1. Introduction
The study of plant responses to changed atmospheric composition can be conducted using leaf cuvettes for the consideration of short-term effects on leaf physiology prolonged for minutes or a few hours. For the study of a whole plant and plant community (crop) responses observed in the medium-term, including weeks to several months, larger cuvettes are required. This can be accomplished using controlled environment cabinets, also known as phytotrons, which involve an artificial environment with a high level of control of the temperature, humidity, light levels and gas levels, but these involve high capital and maintenance costs [
1].
An alternative is the Free Air CO
2 Enrichment (FACE) design [
2]. This design could be used for any gas, but it has achieved some popularity in CO
2 enhancement research. The FACE design involves modification of the atmosphere of an open (wall-less) area of a crop or native ecosystem providing insights on plants’ phycological responses that are noticeable in the long term, typically from one to more years. A pipe system is used for distribution within a typically circular area, with a control system adjusting gas (CO
2) release in response to wind speed and direction and consequent ambient CO
2 levels. The wall-less nature of these systems is attractive in terms of maintaining ambient conditions other than the gas of interest; however, their CO
2 use is high. Also, the control of CO
2 levels could be compromised in windy conditions. For instance, Allen et al. [
3] reported that fluctuations in FACE experiments can be >200 ± 50 µmol mol
−1 above the target [CO
2], although some designs showed that concentrations can be maintained at or higher than 90% of the target for 93–98% of the time [
4].
OTCs are semi-open framed structures designed to maintain the desired concentration of gaseous pollutants and represent a compromise between cost and environmental modification [
5,
6]. An area of ground is enclosed by greenhouse film or transparent sheeting, with the top left open to the atmosphere. A concentrated gas source is ducted into a mixing chamber, mixed with ambient air, and the mixed gas is then ducted into the base of the OTC. The gas concentration inside the open cuvette is usually monitored to control how much gas is delivered. For elevated CO
2 studies, the source of CO
2 is typically a cylinder of compressed pure CO
2, which is diluted to the desired concentration with ambient air before introduction to the cuvette.
Early examples of studies employing OTCs include investigations on the impact of air pollution on plants, i.e., the effect of ozone (O
3) on tobacco [
7] and the effect of elevated CO
2 [
8,
9,
10]. OTC designs generally comprise a framework support for a transparent cover, e.g., polyvinyl chloride, plexiglass, or LDPE glasshouse film, forming an open-topped chamber. The chambers are typically cylindrical [
11,
12,
13], although square or rectangular [
14], hexagonal [
15,
16], and octagonal [
17,
18] shapes have been used.
The chamber size is varied, from small conical units enclosing single plants to multiple plants and larger plant communities enclosed in structures having >10 m diameter [
5]. In the context of climate change, OTCs represent a valuable tool for studying the effects of elevated CO
2 on plants, but assessing their long-term sustainability and scalability is crucial. The advantage of using OTCs relies on their sustainability due to both low environmental impact in terms of carbon footprint and costs, including reasonable initial investments and ongoing operational expenses.
Moreover, OTCs can be effectively expanded from small-scale experiments to larger ecosystems or broader research applications involving growth environment manipulation, such as, for instance, nitrogen sink and source experiments [
19]. However, while on one hand, larger structures with wider openings allow for studying the long-term impacts of elevated CO
2 on trees and forest ecosystems, on the other hand, they are more exposed to wind interference, which consequently makes it difficult to maintain the gas concentration close to the target involving higher capital and operational costs [
20,
21].
The chamber can also cause a heating effect, mitigated by the high volume of air passing through the ventilation system. However, it has been noted that the temperature inside OTCs dedicated to temperature-manipulation experiments is uniform across vertical and horizontal space [
22].
With CO
2 levels currently rising at ~2.17 µmol mol
−1 per annum [
23], there is a wide interest in undertaking studies to anticipate the impact of this change on crop performance. Such studies need to be undertaken at the whole plant/crop level to evaluate not only the effect of elevated CO
2 on leaf-level photosynthetic rates but also on plant functions such as assimilate partitioning between organs. In our case, interest in the effect of elevated CO
2 on peanut growth in the context of a ‘graze and grain’ strategy, i.e., off-take of a hay crop before flowering and pod set, initiated our interest in the use of an OTC design.
