Soil Health Monitoring

Soil health is a phrase that describes the physical, chemical, and biological functions of soils in the context of production agriculture. Certain agricultural practices (like no-till, cover cropping, and managed stock grazing) have been scientifically proven to improve the function of agricultural soils. Over the past two decades or so, the concept of “soil health” and its associate practices have become increasingly accepted across the North American agriculture sector.

However, if practices are adopted with the goal of improving soil health, it is also important to accurately measure how these practices are changing soil properties. Some soil properties can respond to agricultural practice changes in just one year (e.g. active carbon). Others can take 5 years or more (e.g. SOM, SOC). Additionally, soil sampling should be conducted in a spatially representative manner.

The following resources provide information on key soil health measurements and developing spatially-informed soil sampling methods:

Comprehensive Assessment of Soil Health – The Cornell Framework
Soil Sampling Strategies – MSU Extension

Documenting Producer-Driven Innovation in Field Crop Production

Agronomic research is important for improving the profitability and sustainability of field crop production. Production improvements are also driven by field crop producers through on-farm innovation. Using observation and on-farm trials, producers do the work of translating scientific findings into profitable and practical production practices at the field and farm scale. In North America, field crop producers play a significant role in ensuring that scientifically-verified practices like no-till/high residue systems and intercropping. are viable in production systems.

Often, such innovations are not documented and made available in a format where information can be readily shared. Social science research methods can be used to address this issue. However, social science research methods are rarely deployed in such a manner. This issue is not isolated to agronomy alone, as it has recently been observed by (Martin 2019) that there is a troubling lack of integration between social sciences and environmental and earth sciences.

Prescribed burning resources for agricultural landowners and land managers

Grasslands of the Canadian Prairie provinces and U.S. Plains states have a history of management by Indigenous peoples. One of primary management tools used by Indigenous peoples is cultural burning. Cultural burning can promote the growth of economically useful plant species, improve fodder for game species, and reduce the risk of catastrophic wildfires (Roos et al. 2018).

The poor understanding Indigenous land management methods and grassland ecologies by settler governments and land managers led to a long period where fire was not widely used as a grassland management tool (1880s to 1980s) (Stubbendieck, Volesky, and Ortmann n.d.). In the 1980s, increasing autonomy won by Indigenous groups in Canada and the U.S. and an increasing acceptance of fire as a land management tool by government agencies led to an increase in the use of cultural and prescribed burning.

Private agricultural lands are typically not managed using prescribed burning. The lack of prescribed burning on private agricultural lands has significant implications for livestock fodder quality, invasive weed control, the encroachment of woody vegetation, flora and fauna biodiversity, and grassland regeneration (Stubbendieck, Volesky, and Ortmann n.d.).  However, there are some significant barriers to implementing prescribed burning plans on private land. These include costs related to insurance and fire crew deployment along with a lack of accessible prescribed burning expertise.

Most provinces and states in the Canadian Prairies and U.S. Plains states have well-defined rules and regulations for prescribed burning. Some jurisdictions may also provide technical and financial support for agricultural landowners and managers. Below are some prescribed burn resources provided by non-profit and government agencies:


Course: Introduction to Prescribed Fire in the Grassland Environment
Canadian Prairies Prescribed Fire Exchange


Kansas Prescribed Burning – Rules and Regulations
Kansas Flint Hills Smoke Management

Prairie-Plains Resource Institute
Nebraska Prescribed Fire Council
Loess Canyons Rangeland Alliance

Oklahoma Prescribed Burning Handbook

North Dakota
North Dakota Prescribed Fire Cooperative

South Dakota
Mid-Missouri River Prescribed Burn Association

Prescribed Burn Program – Department of Agriculture

Web-based Mapping Platforms

PC-based Geographic Information System (GIS) software is the most robust tool available for compiling, analyzing, and representing geospatial data. The leading commercial GIS package, ArcGIS, enjoys broad use and specialized extensions. Similarly, QGIS is one of the most widely used open-source GIS packages and has many similar capabilities as ArcGIS. ArcGIS only runs on Windows while QGIS is maintained across Windows, MacOS, and Linux platforms. Since QGIS is open source, there is robust third-party support and feature development.

