1. Introduction
Given climate change and attention to environmental issues observed in recent years, international food supply chains are increasingly focused on sustainability and reducing environmental impact. In this context, calculating greenhouse gas (GHG) emissions in agricultural production systems and determining possible mitigation points for these gases is paramount [
1].
Therefore, there is a need to mitigate the environmental impact of products and make food production more sustainable [
2]. Studies on the environmental footprints of agricultural products have become essential tools to assess and improve the efficiency of these chains, as they encompass important environmental impact indicators such as the carbon footprint (CF) [
3].
The reduction of CF is attributed to agriculture, which facilitates reaching internationally established targets to combat climate change by reducing GHG emissions, especially when these studies follow international standards established by the International Standard Organization (ISO), the Intergovernmental Panel on Climate Change (IPCC) and sustainable development goals (SDGs).
Several methodologies have been developed and can be used to calculate GHG emissions related to agricultural production chains. Life cycle assessment (LCA) studies focused on CF cover impact categories of global interest, such as climate change [
4,
5].
LCA is a methodology used to investigate the potential environmental impacts of producing a product or service according to ISO 14040, 14044 [
6,
7]. ISO 14067 [
8] details the methodological procedures for conducting CF studies. These studies allow identifying the origin of a product’s main impacts, thus enabling the development of agricultural activities with greater environmental sustainability.
Thus, it is essential to use methodologies for quantifying direct GHG emissions from agricultural production. To improve the efficiency of calculating the emissions mentioned above and consequently, the evaluation of the CF of agricultural products, operational tools based on different methodological procedures, and/or characterization models are being developed worldwide.
Recently, in Brazil, Embrapa Environment released ICVCalc, a tool designed for building agricultural inventories for LCA studies [
9]. This tool consists of a set of spreadsheets that estimate emissions using various methodological protocols, thereby generating comprehensive process inventories. The methodologies used are internationally recognized, such as Nemecek-Calc [
10], World Food LCA Database (WFLDB) [
11], IPCC-Calc [
12], BR-Calc [
9], and Agri-footprint [
13].
Thus, the BR-Calc methodology is an initiative aimed at addressing the need to develop tools that use characterization factors adapted to the reality of tropical climate countries, enabling more accurate estimates of GHG emissions under these conditions. Comparison with international methods can help identify areas for improvement in BR-Calc, such as incorporating best practices from other methodologies or adjusting to better reflect Brazilian conditions. Since concerns about climate change and sustainability are global, emission estimation methodologies must be internationally comparable.
However, studies evaluating and comparing the estimation of GHG emissions and the impact on the CF of agricultural products are incipient in the literature, especially when it comes to studies conducted for tropical climate regions and crops, such as Brazil and the cashew crop.
Additionally, while various methodologies are available, few studies provide a systematic comparison of their applicability to tropical conditions [
14]. As a result, there is still a limited understanding of how different empirical equation-based methods perform in estimating GHG emissions for cashew production in Brazil, considering regional data availability, scientific robustness, and methodological transparency. This gap in the literature justifies the need for a comprehensive assessment of different GHG quantification methods applied to cashew production, ensuring that the chosen approach accurately represents the environmental impacts in this specific context.
The novelty of this manuscript lies in its focus on providing a comprehensive comparison of different methodologies for GHG emissions estimation, specifically applied to a tropical agricultural system. By considering regionally adapted approaches and a crop that has great economic importance in Brazil, this study fills a crucial gap in the literature, offering insights that are directly relevant to decision-making for sustainable agricultural practices.
Thus, the objective of this study was to evaluate and compare methods for quantifying GHG emissions based on empirical equations for life cycle inventories, using the conventional cashew production system in Brazil as a case study.
5. Discussion
The evaluation of GHG emissions quantification methodologies based on criteria analysis revealed significant differences between the methods analyzed, reflecting variations in clarity, scientific robustness, and consideration of regional data. The IPCC-Calc [
12], BR-Calc [
9], and WFLDB [
11] methods achieved the highest scores due to a balance between these criteria, while Agri-footprint obtained an intermediate score and Nemecek-Calc presented the lowest overall performance.
IPCC-Calc [
12] stood out for its high clarity, as its guidelines are well-established and extensively documented. In addition, its scientific robustness is widely recognized, being one of the most widely used methods for national emissions inventories. However, its consideration of regional data is limited, since the emission factors are global and require calibration for local conditions, which may compromise its accuracy for tropical systems such as cashew cultivation in Brazil.