The present study outlines the design, specifications, and performance of an open-top chamber (OTC) of 1.2 m
2 area tailored specifically for investigating the impact of elevated CO
2 levels on smaller broadacre crops. This study builds upon the designs implemented in previous research, refining and adapting these earlier models to better suit the specific requirements of assessing CO
2 effects under controlled environmental conditions, particularly Messerli et al. [
15].
The key point of our approach lies in presenting a cost-effective open-top chamber (OTC) design. These modifications not only aim to enhance the accuracy and reliability of the results but also provide critical insights into the effects of increased CO2 concentrations on crop growth and development in smaller agricultural settings. This practical, economical solution allows for detailed observation and analysis, ensuring that our findings are both accessible and applicable to a wide range of agricultural research scenarios.
Although this facility was specifically designed to conduct CO2 enrichment experiments, the design provides substantial versatility to accommodate studies involving a range of gaseous pollutants. For instance, the chambers can be suited for investigating the impacts of nitrogenous compounds such as ammonia (NH4+), commonly emitted from agricultural practices and livestock operations, as well as sodium fluoride, which often originates from industrial processes including aluminium production and phosphate ore processing.
Additionally, the facility could be adapted for experiments with ozone, a pollutant produced when nitrogen oxides (NOx) from vehicle exhaust and industrial emissions reacting with sunlight. To facilitate ozone research, significant modifications would be necessary, such as the application of PTFE coatings to protect the structural integrity of the chamber from ozone’s highly reactive nature.
3. Build Instructions
3.1. Bill of Materials
A Bill of Materials is presented in
Table 1, describing the costs for the OTC framing, ventilation, and CO
2 regulation systems and irrigation.
3.2. Garden Beds
The garden beds (125 cm diameter × 72.5 cm height; from Plastic Forests, North Albury, New South Wales, Australia) were manufactured from a stabilized UV-resistant recycled plastic material (Simply Cups recycling program). Each garden bed was divided into two halves by inserting a plastic board, to allow for sub-plot treatments, e.g., of water or fertilization. Each bed was filled with 1.0 m3 of agronomically relevant soil, in this case, a ferrosol (according to the Australian soil classification), acquired from a farm located in Rossmoya, Queensland. The soil was excavated from the horizons between A1 and B2 (0–600 mm).
3.3. Chamber
The construction of each chamber was based on eight 3.6 cm square, 180 cm long recycled-plastic pickets (Plastic Forests, North Albury, NSW, Australia). These uprights were placed in an octagonal pattern against the inside wall of the garden bed ring, driven through the garden bed soil to ground level, and fixed to the bed using 3 cm hex-head screws (
Figure 1b). The posts were connected at their tops with a polyethylene (PE) pipe ring (392 cm circumference, 3 cm diam), providing more rigidity to the structure, and a truncated conical frame was then mounted onto this structure.
A glasshouse low-density polyethylene (LPDE) film of 180 µm thickness and >91% light transmission (Hortiplus180, from Redpath, Bendigo, VIC, Australia) was placed around the frame and fixed with double-sided clear mounting tape. As specified by the manufacturer, the plastic film contains surface tension modifiers to prevent condensation and dirt accumulation that could degrade light transmission and is UV-stabilized, with a 30-month warranty provided. An opening allowed operator access into the OTC. Initially, ‘door’ closing was facilitated by adhesive “hook and loop” strips with double-sided tape, applied on an opening flap of plastic film and on an upright frame. However, this arrangement was not robust. A simple hinged door frame (approx. 100 × 45 cm) was constructed of PVC angle building molding (2.5 × 2.5 cm).
