In the first part of the study, publications were searched on the Web of Science® platform. The query string was defined considering terms related to SOM dynamics in different agroecosystems, such as annual cropping systems, pasture, and integrated agricultural systems, as well as diverse management practices (e.g., tillage) that may occur within these systems. The search considered terms mentioned in each record’s “title, abstract, and keywords” (Topic Field), and it was limited to peer-reviewed articles, as they represent most high-quality published documents. The query string used was as follows:

The search was restricted to the American region (countries from Canada to Argentina, including the Caribbean) using the “countries/region” filtering tool available on the platform. The obtained documents were manually analyzed to quantify (i) how many were conducted in the region and (ii) how many assessed soil C stocks. Studies meeting both selection criteria that covered the first 30 cm soil profile depth were identified. This information was used to build up a spreadsheet accounting for the number of studies that measured C stocks in the continent.

The literature review suggests that, despite many documents retrieved, only a little information is available on C sequestration in agroecosystems of the Americas. From more than 13 thousand papers considered initially for this study, only around 1,189 documents (<9.2%) presented data on C stocks in the soil for the three evaluated systems (Figure 1); and the number was even lower for those reporting the data to 30 cm depth, i.e., 598 documents or 4.6% (Table 2). The bibliometric analysis extracted from the literature review revealed a set of 100 most used keywords, distributed in 5 groups, as shown in the keyword network co-occurrence diagram (Supplementary Tables S1, S2). When describing the studies, the authors used the terms and topics including soil organic matter, soil organic carbon, soil carbon, carbon sequestration, land use change, climate change, no-tillage, carbon dioxide, nitrogen, and phosphorus with greater frequency and intensity. Therefore, these terms are crucial and central to the topic addressed.

Due to limited data availability for conducting a meta-analysis, specific studies were identified within our database that provided the C sequestration rates needed to perform further estimates for the studied region. More specifically, for regions where meta-analysis are available in the literature (e.g., North and South America), C sequestration rates obtained in these studies were prioritized (Conant et al., 2017; Frazão et al., 2023; Nicoloso and Rice, 2021; Viglizzo et al., 2019; Maia et al., 2022). For regions where meta-analysis were not available, and the identified studies were not sufficient to proceed with our originally chosen approach, C sequestration rates reported in specific (individual) studies were adopted (Table 1). It was found that the highest rates of soil C sequestration are obtained in integrated agricultural systems (IAS), especially in the Central and Andean regions (Table 1). Accumulation rates (Mg ha⁻¹ year⁻¹) for IAS ranged from 0.79 to 1.16 for the five regions, 0.25 to 1.03 for improved pastures (IP), and 0.42 to 0.67 for conservation agriculture (CA). The average rates of SOC accumulation (Mg ha⁻¹ year⁻¹) for the three selected BMPs were 0.50 (IP), 0.46 (CA), and 0.97 (IAS). Given the lack of data for the Caribbean region, rates for this region were assumed to be the same as those for the Central region, considering the criteria of the same geographic position (latitude) and climate classes (Supplementary Figure S2). Furthermore, the values of the C sequestration rates were extracted from the documents which either contained these data, or provided the necessary information (e.g., timespan of management) to calculate the rates. The Supplementary Figure S4 illustrates the geographic location of the studies used to derive the soil C sequestration rates.

Upscaling BMPs over 334 Mha in the Americas (30% of the land dedicated to agriculture and livestock, presumed degraded for the region) would sequester up to 5.2 (±2.9) Pg CO₂eq by 2030 (8-year period from 2022 to 2030) and would present an overall potential of 13.1 (±7.1) Pg CO₂eq over 20 years (Figure 2). The northern and southern regions had the greatest potential for soil C sequestration, contributing to 38.9 and 38.6% of the total C sink capacity, respectively. The highest potential of soil C sequestration in the southern region was observed in Brazil, comprising 3.07 Pg CO₂eq over 20 years (60.4% of the total potential for the southern region; Supplementary Table S2). In the northern region, the USA has the greatest C sink capacity, with a potential of 3.7 Pg CO₂eq (72.3% of the total capacity of the region). Together, Brazil and USA represent 51.5% of the total capacity of the Americas (13.1 Pg CO₂eq; Supplementary Table S2). The Andean region has the third largest potential (20.6%), followed by the central (1.4%) and the Caribbean regions (0.5%; Figure 2).