BR-Calc [
2,
9], presented an equally high score, standing out mainly in the consideration of regional data, as it was developed specifically for edaphyclimatic conditions in Brazil. Its clarity was also well evaluated, being a potentially more accessible model for the national context. In addition, its scientific robustness was considered high, as it is based on local data and specific adaptations for the Brazilian agricultural reality, making it a relevant option for analyses in tropical conditions. In addition, this methodology has been reviewed by experts in the field.
WFLDB [
11] also obtained a high score due to its detailed documentation, which ensures satisfactory clarity. Its scientific robustness was evaluated as high since it is widely used in LCA of agri-food production. However, its consideration of regional data is only average, since, although it has broader information than Nemecek-Calc [
10], it is not fully adjusted to tropical conditions, requiring additional adaptations.
Agri-footprint obtained an intermediate performance, being a method used in agri-food LCA and with good scientific robustness. However, its clarity was evaluated as average, since the documentation can vary depending on the version used. Furthermore, its consideration of regional data was also intermediate, like the WFLDB [
11], indicating the need for adjustments to better represent agricultural conditions in Brazil.
On the other hand, Nemecek-Calc [
10] had the lowest overall score, mainly due to its low transparency in calculations. Despite being one of the most reliable databases in LCA and having high scientific robustness, its clarity was considered only average, as its methodologies and emission factors are unclear. Furthermore, its consideration of regional data is limited, as the emission factors are based on temperate conditions, requiring significant adjustments to be applied in tropical regions.
Thus, the results reinforce that the choice of the most appropriate method should consider a balance between clarity, scientific robustness, and regional adaptation. While IPCC-Calc [
12], BR-Calc [
9], and WFLDB [
11] proved to be the most recommended options, Agri-footprint can be useful depending on the context of the analysis, and Nemecek-Calc [
10] requires greater caution due to its limitations in transparency and adaptation to tropical conditions.
The comparison highlights substantial variations in emission estimates produced by each methodology. These discrepancies underscore the influence of methodological choices on the results obtained, consequently affecting the projected impact of emissions on climate change. This emphasizes the importance of selecting appropriate methods for accurate assessments.
Furthermore, these limitations may impact the assessment of the carbon footprint of the analyzed system. Methods with less transparency in calculations, such as Nemecek-Calc, may hinder the traceability of estimates, generating uncertainty in the quantification of emissions, and compromising the reliability of the LCA. Although they do not adequately consider regional data in their standard approach and use global emission factors that do not reflect the particularities of the cashew production system in Brazil, IPCC-Calc [
12] and WFLDB [
11] presented satisfactory results in terms of carbon footprint. These are like those obtained in BR-Calc, which, although it has regionalized factors for Brazil, for this study, there was no significant influence from them.
Thus, the choice of methodology directly influences decision-making, reinforcing the need to select approaches that offer greater data representativeness, transparency in calculations, and adaptation to the context analyzed.
To explain differences in results, we compared equations and emission factors adopted in each methodology (
Table 7 and
Table 8).
CO
2 emissions (
Table 4) are related to the use of urea, limestone (CaCO
3), and dolomite ((CaMg)CO
3). Urea application leads to the release of CO
2 previously fixed in the fertilizer molecule. This occurs due to urea hydrolysis, where the carbon in its structure is converted into CO
2. Since approximately 20% of urea’s mass consists of carbon, applying the stoichiometric conversion to CO
2 (44/12) results in an estimated emission of CO
2 per kg of urea. On the other hand, the carbonates in limestone and dolomite decompose in the soil, releasing CO
2. In the case of limestone, about 12% of its mass is carbon, while in dolomite, this proportion is approximately 13%.
The IPCC-Calc [
12] and BR-Calc [
9] methodologies produced identical results as they accounted for urea and limestone emissions, using the same emission factors (0.12 for limestone and 0.20 for urea) and additionally included emissions from fossil fuel (diesel) combustion. The Nemecek-Calc [
10] estimated lower emissions because it did not consider limestone and dolomite emissions, whereas the WFLDB [
11] reported higher values due to the full accounting of carbon emissions from these inputs. The Agri-footprint methodology [
13], despite using the same emission factors for urea and limestone, did not include diesel combustion emissions (
Table 7), leading to different results from those of IPCC-Calc and BR-Calc [
9,
12]. Consequently, the inclusion of diesel emissions in IPCC-Calc and BR-Calc [
9,
12] significantly increased CO
2 emissions and, therefore, the carbon footprint (
Figure 2).