3.4. CO2 Injection System
Bottled CO2 (size G, cylinder water capacity 50 L containing 31 kg CO2, or 18 m3 at 25 °C) and 1 atm Industrial Grade (gas code 081, >99.9% CO2, initial pressure 20 MPa, 2.0 Sm3 at 15 °C and 101.3 kPa) was used. Outlet pressure was set at 90–100 KPa (12–14 psi) using a CO2 regulator (BOC 801325, NSW, Sydney, Australia). The regulator outlet was fitted with a dual split connector on which two flowmeters (BOC 10521, NSW, Sydney, Australia) set at 1 L min−1 were installed. Plastic tubing (65 mm diameter) was used in the connection of the flowmeter output and the solenoid inlet and in the connection of the solenoid outlet to the mixing box of the respective OTC chamber.
The gas mixing box consisted of a plastic container (60 × 40 × 35 cm) with an opening of 5 × 5 cm width cut into the base of the container to allow air entry. A neoprene adhesive foam tape was placed along the top edges of the box to affect a seal with the lid. An electric blower (73 W, 270 m
3 h
−1, 2830 r min
−1,
Figure 3) (VKM 100, Fanco) was mounted inside the container with the intake facing the internal area of the mixing box which was employed to propel CO
2 through a lay-flat aluminium/plastic flexible ventilation ducting (~50 cm in length; 10 cm diam) connecting the blower to the chamber with a PVC dual split coil (10 cm diam), and finally, a ducting made of the same material (400 cm long; 10 cm diam) was connected to both ends and placed along the OTC circumference to serve as a plenum.
The plenum was perforated at regular intervals (~5 perforations of 22 mm diameter per linear meter), and a plastic irrigation barbed plug having 19 mm internal diameter was placed in each hole as a nozzle and fixed with reinforced insulation tape. The decision to add a nozzle instead of just leaving an open perforation in the ducting was justified by several considerations. Having outlets in each perforation ensures a steady distribution of the air mix, allowing for consistent levels of gas injection. Additionally, the angle of each nozzle can be adjusted as needed, which allows for a more targeted dispersal of CO2 in the OTC environment. Lastly, this setup prevents the perforation from enlarging with time and prevents water entering the plenum and, consequently, into the mixing box.
To improve CO2 level stability during windy periods, a corrugated PVC pipe (50 mm diameter) (Vinidex, Sydney, NSW, Australia) was installed around the chamber’s top opening, with a connection to the gas mixing box. This allowed a partial capture of exiting gases, recycling into the air mix cycle. This backflow pipe was perforated at 10 cm intervals with 19 mm diameter holes and equipped with nozzles for the plenum.
The CO
2 sensors (GMP252, Vaisala, Helsinki, Finland) were mounted inside the four elevated CO
2 chambers. The sensors (
Figure 4) were connected to two control systems (Indigo520, Vaisala, Helsinki, Finland). The CO
2 controllers were programmed to log the CO
2 concentration data at two-minute intervals, with data downloaded through a LAN linked to the Vaisala web interface and saved as a .csv file. The transmitters also activated solenoid valves (Bürkert 6013 G1/4 24V) to regulate CO
2 release on a set point of 650 μmol mol
−1, with the hysteresis value set at 150 μmol mol
−1. The value of 150 μmol mol
−1 was chosen with the purpose of avoiding short cycling of the system which can cause wear and tear of the solenoid components and over-heating. Therefore, the solenoids were switched off automatically when the internal [CO
2] exceeded 800 μmol mol
−1 and switched on when dropping below 500 μmol mol
−1.
4. Operating Instructions
4.1. Set up and Run
In our experience, the operation of the CO2 dosing chain to the OTCs can be trouble-free, although maintenance of and access to a stock of spare components (blower, sensors, valves) is recommended, as for any equipment. The primary operational issue is allowance for purchasing and delivery time of CO2 cylinders. Longer-term (over several years) maintenance includes the replacement of the plastic sheeting, yearly calibration of sensors, and checking of electrical components, e.g., blower.
The gas supply to the four OTCs was carried out using two 31 kg gas cylinders which were monitored daily for timely replacement, as needed, to avoid CO2 enrichment gaps. The size of the cylinder was relatively easy to handle; however, gas supply involved manual replacement of cylinders, which is time-consuming and labour-intensive. Such limitations could be overcome by using an automatic electronic switching system able to open the valves of another couple of cylinders, by using high-capacity tanks or by producing CO2 on-site.
