Liming acidic soils creates profits, land use options but often more emissions

Study region

The study region (pale yellow to orange shades in Fig. 1) in Australia’s southwest is dominated by a Mediterranean-type climate, which supports annual winter cropping and pasture production. The average growing season rainfall in the study region ranges from 450 mm near the region’s wetter margins down to only 175 mm on the drier inland margins. The size of farms in the study region is negatively correlated with growing season rainfall. The northern and eastern parts of the study region are dominated by large crop-dominant farms whereas smaller mixed crop and sheep farms are common in the southern parts of the study region.

Fig. 1.

The study region and locations evaluated. Colour legend shows rainfall in mm.

Fourteen different locations (see Fig. 1) across the study region were selected to represent different rainfall and crop-growing environments. The study region supplies around half of Australia’s wheat exports (DAFF 2023), equivalent to almost 4% of global wheat exports (Observatory of Economic Complexity 2024). In 2022 the study region produced 26 million tonnes of grain (GIWA 2023) but in 2023, a year of very low rainfall, grain production was 15 million tonnes (GIWA 2024). The region also supported 12 million sheep in 2021 (DPIRD 2021) but the sheep population is trending downwards in response to the intended closure of the live sheep export trade (MLA and AWI 2024).

Modelling

The Economic Valuation of Alternative Land Use Sequence (EVALUS) (Kharel 2022; Kharel et al. 2022) model was used to study the impact of liming on land use sequence profitability and emissions. At each of the 14 locations, for a representative farmer’s field, its land use sequences, profits and emissions were simulated for 30 years. The model tracks fertiliser uses and soil indicators, including soil acidity, over the specified period. The Optlime model (Gazey 2008), later known as iLime (Fisher et al. 2019; DPIRD 2019), was incorporated into EVALUS to study the relationships between fertiliser applications, acidification, and liming practices.

The EVALUS model calculates the fertiliser demand based on a yield target, as explained in d’Abbadie et al. (2024a). Based on the fertiliser input and crop yield, the Optlime model within EVALUS calculates the acidification of the soil, based on factors like soil type and soil carbon. Then the available aluminium at each pH level is calculated and yield is penalised based on the crop’s tolerance to aluminium. When lime is added, based on different strategies, it reacts with hydrogen ions in the soil to neutralise the soil and increase soil pH over time. The Optlime model also simulates the movement of pH through the soil profile depending on rainfall and the neutralisation rate of the surface soil.

The lime used in the model is Generic West Cost Limesand as defined in the iLime app (DPIRD 2019) where the net weight of CaCO₃ per tonne of lime is 89.3% and the surface area is 21,837 m² per tonne for applied CaCO₃. The lime is assumed to be sourced from the nearest coastal lime pit.

The gross margin data generated in EVALUS reflects crop yields that are gradually affected by the changes in soil pH as influenced by applying lime at various amounts and frequencies or by not applying lime at all.

Greenhouse gas emissions

In this study, we report the three categories of greenhouse gas emissions, as stated in the Greenhouse Gas Protocol Agricultural Guidance (GGP 2015). Scope 1 emissions are those produced directly by farm operations, whereas Scope 2 emissions are those associated with energy purchases. Indirect emissions in the value chain are the source of scope 3 emissions. Scope 3 emissions in this study refer to emissions generated during the manufacture of fertilisers and other agricultural chemicals as well as from transport of farm inputs. The Scope 1, 2 and 3 emissions in this study were calculated using the framework for greenhouse gas accounting created by the Primary Industry Climate Change Centre (PICCC) (Lopez et al. 2022, 2023). Within those three categories of emissions, this study accounts for three gases: Carbon Dioxide (CO₂), Methane (CH₄) and Nitrous Oxide (N₂O).

In the natural environment, lime dissolves over a few years, depending on the soil’s acidity, and thereby produces carbon dioxide. However, the National Inventory on emissions and the Intergovernmental Panel on Climate Change (IPCC) assumes all emissions associated with lime in agriculture are accounted for in the year of application (Barton et al. 2014; DCCEEW 2023). This study reports the impact of liming on emissions over 30 years and uses the total emissions over that period when comparing the various land use sequences and liming strategies. This study has not considered any abatement cost of emissions, so land use gross margins are not affected by any assumed cost of emission abatement. However, gross margins are affected by the likely gradual change in soil pH that is associated with the practice or absence of liming.

In summary, this study employs the IPCC and National Inventory assumption that all emissions associated with lime in agriculture are accounted for in the year of application. We realise that in practice those emissions would in fact be spread over several years. However, by aggregating emissions over a 30 year period the sum of assumed pulses of emissions will be very similar to the sum of a pattern of more graduated emissions. Importantly, the calculation of land use gross margins does reflect the graduated change in soil pH associated with liming. Soil organic carbon sequestration is not included in this study. We link changes in crop yields to changes in soil pH, although rainfall, crop sequence and soil type are also key yield factors considered in this study.