There is little information regarding the status of soil degradation in the Americas. However, the few information available indicates that around 30% of the land is degraded in the region, which could be restored and increase soil C sequestration over time (FAO, 2015; Gardi et al., 2013; UNCCD, 2019). In such case, the adoption of BMP in those 30% of the land dedicated with agricultural practices (crops and livestock, i.e., 334 Mha) would offset 39.2% (i.e., 0.66 Pg CO₂eq year⁻¹) of the annual agricultural GHG emissions (i.e., 1.68 Pg CO₂eq year⁻¹) over 20 years (Figure 3). For a complete offset of annual emissions, however, it would be necessary to produce with the selected BMPs in 915 Mha (Figure 3), corresponding to 82.4% of the total agricultural area (i.e., 1.11 B ha, per the World Bank, 2023). The south and north regions have the greatest need for expansion of the BMP to achieve neutrality of total GHG emissions from agriculture, i.e., 323 and 507 Mha, respectively (Figure 4). On the other hand, although the Caribbean and the Central regions require a smaller area (11 and 23 Mha, respectively) to mitigate the sector’s emissions, these exceed the agricultural area available in these regions by 30 and 74%, respectively. Among all regions, the Andean region requires the least area to completely offset emissions from agriculture, needing about 36% of the available agricultural area (50.5 Mha; Supplementary Table S2), exceeding by only 6% of the baseline value adopted in this study (30% of the total agricultural area) as an available area for upscaling of BMPs. In contrast, the northern and southern regions need approximately 90 and 83%, respectively. For a short-term mitigation scenario, with 50% GHG reductions by 2030, the Andean region is the only one that could reach the target without exceeding the reference area (Figure 3; Supplementary Table S2). It is worth noting that the higher the level of adoption of BMPs in the region, the shorter would be the time needed to achieve net-zero emissions targets, as the need to scale up the selected technologies depends on the current level in the region.

The systematic literature review showed that despite the large number of studies retrieved in the search, only a small fraction performed field measurements for soil C data evaluation, mostly conducted in the north and south regions (Table 2). This discrepancy suggests that despite SOM-related terms being systematically mentioned in scientific research articles, the fraction of studies that measure it in the field is limited and does not present a good coverage of the whole continent. Indeed, the agricultural area in the Americas is largely represented by the South and North regions (952 Mha; World Bank, 2023), which helps to justify the high concentration of studies in these areas. Nevertheless, despite the lowest individual contribution, Central, Caribbean, and Andean regions sum together 161 Mha of agricultural land (World Bank, 2023), therefore accounting for a significant land proportion. Furthermore, the variability for soil, climate, and agroecosystems management encountered in these regions are considerable high (Alavi-Murillo et al., 2022; Owens et al., 2020), which requires an appropriate sampling density for properly identifying the impact of different management/climate conditions on soil C dynamics.

In addition to the reduced number of studies identified, lack of sampling standardization was also a major limitation observed in the dataset, since only 4.6% (out of the 13 thousand studies retrieved in the search) of the studies measured soil C up to 30 cm soil deep. The IPCC recommendations for soil C evaluation suggests a minimum sampling depth considering the first 30 cm of the soil (IPCC, 2006), since this layer is frequently the most affected one under cropping systems areas and is predicted to account for ~45% of total soil C stocks found in the first 1 m of soil (1,505 Pg of C; Lal, 2018). Moreover, new studies have recently suggested more detailed approaches, emphasizing the need for sampling deeper soil layers (Button et al., 2022; Olson and Al-Kaisi, 2015). The need for more detailed sampling protocols relies on the fact that studies showed that cropping management can affect soil C in deeper layers (Olson and Al-Kaisi, 2015). In the same line, C sequestration in the sub-soil layers have been advocated as a promising strategy for long-term soil C stabilization (Button et al., 2022); since deeper soil layer have a reduced microbial activity and a higher C saturation deficit, therefore favoring C binding onto mineral structures and protection against microbial oxidation (Chenu et al., 2019; Sokol and Bradford, 2019).

The search also suggested that a reduced number of studies focused on BMPs, as indicated by the co-occurrence map of the keywords. The most cited terms were related to SOM and overall processes such as “land use change,” suggesting that this is a largely studied subject. Apart from that, only “no-tillage,” which is a BMP, had a significant contribution/interaction in the map. Other important keywords such as “agroforestry” and “cover crops” appeared but indicated to have a small contribution. This highlights that despite the wide recognition that these practices are great strategies for promoting soil C accrual and restore soil health (Jian et al., 2020; Lange et al., 2015; Nwaogu and Cherubin, 2024; Steinfeld et al., 2023; Yang et al., 2024), the number of research in these topics is still limited. Unlocking the potential of the Americas for improving the sustainability of its agricultural areas depends on obtaining high-quality field data for soil C sequestration strategies. Best management practices (e.g., cover cropping, crop diversification) are suggested to be a key-tool for achieving these goals, and the literature review suggests that this is a path that still needs to be paved.