Also, it is recommended to progressively raise the CO2 sensors within each chamber to a constant distance above crop canopy height throughout the growing season, to ensure that the foliage does not cover the sensor.
The use of soil moisture probes, not discussed in this report, is recommended for agronomic studies, given the biomass and water use efficiency (WUE) differences, and thus water use, of plants between treatments.
4.2. Maintenance
The ducting used for the plenum and the clear plastic sheeting are affected by the UV component of solar radiation, with degradation obvious after 12 months in the harsh tropical location of the current experiment. Metal ducting would be resistant to UV and extreme weather conditions but would result in higher cost. Annual replacement is the alternative. The chamber plastic film required cleaning with a soapy solution to remove soil dust, insects, and bird droppings. However, the light transmissibility of the clean plastic decreased to 71% after 30 months, presumably due to UV damage, and replacement (nominally at 2 years) is recommended.
4.3. Troubleshooting
During severe precipitation events, the CO2 control system showed low input signal from the CO2 probes, Vaisala GMP 252, due to condensation on the optical surface of the sensor. The issue was solved by letting the sensor dry for some hours. In some situations, it was required to remove the filter for drying (using a compressed air instrument followed by placement in a sealed container with dry silica for 12 h).
During the trials conducted on peanuts from 2021 to 2024, it was observed that crops inside OTCs were slightly more prone to pests and diseases with mild infestations of Cowpea aphids (Aphis craccivora) and peanut mites (Paraplonobia spp.)
Finally, in our experimental trials, we observed that crops cultivated in garden beds without OTC tended to grow beyond the bed edges, benefiting from unrestricted space. To mitigate the edge effect and maintain standardized growth conditions, we suggest using guard plants planted in the outer areas of the beds and regularly clipping overgrown branches during the season. These guard plants must not be considered during in-season measurements and must be discarded at harvest. Alternatively, we suggest installing short barriers, e.g., plastic or metal mesh, around the garden beds to contain the plants, especially if using crops with a decumbent or prostrate habitus like peanuts.
5. Validation
5.1. Airflow
Airflow direction was visualized using a non-toxic smoke-emitting tool (Smoke-Pen, Bjornax, Sweden), which is used for testing airflow in industrial ventilation systems. Airflow speed and temperature were recorded for each point with a handheld thermo-anemometer (Protech, model QM1646, Electus distribution, Sydney, Australia). CO2 concentration was measured using a CO2 probe, Vaisala GMP 252. Measurements were taken at regular intervals (20 cm) from ground level to above the open chamber top, both along the side wall and in the centre of the OTC. Measurements were conducted during a calm period and a period with a constant (2.78 m s−1 at time of measurement) wind. Measurements were also made of a chamber without the open conical collar top on the day of constant wind.
Chamber CO
2 concentrations were not different (
p ≤ 0.05) in the calm and windy period. CO
2 concentration profiles and airflow patterns were quite different for an OTC without the conical collar, with much less uniformity in CO
2 concentration within the chamber (
Figure 5). Air from the ventilation ducting flowed in horizontally at the base of the chamber, then tended to flow upwards through the centre part of the chamber and out into the general atmosphere. Without the collar, a downward draught along the inner wall of the chamber brought atmospheric air into the OTC, lowering CO
2 concentration (
Figure 5a). This did not occur with the collar installed, providing more consistent elevated CO
2 levels within the OTC (
Figure 5b).
The position of the CO
2 control system sensor was considered. The target concentration of 650 µmol mol
−1 was maintained at positions up to 60 cm from ground level, which exceeds the canopy height expected for this crop. A sensor height of between 30 and 45 cm was adopted, a choice that was driven by the optimal performance of the system to maintain the target concentration enhancing the accuracy in monitoring the CO
2 levels (
Figure 6).