Lime scenarios

The simulation model assumes that the soil pH across the soil profile (from surface to 30 cm) is 4.5. This is considered a degraded and acidified profile. In the analysis a baseline and three other scenarios are assessed:

Fig. 2 illustrates how and when liming occurs under the various scenarios, using the location, York, and its rotation sequences, as an example. Over the 30 years, in various years under each scenario for the initial wheat:barley:canola (WBC) rotation, liming applications are triggered by reductions in yields of the various crop phases caused by acidification of the soil profile.

Fig. 2.

Comparison of the Baseline and Scenarios 1, 2 and 3, for the location of York, considering an initial 100% cropping rotation based on Wheat (W), Barley (B) and Canola (C). Due to soil acidification, the baseline rotation shifts to a rotation that includes Lupin (L), and then finally transitions to volunteer pasture (VP). Scenarios 1, 2 and 3 show the tonnage of lime applied in two ways: a surface application (s) or a deep incorporation application (d).

Rotations

The crop rotations assessed at the 14 locations are listed in Table 1. The rotation selected at each location was the rotation in common practice at that location, according to industry experts (Harries et al. (2015) and Planfarm (2016–2020)).

Liming is profitable at almost all locations and for all rotations considered

Fig. 3 shows the net present value of gross margin differences across the 30 years between each scenario and the baseline (no lime) at each location. A 7% discount rate was applied to the calculations of net present values. For 100% cropping rotations, across all the locations, an economic incentive to apply lime, irrespective of the scenario, was revealed. Moreover, the economic incentive was greater in higher rainfall locations compared to low rainfall locations.

Fig. 3.

Net present value (NPV) of the net margin difference between the three scenarios analysed versus the baseline at each of the 14 locations. The NPV of the net margin of the baseline scenario, i.e. no lime, was subtracted from the NPV of each scenario and the box and whisker present the difference. Data are presented as box and whisker charts, in which the centre value is the median. The start and finish of the whiskers are the 25th and 75th percentiles. (a) 100% cropping, (b) 75% cropping, and (c) 50% cropping.

For rotations that featured less emphasis on cropping (i.e. 75% and 50% cropping), the incentive to apply lime generally remained. Only at a couple of locations was there little incentive to apply lime. For instance, at Merredin with 50% cropping, there was virtually no incentive to apply lime under any of the scenarios.

At locations with higher rainfall, there was a much wider spread in the net present value (NPV) of gross margin differences across the 30 years between each analysed scenario and the baseline (no lime). For some farmers, this variation in the commercial attractiveness of liming could complicate their liming choices. Nonetheless, the positive and large expected returns to liming, mostly in high and medium rainfall locations where crop-dominant land use occurs, suggest that liming is a commercially sensible business choice. Moreover, the much higher net present values from liming applications at higher rainfall locations indicated liming helped protect the high values of farmland at those locations; noting that farmland valuations are mostly conditional on the magnitude and durability of profit streams from agriculture.

At low rainfall locations, the NPVs of the per-hectare net margin differences between rotations inclusive of liming versus the absence of liming were small. These results suggest that liming is far less commercially attractive at these low rainfall locations. Over 30 years, the average NPV of benefits provided from liming was AUD 280 $/ha in Merredin (low rainfall) for 100% cropping whereas it was approximately AUD 1468 $/ha in Katanning (high rainfall). However, unlike the situation at high rainfall locations, in low rainfall locations farm sizes are often large (Fig. 4). Accordingly, when small positive per-hectare margins are multiplied by thousands of hectares, as is the case at several low rainfall locations, then the aggregate effect on farm business finances could be large. As shown in Fig. 4 there can be a three or four-fold difference in farm size between low and high-rainfall locations (e.g. Kalannie versus Katanning), causing small per-hectare commercial advantages to translate into sizeable business returns.

Fig. 4.

Average farm size at each location.

The results in Fig. 3 highlight the importance of understanding the framing of a farmer’s liming decision. Is it simply a paddock-based decision where the focus is solely on returns from liming in that paddock; or is the liming decision one that can be replicated across the farm and therefore it is the aggregate impact of liming on the entire farm business that needs to be stressed? As farms are not only agricultural systems but also business entities, our view is that the impact on the entire business should be given precedence rather than simply highlighting a per-hectare impact.

The source of the additional profit generated by liming is worth emphasising. In the absence of liming many current farm practices gradually worsen soil acidity and eventually restrict land use to less profitable grazing options. Liming, by lessening the productivity restrictions imposed by soil acidity, enables a wider and more prolonged profitable sequence of land uses. Hence, liming is a profitable and productivity-enhancing soil management practice in almost all situations where problematic soil acidity is a feature or is poised to soon become a problem.

Liming increases emissions, with some exceptions

Fig. 5 shows the summation across the 30 years of the difference in emissions between the three scenarios analysed and the baseline, at each of the 14 locations for the three intensities of cropping. At many locations, irrespective of the initial land use strategy (i.e. 50%, 75% or 100% cropping), including lime in the rotations increased emissions. However, at some locations given some land use strategies, reductions in emissions occurred when liming was practised.

Fig. 5.