PC-based GIS has the capability to interface with networked geospatial databases. This connectivity allows for distributed data input and analysis, as well as map sharing. However custom geospatial databases must be developed and maintained – which is costly. Web-based GIS platforms only provide a small number of features relative to PC-based GIS software. But web-based GIS platforms permit the sharing of static and dynamic maps quickly, easily, and securely, without the need for custom database solutions.

Some popular web-based GIS platforms include:

1) Google MyMaps (free)

Includes a basic set of GIS features such as distance measurement, area delimitation, georeferenced point identification, and map layering. Offers a number of basemap styles. Supports data import via Excel files. Map access permissions are controlled via Google’s user account system.

2) ArcGIS Online (subscription)

A web-based version of ArcGIS. Includes a more robust set of basic GIS features compared to Google MyMaps. Includes a broad set of map layers and imagery for the creation of unique base maps. A smartphone and tablet app, Collector for ArcGIS, allows for users to add data directly to ArcGIS Online maps from the field. ArcGIS online provides robust security measures that can limit shared maps and data to specific users within an organization, but also allows for the public sharing of maps and data.

3) BatchGeo (subscription)

BatchGeo provides an extended feature set for Google Maps, such as automated geospatial data import, the embedding of dynamic maps into websites, and basic spatial analysis.

Land cover and carbon sequestration in agricultural landscapes

Over the past 200 years, large tracts of diverse biomes such as North American grasslands, the northern Kazakh steppe, the Brazilian cerrado, or Ontario’s Carolinian forest zone have been converted to commodity-based agricultural production (Comer et al. 2018; Petrick, Wandel,and Karsten 2013; Oliveira and Hecht 2016; Bowley 2016). The process of clearing more “natural” forms of land cover for commercial agricultural use has significantly compromised the carbon sequestration capacity of terrestrial biomes at the ecosphere scale (Zomer et al. 2017).

In most cases, the carbon sequestration potential of agricultural land cover categories cannot match the potential of well-managed, relatively natural forms of land cover (Deng, Liu, and Shangguan 2014; Yang et al. 2019; Purakayastha, Huggins, and Smith 2008; Castaño-Sánchez et al. 2021). For example, field-based research conducted in a south-central Ontario agricultural landscape demonstrates that areas of remnant forest contained 25% of farm-wide soil carbon even though remnant forest comprised only 15% of land cover at research sites (Mazzorato, Esch, and MacDougall 2022). These findings suggests that field boundary areas under perennial land cover such as trees and grasses provide more carbon sequestration capacity than areas under annual field crops.

Agricultural landscapes may serve as atmospheric carbon sinks if managed using practices that have been scientifically verified to sequester more carbon than they emit. It is important to stress that climatic and other environmental conditions may override the effects of changes to land use and land cover or shifts made in land management practices (Spengler 2011; Zomer et al. 2017). Recent research also points out that the agricultural sector continues to be a net source of GHG emissions and any practices that may sequester carbon over the long term may not provide additional capacity to sequester carbon from other economic sectors (Schlesinger 2022).

Carbon sequestration occurs in agricultural landscapes via two pathways: above-ground biomass growth and soil-based carbon sequestration. Above-ground biomass growth ties up carbon for relatively short timeframes (even woody biomass) because carbon stored in biomass is released back into the atmospheric carbon pool as plant matter decomposes. Therefore carbon stored as above-ground biomass is not seen as a viable means of long-term carbon sequestration (Zeng et al. 2013).

The second pathway is via soil-based carbon sequestration. Soil-based carbon sequestration occurs when organic forms of carbon are stored in soils for relatively long timeframes. The primary mechanism for sequestering carbon in the soil at longer timescales is via exudatessecreted by plant roots into the soil rhizosphere. Exudates are short-chain starch molecules that serve a variety of purposes in the rhizosphere. In relation to carbon sequestration, exudates can be converted to stable forms of soil carbon by soil microbiota. A secondary mechanism of soil-based carbon sequestration is the decomposition of plant biomass at the soil surface and in the soil profile (Kell 2012; Jones, Nguyen, and Finlay 2009).