The CO
2 target was not maintained when the control sensor was positioned at 90 cm above ground level, due to the dilution of CO
2 in the chamber air with ambient air (
Figure 7). In this circumstance, the control system attempts to maintain the set point, resulting in excess CO
2 levels within the crop.
The suction in the return line situated on the inside of the base of the plenum was visually determined using a smoke pen (
Figure 8). The suction strength was greatest in the inlets located in the proximity (30° to either side) of the junction with the main pipe, medium at 30°–120° either side of the junction, and low in the range 120°–180° on either side of the junction.
5.2. CO2 Concentration
CO
2 measurements were acquired at 2 min intervals over a 12 h period. The concentration inside the chamber was within 50 µmol mol
−1 (±8%) of the target 650 µmol mol
−1 in 43% of observations, within 100 µmol mol
−1 (±15%) of the target in 79% of observations, and within 150 µmol mol
−1 (±23%) of the target in 92% of observations (
Table 2).
5.3. Air Temperature
OTC internal temperature was recorded at hourly intervals with temperature dataloggers positioned above canopy height (45 cm) next to the CO
2 probes. There was no difference in day time (05:00–18:00) temperature averages for OTC operated at ambient and elevated CO
2 levels. However, these chambers were warmer (with approximately 2.3 °C higher temperature) than the no-chamber control plots (
Figure 9). The within-OTC and control temperatures gap was narrower in cloudy or rainy days. The night air temperature (19:00–4:30) was increased by 0.6 °C inside the OTCs (23.99 °C), compared to unchambered (control) plots’ temperature (control), with an average value of 23.40 °C. These temperature variations could alter biological processes such as plant growth, metabolic rates, and phenological events. OTCs also modify other aspects of the microclimate, including humidity, light exposure, and wind flow. Such differences, particularly between day and night, introduce additional variability that may not accurately reflect natural conditions, thereby challenging the ecological validity and reproducibility of the results across different settings. To mitigate these effects, it is recommended to standardize OTC design, accurately monitor environmental conditions, and employ ambient OTC for direct comparison, while unchambered plots can serve as a reference for current environmental conditions. This “chamber effect” can be easily detected by investigating changes in morphological traits on plants such as, for instance, flower count, fresh weight measurements, and physiological parameters such as photosynthetic rates and stomatal conductance [
25].
When the ventilation system of an OTC was turned off at the start of a day, the average air temperature through the monitored day was higher by 0.7 °C within the chamber without ventilation than the chamber with ventilation (
Figure 10).
5.4. System CO2 Consumption
With CO2 injected to maintain a concentration of 650 µmol mol−1, CO2 consumption was, on average, 3.15 L min−1 (0.22 m3 h−1) per chamber for the non-collared OTC design, operated 24 h per day. Consumption was decreased to 2.67 L min−1 (0.16 m3 h−1) when the system was operated during daylight hours (0530–1800) only. With the installation of the conical collar and recirculation of a portion of the airflow, consumption was decreased to 1.38 L min−1 (0.09 m3 h−1) per chamber, requiring the replacement of a 31 kg CO2 cylinder replacement every 8–9 days for the operation of four chambers.
A single size G compressed CO2 bottle contains 31 kg CO2, or 704 moles CO2, which, at 25.4 L mole−1, represents 17,881 L at 1 atm. This source can be diluted with ambient air to produce approximately 17,881/0.00065 = 27.5 M L of mixture, or 27,509 m3 (×2 = approx. 60 ML or 60,000 m3 of gas for 650 µmol m−2 s−1). With the blower operating at 270 m3 h−1, one bottle should, therefore, support operation for 60,000/270 = 300 h for a single chamber or 75 h for four chambers. Allowing for 35% recirculation of chamber air, the operation should be extended to 105 h for four chambers. With operation at 12 h a day, it should operate for approximately 9 days, consistent with observed usage.
The capital costs, OTC specifications, CO
2 consumption rates, temperature, and overall performance results are reposted in
Table 3. The OTCs performance is compared to the findings presented by Messerli at al. [
15], providing a comprehensive analysis of how our current results align or differed with the previous studies.