Summation across the 30 years of the difference in greenhouse gas (GHG) emissions between the three scenarios analysed and the baseline for each location at different intensities of cropping. (a) 100% cropping, (b) 75% cropping, and (c) 50% cropping. CO₂e, CO₂ equivalents.

For example, at some northerly locations with deep sands (Northampton, Moora), their relatively high rates of application of nitrogenous fertilisers and low soil carbon content led to their high rates of soil acidification. In the absence of liming, soil acidification eventually led to an inability to profitably grow canola and subsequently wheat, forcing the introduction of volunteer pasture for sheep production. The upshot of the transition into sheep production caused an increase in emissions. Hence, over the 30-year period of land use at these sites, the overall emissions from land use with liming were less than emissions in the absence of liming.

Similarly, at southern sites like Katanning and Kojonup with their higher rainfall, in cases where their farming systems might already be crop dominant (i.e. 75% and 100% cropping), if soil acidification was not prevented via liming, then eventually those initially crop-dominant systems became less viable. Absence of liming at these locations caused a gradual reduction in yields of canola and then wheat, forcing a land use change into sheep production that then generated more emissions. Liming, by preventing a land use change towards sheep production, caused lesser emissions when crop-dominant land use was able to be maintained.

However, at the southern sites, Katanning and Kojonup, in cases where 50% cropping occurred, the higher rainfall firstly supported sheep enterprises with higher stocking rates yet secondly caused more nitrogen to be fixed during pasture phase, thereby reducing the need for nitrogen fertiliser and hence slowing the rate of acidification. The overall impact was that at 50% cropping, the average emissions associated with liming were greater than emissions in the absence of liming, despite the higher stocking rate.

Overall, at many locations, and especially with crop-dominant land use sequences (i.e. 75% and 100% cropping) at medium and low rainfall locations, lime applications generated more emissions over the 30 years of analysis. By illustration, at Cunderdin (medium rainfall) and Merredin (low rainfall), lime strategies generated an average of 1.82 and 1.07 t CO₂e/ha respectively more emissions over 30 years compared to no lime at 100% cropping and 2.64 and 1.46 t CO₂e/ha more emissions at 75% cropping. At medium and low rainfall locations often rates of application of nitrogenous fertilisers were lower and rates of soil acidification were lower. Hence at these locations the forced alteration of land use due to acidification towards sheep production was far less likely and so on average liming increased emissions.

Liming allows greater land use flexibility

Barley, canola and lupin are major break crops in Western Australia’s wheat-dominant farming systems that characterise much of the study region (Seymour et al. 2012; Harries et al. 2015). Pulses like faba beans, chickpeas and field peas play much smaller roles (Paterson and Wilkinson 2023) as break crops. All these traditional break crops are sensitive to soil acidity, with aluminium toxicity being a major limitation on crop yield (French and White 2005; Wang et al. 2006; Lofton et al. 2010). Under acidic conditions choice of break crop becomes problematic, forcing farmers to apply higher rates of herbicides and fungicides as the break crop option for pest management becomes limited. The increase in chemical use increases both farm costs and emissions. Furthermore, the limited choices affect the farmer’s ability to respond to market and weather signals.

Soil acidity not only limits farmers’ crop choices, but it also decreases crop competitiveness against weeds. The survival rate of weeds increases under acidic conditions and crop herbicide efficiency decreases (Hashem and Borger 2018; Borger et al. 2020). Hence, liming a soil can increase the competitiveness of crops and decrease weed burdens, enabling more effective weed control (Gazey et al. 2014).

Although legume crops like lupins are not specifically considered in this study, it is worth noting that the ability to reduce farm emissions is aided whenever legume crops can be profitably included in land use sequences, as these legume crops sequester nitrogen for following cereals (d’Abbadie et al. 2024b) and lessen reliance of applications of nitrogen fertilisers with all their associated emissions generated during their manufacture and transport. Acidity in a soil reduces the chances of rhizobium inoculation that forms nitrogen-fixing nodules on legume crops, especially when the pH is less than 5.5 CaCl₂, hence reducing the nitrogen available for a following crop (GRDC 2022). Poor nodulation causes increases in fertiliser applications in cereal phases that follow legume crops. Furthermore, whenever plant growth is limited by acidity then resulting ground cover can also reduce, thereby increasing the risk of soil loss or degradation from wind or water erosion. Loss of topsoil can generate expensive long-lasting impacts. However, in this analysis we ignore this additional environmental benefit (i.e. wind erosion costs avoided) of liming.

Caveats

Soil systems are inherently complex, characterised by intricate bio-physical-chemical interactions that surpass the scope of this paper’s models. The primary aim of this study is not to model the biogeochemical carbon cycle but to assess the economic aspects of liming and greenhouse gas (GHG) accounting. Specifically, this work evaluates the effect of lime dissolution on soil pH over several years and its subsequent impact on crop yield, while also considering aggregate emissions. Additionally, it is important to note that CO₂ release in biological systems can occur through various pathways (Page et al. 2009). However, this study assumes that all emissions originate directly from the amended soil.