The calculations exercises present a first attempt to estimate both the capacity of agricultural land from the Americas to sequester C and the necessary area needed to completely offset GHG emissions using only three nature-based solutions or technologies: IP, CA, and IAS. The estimated soil C sequestration potential (13.1 ± 7.1 Pg CO₂eq) supports the previous studies (Lal, 2004, 2003; Minasny et al., 2017; Paustian et al., 2016; Smith, 2008), which showed that BMPs are efficient strategies of promoting soil C sequestration and are a potential tool to mitigate the anthropogenic climate change. Adoption of BMPs could sequester up to 8 Pg CO₂eq year⁻¹ globally (Paustian et al., 2016) and have a large positive impact on GHG mitigation (IPCC, 2022). Other researchers have estimated a mitigation capacity of 1.5 to 5.3 Pg CO₂eq year⁻¹ (Smith, 2008). The estimate presented herein suggests that up to 1.02 (0.66 ± 0.36) Pg CO₂eq year⁻¹ would be mitigated in the Americas considering only one-third of the available agricultural area, which represents roughly 12 to 44% of the global estimated capacity according to Smith et al. (2008), and ~ 8% according to Paustian et al. (2016). The given scenario places the agri-food systems of the Americas as a major component to mitigate climate change as well as farmers as a key actor to enhance global food security.

The greatest potential of soil C sequestration observed in southern and northern regions depends on the available agricultural area in the countries of these regions. For the Andean region, however, despite the lower contributions compared to southern and northern regions, the reviewed rates for soil C sequestration are considerably high, especially for the practice of IAS, which also presented high values for this modality (1.16 Mg ha⁻¹ year⁻¹) in the central region. This highlights the potential for soil C sequestration for this region when considering each area unit. In practical terms, these higher rates for soil C sequestration may be explained by two main reasons: (i) the Andean region is comprised of high-altitude area which contribute to promoting soil C accrual (Alavi-Murillo et al., 2022; Segnini et al., 2011); (ii) the studies recovered in the present search for the Andean region assessed mostly agroforestry systems for the IAS modality. Agroforestry systems are generally more complex when compared to other IAS studies that were considered in the search for other regions (e.g., integrated crop-livestock systems). The high complexity of these systems may contribute to increasing soil C sequestration potential, as stated by Steinfeld et al. (2023) and Lange et al. (2015). Once it promotes a greater capacity for C accrual in the soil, this feature is highly interesting in the context of C market initiatives. In general terms, the high rates for soil C sequestration increase the region’s capacity for mitigation of GHG emissions. Despite the smaller contribution of each country when compared to the other regions, Andean, Caribbean, and the Central American countries together contribute up to 23.6% of the total potential, representing 3.1 Pg CO₂eq.

The predicted potential was based on C sequestration rates reported in the literature (Table 1), where three different strategies were adopted in proportions that are more feasible to be implemented in the field, i.e., 56% of IP, 24% of CA, and 20% IAS. Due to this conservative scenario, a significant proportion of the agricultural area in the continents (i.e., 915 Mha – 82.4% of the total) would be needed to completely offset agricultural GHG emissions. Prioritizing strategies that are more efficient in promoting soil C sequestration, like IAS (Table 1), could be beneficial for reducing the required area to achieve C neutrality since it could increase the capacity for promoting CO₂ removal from the atmosphere. In the same line, combining other approaches not considered in the calculations presented here, such as biochar, could also be beneficial. Biochar has been reported to greatly contribute to soil C sequestration, besides reducing N₂O emissions and improving soil health (Beillouin et al., 2023; Kaur et al., 2023; Mukherjee and Lal, 2013). Available meta-analyses indicated increases on soil C levels ranging from 39 to 64% under biochar application (Bai et al., 2019; Beillouin et al., 2023; Chagas et al., 2022). Alongside, other conservation strategies such as reduced tillage (or minimum tillage) also can potentially contribute as a BMP strategy (Jordon et al., 2022; Krauss et al., 2017). The soil C sequestration potential calculations performed here were based on NT practice due to the great data availability (i.e., soil C sequestration rates) in the literature. Nevertheless, adopting other intermediate strategies such as minimum tillage could also be positive, especially in field conditions where management or technical restrictions prevent the application of more robust approaches such as NT.