N
2O emissions were estimated in all methodologies considering direct and indirect emissions, as proposed by based on [
24]. Direct N
2O emissions are from the use of mineral and organic fertilizers, decomposition of crop residues, and mineralization associated with loss of soil organic matter resulting from changes in land use or management of mineral soils.
Indirect emissions regard the conversion of NH
3, NOx, and NO
3⁻ into N
2O. However, differences were found in the methodologies regarding emission sources and factors. The lowest emission of N
2O was obtained using the methodology of Nemecek-Calc [
10] due to disregarding the mineralization of soil organic matter (Nsom) and the indirect emission from NOx. WFLDB [
11] adopts equations with modifications and factors considering direct emissions from fertilizer applications (mineral and organic), crop residues, and decomposition of soil organic matter, as well as indirect emissions from NH3 volatilization, NO
3 leaching, and NOx emissions. On the other hand, the highest N
2O values were from IPCC-Calc and BR-Calc [
9,
12], justified by the greater specificity of the equation (Tier 2) for N
2O emissions and by the lack of both considering nitrous oxide emissions from diesel combustion [
9,
12].
Considering the carbon footprint results obtained from each methodology, IPCC-Calc and BR-Calc [
9,
12] showed the highest impacts. This is because, in addition to fossil CO
2 and N
2O emissions, these methodologies account for additional GHG emissions resulting from diesel combustion, a criterion not evaluated by the other methodologies. Even without evaluating emissions from burning diesel, the WFLDB methodology [
11] showed a lower level of only 9.8% in relation to the greater impact, justifying the fact that this method uses a greater emission factor for urea, directly contributing to the impact on climate change.
Nitrogen fertilization directly contributes to the increase in the carbon footprint, where urea, the main fertilizer source of N, increases CO
2 emissions due to its synthesis from ammonia and CO
2, increasing GHG emissions by about 40% due to its high carbon content [
25].
BR-Calc [
9] is inserted as a regional alternative adapted to Brazilian conditions, seeking to fill gaps left by global methodologies such as the IPCC [
12] and more detailed models, such as the WFLDB [
11], incorporating emission factors and parameters adjusted to climatic, soil, and agricultural management particularities in Brazil, making it especially relevant for local and regional studies [
2]. It is also noteworthy that BR-Calc [
9] also calculates emissions other than GHG-related ones, such as heavy metals and phosphate emissions.
When the methodologies of Nemecek-Calc [
10] and Agri-footprint [
13] were used for the quantification of GHGs, lower values of environmental impact were obtained in the climate change category. This result occurred mainly due to the simplification of calculation present in these methodologies, especially in Nemecek-Calc [
10], which presented underestimated results for most of the quantified emissions (
Figure 2).
Currently, the most frequently used methodologies for estimating GHG emissions from agricultural systems in LCA studies are those adopted by the IPCC-Calc [
12] and the WFLDB [
11], especially the joint use of both, according to [
26,
27].
It is important to stress that more specific models such as BR-Calc [
9] and Agri-footprint [
13], although they allow the estimation of emissions within a limited time and with a small financial budget, require knowledge of local ecosystems to calibrate specific variables for the correct modeling [
4].
Based on the IPCC Guidelines [
12], the IPCC-Calc quantifies only greenhouse gas emissions, considering N
2O from fertilizers and crop residues and CO
2 from limestone and urea application. However, this methodology has limitations that may affect comparisons with other methods, as it does not include the indirect formation of NO
2, the radiative forcing of NO
x, water vapor, and sulfates, nor the indirect effects proposed by the IPCC, which may underestimate climate impacts. Additionally, it does not account for the conversion of CO-to-CO
2, a relevant factor under certain agricultural conditions [
28].
The BR-Calc [
9] corresponds to a protocol developed by Embrapa Environment, which compiles methods consolidated in the literature for estimating emissions from agricultural activities, selected for better representing tropical conditions. Additionally, the tool includes a climate and soil database specific to Brazil, allowing users to choose the most suitable geographical scale for their study. However, this protocol uses the same equations and emission factors as the IPCC-Calc for GHG emission quantification, resulting in little mathematical differentiation between the methods. Therefore, future studies should focus on developing GHG emission factors adapted to Brazil’s edaphoclimatic conditions to improve estimation accuracy.