5.5. Plant Performance
Commercial peanut plants (
Arachis hypogaea L., cultivar Sutherland) were planted in December 2021 into the 12 beds, including four OTC with 650 µmol mol
−1 CO
2 (EC), four OTC with ambient CO
2 levels (AC), and four control beds. The injection of CO
2 to the four chambers (EC) started when the first leaves were fully developed (V-3) [
26], and injection was terminated at harvest, 30 days after planting, when the peanuts reached the flowering stage (R1). The chamber’s internal temperature was monitored with temperature probes (HOBO dataloggers, Bourne, MA, USA). Leaf gas exchange parameters were measured on one per plot of the last fully expanded young leaflet at the canopy top, one day before harvest, using a portable photosynthesis system (LI-COR 6800, LI-COR Inc., Lincoln, NE, USA). Measurements were conducted with a solar photosynthetic photon flux density (PPFD) manually set at 2000 µmol mol
−1. Humidity reference was set at 60%, fan speed at 8000 rpm, air flow at 500 mmol
−1, and cuvette temperature at 30 °C. The cuvette [CO
2] was adjusted based on the growth condition: 400 µmol mol
−1 for AC OTC and 650 µmol mol
−1 for EC OTC. Measurements were taken between 9:00–11:00 a.m. and 14:00–16:00 p.m. The intrinsic water use efficiency (WUE) was calculated as a ratio between net CO
2 assimilation rates and the transpiration. The crop performance results are reported in
Table 4.
Over the 30-day growth period, the accumulated above-ground biomass, measured as the dry weight of 10 plants per chamber, was similar between the control (no chamber) and the OTC with ambient CO
2 levels (AC), with values of 61.4 g (no chamber) compared to 60.9 g for AC. In contrast, the accumulated above-ground biomass was ~40% greater in the OTC with elevated CO
2 (EC) compared to the chambers with ambient CO
2) (
p = 0.0060) (
Table 4). Similarly, net CO
2 assimilation rates were greater for EC-grown peanuts compared to peanuts grown under ambient control and AC. This difference, however, was not significant. In contrast, intrinsic WUE was greater for EC-grown peanuts compared to control and AC-grown peanuts, and this difference was significant (
p = 0.0024). These results are in close agreement with previous studies reporting the stimulation of biomass, photosynthesis, and water use efficiency of crops when grown under elevated CO
2 [
27].
6. Conclusions
The design of an open-top chamber facility described in this research effectively showcases the impact of elevated CO2 levels on crops, combining precision with cost effectiveness for a sustainable approach to agricultural research. Details of its construction and the results provided for the facility evaluation show that it represents a viable and practicable alternative to costly methods of ascertaining the effects of CO2 on plants.
The test performed on the OTCs and CO2 concentration control equipment highlighted the system’s efficacy in maintaining the internal target concentration throughout the trial, which, for this research, was set to 650 ± 50 µmol mol−1. Additionally, recycled chamber air was used, reducing overall CO2 usage from 2.68 min−1 to 1.38 L min−1.
The daily average CO2 consumption observed during daylight hours injection in this study was 1.56 kg m−2.
Some adjustments and system fine-tuning were required to improve CO2 consumption rates and limit wind interference. Simple modifications implemented to the OTC structure, such as partially modifying the opening with a conical top opening (frustum), resulted in effective measures for mitigating the issues caused by wind and improving the system’s functioning. However, the use of a conical frame for the frustrum could lead to changes in the indoor–outdoor continuum as a trade-off, modifying the amount of received rainfall. However, such an issue can be overcome with automated irrigation systems set to release the amount of water according to the crop need, geographical location, and soil characteristics. A potential limitation that could affect the interpretation of the effects of elevated CO2 is the increased temperature inside the chambers, which could confound the effects attributed solely to CO2. As a mitigation measure, each chamber ventilation system was equipped with a blower which effectively contributed to reduce temperature through the high-volume air produced. Furthermore, the facility included OTCs with ambient CO2 levels for direct comparison and “control plots” without chambers which are recommended to reflect current CO2 and environmental conditions accurately.