natural conditions for long periods.

The Intergovernmental Panel on Climate Change has indicated that the observed trend in atmospheric CO2 will continue to rise, and during this century may well exceed 550 ppm. Because CO2 is a primary input into photosynthesis, an enriched atmosphere will affect plant growth and the effects will be confounded by changes in water use, nutrient uptake and temperature. FACE systems are used to investigate how crop, forest or natural systems will respond to elevated CO2 levels. While there have been evaluations of existing systems models to predict climate change impacts on wheat production (Asseng et al. 2003), there is a need for good quality validation data sets to refine and improve the routines in these models. Once calibrated, these models can be used with greater confidence to investigate the impact of future climate change. The FACE experiments also become a platform for multi-discipline ecosystem scale research, such as on soil quality, pest and disease responses and product quality.

In 2006, the Australian Greenhouse Office (now Department of Climate Change) released a report that identified that there was a significant knowledge gap in the effects of elevated CO2 on water and nutrient limited farming systems in Australia (Hely et al. 2006). This report considered a range of strategies to generate information to fill this gap, and proposed the establishment of a limited number of large-scale FACE systems located across Australia, supplemented by a number of mini-FACE systems in more remote areas would provide the most appropriate design to obtain the required data on the impacts of climate change on rainfed production systems.

This paper aims to provide some background to the establishment of this facility as well as describing the research facility of eight large CO2 enrichment rings and eight control areas established at Horsham in 2007. An additional satellite site has been commissioned at Walpeup in the Victorian Mallee with eight 4 m elevated CO2 rings and 8 control areas.

The Australian Grains FACE Array


Infrastructure for the Horsham FACE consists of 45 tonne CO2 storage tank, a passive heat exchanger to warm liquid CO2 to ambient temperatures, gas regulators and 3 km gas pipes to deliver CO2 to plots in the field. This infrastructure can deliver up to 1 t CO2 per hour to the rings. Irrigation water pipes (2 km) and electricity supply for CO2 controllers (2 km) is also reticulated across the field site.

Eight 12 m diameter elevated CO2 rings have been constructed at Horsham, each with separate controllers that measure wind direction, wind speed, and CO2 concentration at the centre of the ring. The controllers maintain CO2 concentrations in the crop within the desired range (550 ppm) by adjusting CO2 release from the three upwind segments of the ring. As the wind direction, alters the segments venting CO2 change. CO2 is only released when photosynthesis is occurring. Ambient CO2 concentration is monitored within two control plots. Figure 1 is a diagrammatic representation of the elements of the FACE system.

Figure 1. Diagramatic representation of the elements in an AGFACE ring in the field.

Control of CO2

Carbon dioxide is released during daylight hours only and commences once the crop has emerged and terminated at physiological maturity. Ventilation of CO2 will cease when ambient temperatures fall below 4C and also when wind speeds are greater than 10 m/s.

Figure 2. CO2 concentration at the centre of the AGFACE ring graphed against windspeed from data logged on 24 October 2007.

The central sensor measures CO2 concentration every second, and every 4th recording logged along with one minute average concentrations. Figure 2 shows a trace of the central CO2 average one minute concentration in one ring on one day graphed against windspeed. The set point recorded from the centre of this ring was 55220 ppm.

When taken over a longer period, the control system maintained the concentration of CO2 at the centre of the ring at ≥ 90% of the target (ie. 495 ppm) for 97% of the time. While operational (during the daylight hours), the mean CO2 concentration inside the ring was 549 ppm with the standard deviation of 25 ppm.

Experimental Treatments

The experimental treatments in the core experiment will be:

  • a) Ambient CO2 & 550 ppm CO2 – current CO2 is approximately 380 ppm, and the level selected for enrichment is the IPCC prediction for 2050 (scenario A1B). This CO2 concentration is also still on the responsive part of the A/Ci curve for wheat, which is usually saturated at about 650 ppm.
  • b) Sowing times – June and late July – which act as ways to alter temperature during growth, particularly grain growth. In 2007, these sowing times gave a 6oC difference in temperatures at the start of anthesis. Other methods of altering temperature are very expensive (eg large IR heaters) or provide significant artefacts to the treatments especially changing vapour pressure deficit.
  • Ambient water supply & Supplementary Irrigation
  • c) Water regime – no rain out shelters are deployed on this site, but presowing water levels can be altered by using soil covers. Supplementary irrigation is applied to provide a contrast between the treatments. In 2007, an additional 50 mm of water during grain fill provided a 30% increase in grain yield.
  • d) Within each half-ring, two wheat varieties (Yitpi and Janz) are grown, and additional N is added to Yitpi to give three treatments – Yitpi N0, Yitpi N+, Janz N0. ). These cultivars were chosen for this experiment as they are widely grown commercially in southeastern Australia and are also genetically quite different (Ogbonnaya et al. 2006).