In the Nemecek-Calc [
10] methodology, a significant underestimation of GHG emissions is observed, mainly due to the absence of accounting for fossil CO
2 from limestone use, nitrogen mineralization in mineral soils, and the exclusion of indirect emissions of (NO
x) which, according to the authors, correspond to 21% of N
2O emissions. Additionally, this methodology features poorly explained equations and factors that are difficult to understand, compromising the clarity and transparency of the calculations.
The WFLDB [
11] methodology is an update of the method by Nemecek-Calc [
10], correcting most of the issues of its predecessor by providing greater data transparency and enhanced scientific robustness. However, in both methodologies, GHG emissions from fossil fuel combustion are not accounted for, which limits the accuracy of the estimates.
In the literature, it is observed that most of the studies that evaluate GHG emissions resulting from agricultural production in tropical climate conditions use the IPCC-Calc and WFLDB methodologies together, especially when evaluating the emission of gases that indirectly contribute to the greenhouse effect (NO
x, NH
3, and NO
3), as reported in coconut [
3], cashew [
4], melon [
5], macauba [
29], and soybean [
28] crops.
More recently, in the context of Brazilian agriculture, the integrated use of the BR-Calc methodology with the methods mentioned above has been observed, as evidenced by wheat [
2] and banana [
26] crops. However, although all these studies use more than one methodology to assess emissions, they do not compare the different results that each method provides.
Table 9 presents summaries of the main advantages and limitations of all methodologies evaluated in this study, providing a direct comparison between the methods and offering insights into their applicability in different contexts.
The results of this study can guide public policies and sustainable agricultural practices in cashew cultivation. Methods such as IPCC-Calc and BR-Calc, which provide more accurate estimates, allow the search for more effective actions to reduce GHG emissions from cashew cultivation in tropical environments.
For farmers, it is recommended to optimize the use of fertilizers [
1], adopt organic and green fertilizers [
4,
30], and implement agroforestry systems [
31], reducing N
2O emissions and improving the carbon balance in the soil. For policymakers, it is essential to encourage methodologies and emission factors adapted to the Brazilian reality, in addition to promoting research and technologies to strengthen low-carbon agriculture, according to the ABC+ plan [
32].
These strategies are aligned with SDGs [
19]: 2 (zero hunger and sustainable agriculture); 12 (responsible consumption and production); and 13 (climate change mitigation). Thus, the study not only improves the quantification of GHG emissions but also offers guidelines to make cashew production more sustainable.
6. Conclusions
This study contributes to the advancement of LCA and CF in tropical regions by addressing the gap in methodological comparisons highlighted in the work. By evaluating different approaches to quantifying GHG in a tropical agricultural system, our results highlight the importance of choosing methodologies that can be adapted to different agricultural scenarios.
The choice methodology for estimating GHG emissions is a critical factor, as it directly influences the interpretation of agricultural carbon footprints. In this study, the IPCC-Calc and BR-Calc methodologies yielded the highest emission estimates for cashew production under tropical conditions, leading to the largest carbon footprints. In contrast, the Nemecek-Calc methodology reported the lowest values, highlighting the variability among methods and the need for regionally adapted approaches.
Based on the analysis criteria, the WFLDB, IPCC-Calc, and BR-Calc methodologies were the most recommended, balancing clarity, scientific robustness, and regional adaptation. BR-Calc stood out for its consideration of regional data, while WFLDB and IPCC-Calc demonstrated greater scientific robustness. In contrast, Nemecek-Calc showed limitations in transparency, and Agri-footprint had an intermediate performance, proving useful depending on the analysis objective.
Therefore, it is suggested that BR-Calc be used for evaluations in national contexts, especially in situations where factors such as the use of limestone, urea, and local agricultural practices have a significant impact on GHG emissions. This approach can improve the representativeness of the results and provide more precise support for decision-making aimed at mitigating emissions in Brazilian agricultural production.
Furthermore, future studies should consider the integration of primary data and direct measurements to validate the estimates obtained through different methodologies, reducing uncertainties and enhancing the applicability of models in tropical agricultural systems. The adoption of hybrid approaches, combining empirical equations with process-based modeling, could represent a promising alternative to increase the accuracy of estimates and assist in the development of more effective policies for mitigating GHG emissions in agriculture.