The ability of agricultural soils to sequester carbon is determined by two biophysical factors: biotic processes and physical soil structure. Biotic processes of living plants, animals, fungi, and microorganisms drive carbon inputs and outputs in soils. The physical structure of soils is comprised of inorganic and organic particles in the soil profile. Biotic and physical soil properties vary by depth. Upper soil layers closer to the surface are more influenced by the environment and land use and land cover. Most soil carbon is found in soil horizons closer to the surface (O and A soil horizons) (World Bank 2021, 11).

In conclusion, soil-based carbon sequestration is the primary pathway of carbon sequestration in agricultural landscapes. Even with field crop land management practices that promote soil-based carbon sequestration, this may not be enough to offset GHG emissions from within the agriculture sector. More “natural” forms of land cover provide better carbon sequestration capacity than areas of annual crops. To maximize carbon sequestration in agricultural landscapes, more attention should be paid to field boundary areas, pastures, and the conversion of marginal annual cropping areas back to perennial forms of land cover.

Why are the concepts of land use and land cover relevant to agricultural landscape management?

The concepts of land use and land cover are increasingly relevant to agricultural landscape management. Land use refers to how land is utilized based on economic and policy decisions. Land cover refers to vegetation characteristics (USDA ERS 2022) Precision agriculture and the quantification of ecosystems services require that agricultural landscapes be classified into different land use and land cover categories using a process called stratification (World Bank 2021)

Stratification relies on geospatial tools and data along with field surveys to divide areas of land use and land cover into clearly bounded and georeferenced units. Geospatial tools and data include GIS, GPS, aerial photos, and satellite imagery. Field surveys can include direct observations or surveys and interviews with land managers.

There are no set rules for determining land use and land cover categories – they are often tailored to particular applications. Site stratification for quantifying ecosystems services in North American agricultural landscapes might divide land use into three categories with land cover as a sub-category, as shown in the table below.

Land Use CategoryLand Cover Sub-Category
Crop FieldAnnual Crops; Perennial Forage
PastureGrasses; Perennial Forage
Field Boundary Habitat (FBH)Grasses; Tree/Shrub; Wetland

Land use and land cover categories should have clear delineations, but this is not always easy in agricultural landscapes, since crop fields seeded to perennial forage may serve similar economic and ecosystems functions when compared to pasture with perennial forage species.

From a precision agriculture perspective, land use and land cover are important because crop field boundaries must be clearly delineated from other categories of land use. Variable rate applications and yield maps also rely on clear, georeferenced boundaries. From an ecosystems services perspective, land use categories and land cover sub-categories can serve very different ecosystem functions. In order to quantify these functions, crop fields, pastures, and Field Boundary Habitat must be clearly divided with georeferenced boundaries. For example, areas of crop fields that are not devoted to annual crop production such as fencelines, permanently retired headlands, windbreaks, and wetlands can be quite small in area, but still provide critical flora and fauna habitat that crop fields are not able to provide.

In the direct measurement of carbon sequestration in agricultural landscapes, geospatial methodologies are being overlooked

Soil Organic Carbon (SOC) is the primary means of directly measuring carbon sequestration in agricultural soils. SOC is a measurable component of Soil Organic Matter (SOM), which is a commonly measured soil property. SOC can be measured in most scientific and commercial soil labs and is considered a standard soil test. Although SOC measurements in soil samples are straightforward, the development of representative soil sampling methodologies for SOC at the field and landscape scale poses challenges that are reflective of biophysical factors that drive changes in SOC.    