The use of split plot designs was identified as the most efficient for FACE facilities (Filon et al. 2000), with two-way nested designs and split plot designs having similar power in testing CO2 main effects. Those authors recommended the use of a second treatment factor with many levels within each ring in order to obtain split-plot designs that provide a powerful test of interactions between treatment factors.


Because of the number of factors considered and the limited size of the rings, experimental plots are small. Within each half ring there are two plots per treatment each of 1.8*4.0 m per treatment. One plot is used for destructive harvests at DC30 and DC65, while the other is used for non-destructive measurements and a final grain harvest. Non-destructive measurements are water use (neutron probe), canopy cover (ceptometer), NDVI (CropCircle™), canopy temperature (IR Thermometer) and a range of imaging (digital, thermal, multispectral) are taken between the destructive harvests. Leaf photosynthesis and other physiological measurements are also taken.

An Array Concept

It is felt important that this research not be focused on the effects of elevated CO2 alone, but aims to investigate the interaction between elevated CO2 and other factors likely to alter under climate change. The proposed concept was for an array, with a major site at Horsham plus several smaller sites. This would be used to generate data that could be used to calibrate models of Australian and world wheat production systems so they could be run with current and future climate scenarios and investigate interactions among CO2, water and temperature. The approach of calibrating a model over a broad range of environmental conditions to develop interaction models has been very successfully used to analyse interactions among light, N, water, temperature and age affecting the growth of orchard grass in a validated agroforestry system model (Peri et al. 2006).

The cost of establishing and running a small FACE facility is approximately $200 k to establish and $160 k per year to operate (excluding scientific input) and this cost may preclude additional sites being established in Australia. However, the establishment of two sites (Horsham and Walpeup) on contrasting soil types (Vertosol and Calcarosol) provides a broader data set for model calibration and should increase the range of environments for which the models can be confidently be used.

Additional Research within the AGFACE

As well as the main experiment, there are some investigations within the FACE rings on the effects of biotic factors on wheat. In particular, in 2007 the progress of Stripe Rust (Puccinia striformis), Crown Rot (Fusarium spp) and Barley Yellow Dwarf Virus was evaluated under elevated CO2 and this work will continue in 2008. Additional research on N transformations, water use, canopy structure, grain quality and grain filling are proposed for 2008. In future years the impacts of elevated CO2, temperature and rainfall will be investigated for other crops.


The AGFACE array will enable the identification of high CO2 impacts within agricultural systems and provide information on the way management can adapt to take best advantage of these changes. The project will generate high quality data sets on plant growth, water use, nutrient uptake and grain yield of wheat in response to elevated CO2 levels from field experiments across the grain belt of Australia. These data will be used to calibrate existing crop simulation models so they can be run over current and future climate scenarios and include interactions between CO2, water and temperature.


Asseng S, Jamieson PD, Kimball B, Pinter P, Sayre K, Bowden JW, Howden SM (2003). Simulated wheat growth affected by rising temperature, increased water deficit and elevated atmospheric CO2. Field Crops Research 85, 85-102.

Brookhaven National Laboratories (2008). The FACE Program - (accessed May 12, 2008).

Filon M, Dutilleul P, Potvin C (2000). Optimum experimental design for Free-Air Carbon dioxide Enrichment (FACE) studies. Global Change Biology 6, 843-854.

Hely S, Slattery W, Reeves T, Ugalde D (2006). Options for Investigating the Impacts of Elevated Carbon Dioxide on Agricultural Production in Australia, Australian Greenhouse Office 24 pp.

Ogbonnaya FC, Imtiaz M, DePauw RM (2006). Haplotype diversity of preharvest sprouting QTL’s in wheat. Genome 50, 107-116.

Peri PL, Moot DJ, McNeil DL (2006). Validation of a canopy photosynthesis model for cocksfoot pastures grown under different light regimes. Agroforestry Systems 57, 259-272.

Previous PageTop Of PageNext Page