Furthermore, due to limited data availability, these estimates considered only the potential for soil C sequestration in the soil, and do not include the aboveground C sequestration (e.g., in the trees of IASs) and the potential reduction of GHG emissions from the agricultural sector. The present study also did not include the potential positive impacts on agronomic productivity, soil health, and other benefits (Kopittke et al., 2022). The actions taken for climate change mitigation must involve multiple efforts, which would comprise strategies specifically focused on reducing GHG emissions in the productive chain of the agricultural sector, therefore enhancing the mitigation capacity (Lal, 2022). This could lead to a carbon-positive scenario, where the agri-food systems could mitigate more GHG than they generate. For example, the Global Methane Pledge, launched at COP26, in Glasgow and ratified by 30 countries of the Americas, seeks to establish volunteer actions to reduce 30% of CH₄ emissions from 2020 levels until 2030. This target could be achieved using different tools with the potential to reduce GHG emissions allied to soil C sequestration technologies (Johnson et al., 2007; Lal, 2022; Nayak et al., 2015). For example, using additives in ruminants’ diets could reduce enteric CH₄ production by ~34% (±15) (MAPA, 2020; Roque et al., 2019; van Wesemael et al., 2019). This management practice can significantly impact the mitigation of GHG emissions from the livestock sector in the Americas, where five (i.e., Brazil, USA, Argentina, Mexico, Uruguay) of the ten largest cattle herds in the world are present (i.e., ~366 million head, per Souza Filho et al., 2019).

Improved grazing management practices could also be beneficial to reducing enteric CH₄ production (Souza Filho et al., 2019), besides being an effective strategy for increasing nutrient use efficiency (NUE) (Löw et al., 2020). A greater NUE can be beneficial: (i) to increase forage/crop yield and consequently C accrual, and (ii) to reduce the surplus of nutrients in the soil solution that could be leached or converted into GHG like N₂O in the case of N (Venterea et al., 2016). Specifically, for N fertilization, inhibitors can also reduce the production of N₂O (e.g., nitrification inhibitors like dicyandiamide - DCD; Liu et al., 2013). Similarly, crop diversification strategies like crop rotation could improve nutrient cycling, consequently leading to better nutrient usage and reducing the demand for external inputs (Lehman et al., 2017). In more biodiverse cropping systems, the inclusion of cover crops associated with no-till can significantly improve soil physical and biological quality (Blanco-Canqui and Ruis, 2020; Lange et al., 2015; Williams et al., 2020). Soil with better physical conditions can increase air and water diffusion, thus reducing the occurrence of anaerobic sites that are prone to denitrification or methanogenesis processes (Blanco-Canqui and Ruis, 2020). Biologically, a more diverse microbial community can prevent the dominance of specific groups that can be more efficient in producing N₂O and CH₄ (Cezar et al., 2021). Lastly, maintaining healthy soil could increase CO₂ capture in living biomass and, therefore, enhance C sequestration.

The systematic literature review indicated that despite BMPs-related terms are frequently mentioned in the literature, the number of studies that assess soil C under these strategies is reduced. The review also indicated the urgency for standardization of the sampling depth, which should minimally include the top 30 cm layer and preferentially reach 100 cm depth (at least). Evidence suggest that assessing soil C considering only the utmost soil layer may not represent soil C dynamics properly, as the impact of cropping systems can affect deeper layers (Smith et al., 2020). A significant gap on soil C data was observed specifically for Central, Caribbean, and Andean regions. The low availability of data for these regions, combined with the lack of technical criteria, prevented the extraction of C accumulation rates via meta-analysis (Bown and Sutton, 2010; Hansen et al., 2022). Therefore, the scenarios were generated from an extensive literature review focused on identifying C sequestration rates in regions with data availability (Table 1 and Supplementary Figure S4). Most of the extracted rates were obtained from meta-analysis studies per se, as for North (Conant et al., 2017; Nicoloso and Rice, 2021) and South (Viglizzo et al., 2019; Maia et al., 2022; Frazão et al., 2023) America. But for other regions (Central, Caribbean, and Andean), isolated studies were adopted. The use of individual studies for Central, Caribbean, and Andean regions is a shortcoming encountered in the analysis, which unfortunately, represents a bias in the reported data (Lin and Chu, 2018; Rothstein et al., 2005).

Another challenge was identifying the real extension and the most promising location (within each landscape/region) of degraded agricultural areas where implementation of BMPs would be more technical and socio-economically feasible for increasing soil C sequestration. Due to a lack of data, these estimates are restricted to the extension of 30% of the available agricultural area, an average estimate of the degraded land proportion (see Materials and Methods, section 2.2.1). It is uncertain whether this value is below or above the true extent. Further limitations associated with this study are related to the poor availability of soil C data across the region, which would be necessary to account for soil C saturation status. As expected, the reported soil C sequestration rates (Table 1) may not be fully achieved depending on the current soil C levels. The absence of these data prevents obtaining precise estimates of the C sequestration potential, which is an essential step towards establishing public policies aimed at mitigating GHG emissions through agriculture for addressing anthropogenic climate change. Despite this, we consider the results presented here valuable and timely because it provides an order of magnitude of the mitigation potential with the expansion of BMPs in the Americas based on evidence. Such approach can be improved at the pace that more information is generated for the region, enhancing the quality of the analysis.