At the field and landscape scales, SOC tends to have significant temporal and spatial variability. Temporally, SOC levels fluctuate seasonally as a result of annual fluctuations in temperature and precipitation. Over the longer term (five or ten year intervals, for example), SOC levels can be slow to change, even with shifts to climate-friendly BMPs. SOC measurements must be timed in a way that permit accurate sampling over seasonal and multi-year timescales. Spatially, SOC is tied to biophysical factors such as historical and present-day land management practices, soil type distribution, landscape position, and climate (Abram 2020; Department of Agriculture and Food, Western Australia 2013; World Bank 2021)

In order to accurately reflect temporal and spatial SOC variability at the field and landscape scales, sampling methodologies must rely upon geospatial approaches. But in the grey literature and scientific publications on SOC measurement, geospatial approaches are being overlooked. For example, a recent World Bank carbon sequestration guidebook provides detailed information on soil sampling and testing, but glosses over the geospatial work required to ensure representative soil sampling (World Bank 2021). Similarly, a recent study on SOC in Ontario agricultural landscapes utilizes a geospatial approach, but the geospatial methodologies used are not made explicit and do not align with contemporary land use and land cover research (Mazzorato 2022). Geospatial methodologies for accurate temporal and spatial SOC sampling can largely rely upon geospatial data collected in the field or public geospatial datasets. The below table lists geospatial methodologies that should be considered mandatory for building accurate SOC sampling methodologies. These methodologies are fundamental to the fields of remote sensing and GIS and can be developed using standard textbooks and scientific articles in these fields (Campbell and Wynne 2011; Islam, et al. 2019; Bassett and Zuéli 2000)

Geospatial MethodologyData Type(s)Description
site boundary delimitation with GPS + GISGPS field boundaries; georeferenced vertical aerial photographsEnables accurate field boundaries to be drawn using GIS software to process and combine field GPS boundary data and aerial photographs
site stratification via aerial photographsfield GPS boundary data; georeferenced vertical aerial photographsEnables the accurate delimitation of different land use and land cover types at the field scale, such as separating areas of tree or perennial grass cover from cropping areas
soil sampling with GPSGPS soil sampling site coordinatesEnables accurate GPS coordinates to be taken at each soil sampling site. Essential for repeated sampling over multi-year periods
land use and land cover surveys and interviewsquantitative and qualitative data; supplements site stratification process (above)Land use and land cover history is often poorly documented. Robust methodologies which pair land manager interviews and surveys with historical air photos and vegetation maps can be used to gain insight into historical patterns of land use and land cover. It is necessary for surveys and interviews to be conducted in a manner which aligns with the site stratification process (above)

Why is it difficult and costly to directly measure soil carbon in agricultural landscapes?

It is argued that worldwide efforts to shift agricultural land management to practices which sequester excess atmospheric carbon may only provide a small percentage of the sequestration capacity that is required to mitigate climate change. (Schlesinger, 2022). Although the carbon sequestration capacity of agricultural soils may be marginal, the co-benefits provided by managing agricultural soils to sequester carbon are significant. Co-benefits include improved soil function (or soil health), improved production system resiliency, and reduced soil, air, and water pollution.

From land management, scientific, and policy perspectives, it can be important directly measure soil carbon levels. Soil carbon levels are quantified using Soil Organic Carbon (SOC). SOC is a measurable fraction of Soil Organic Matter (SOM) (a commonly measured soil property) (Moorberg and Crouse, 2017). While SOM comprises a small percentage of most soil mass (between 2-10%), it plays an important role in the biophysical function of soils. SOC refers to only the carbon component of SOM (Government of Western Australia, 2021).

Measuring SOC at the field and landscape scales poses significant methodological and cost challenges. Sampling methodologies must be temporally and spatially representative and SOC must be measured at the correct depth in the soil profile. At the same time, the large numbers of soil samples required for proper temporal and spatial representation and relatively deep soil cores would incur significant labor costs for sample collection and lab analysis. A World Bank carbon sequestration guidebook states that methodological challenges and high costs remain a significant barrier to “implementing transparent, accurate, consistent, and comparable methods for [the] measurement” of carbon sequestration (2021, 7).

Over the short term SOC levels tend to fluctuate throughout the year based on seasonal fluctuations in precipitation (Abram 2020). Over the longer term (such as five or ten year intervals), SOC levels can be slow to change, even with shifts to carbon sequestration-focused land management practices (Department of Agriculture and Food, Western Australia 2013; World Bank 2021). SOC measurements must be timed in a manner that allows for accurate quantification over both seasonal and multi-year time scales.

SOC changes can be highly variable across fields and landscapes and are closely tied to biotic and abiotic factors such as historical and present-day farm management practices, soil type distribution, landscape position, and climate (Tautges et al. 2019; Morais, Teixeira, and Domingos 2019). SOC measurements must accommodate for spatial variability at the field and landscape scales.

To account for the spatial and temporal variability described above, it is recommended that SOC measurements be taken at a depth of 30cm or even deeper (Slessarev et al. 2021). Soil samples taken at proper depths require vehicle-mounted hydraulic probes (Franzen 2018).

Sampling SOC at temporally and spatially representative intervals at proper depths at the field and landscape scales requires large numbers of soil samples (NC State Extension 2017). The testing of soil samples for SOC is related to SOM and uses common soil lab equipment. SOC can be measured by most commercial soil labs. However, measuring SOC in lab settings is a labor-intensive process, especially when large numbers of samples are involved (FAO 2019).

The direct measurement of SOC at the field and landscapes scales can be useful for scientific, land management, and policy purposes. However, the methodological and cost barriers posed by representative soil sampling raises questions about the practicality of using SOC as an indicator for measuring the effectiveness of carbon sequestration-focused agricultural practices.

Is the carbon sequestration capacity of agricultural soils being overstated?

Since 1995 the IPCC has promoted carbon sequestration in agricultural soils as a one of many climate change mitigation strategies (IPCC, 1995). In principle, certain “improved” practices such as no-till, cover cropping, and increased crop diversity can result in the long-term storage of excess atmospheric carbon in the soil profile. The adoption of such practices constitutes a change in land management (in contrast to land cover and land use change which have higher carbon sequestration potentials). Government agencies and private enterprises have been keen to promote the development of carbon offset markets in which land managers are paid for carbon sequestered in agricultural landscapes.

One major issue with agriculture-based carbon offsets is that methodologies for quantifying carbon sequestration at the field and landscape scales are poorly developed. Another major issue with agriculture-based carbon offsets is that improved practices most likely can only provide a small amount of carbon storage that can be sold as credits to other economic sectors. Schlesinger (2022) (not open access) states:

“for the United States to mitigate 5% of its current carbon dioxide emissions to the atmosphere, about 0.94 tons of carbon would need to be stored in each hectare of cropland annually……For comparison land abandoned from agriculture typically accumulates carbon at rates of 0.1 to 0.3 mtC/ha/year…….any mitigation of climate change will first require that the agriculture sector become “carbon neutral” so that additional soil carbon sequestration might reduce the content of carbon dioxide in the atmosphere.”

Schlesinger goes on to argue that most improved management practices are not likely to lead to large amounts of net carbon sequestration in agricultural soils due to emissions related to agricultural production and photosynthetic constraints.

Even if agricultural soils have a limited capacity to store anthropogenic carbon emissions, farming practices designed around maximizing carbon sequestration remain important. Carbon sequestration can help the agriculture sector reach “net-zero” emissions (when paired with significant GHG emissions reductions). Additionally, carbon sequestering practices like no-till, cover cropping, and increased crop diversity can provide production benefits which improve soil function such as increased water holding capacity and reduced erosion. From an environmental perspective carbon sequestering practices can reduce soil, air, and water pollution when implemented at broader scales.

In summary, practices designed around carbon sequestration can provide significant benefits. However, the small carbon sequestration capacity of agricultural soils raises questions about the viability of ag.-based carbon offset markets.

Measuring and Modelling Carbon Sequestration in Agricultural Landscapes

Carbon sequestration in agricultural landscapes is promoted as a key climate change mitigation pathway. The IPCC has long promoted carbon sequestration in agricultural landscapes and more recently governments and private companies have promoted commercialized carbon offsetting for agricultural landscapes. However, methodologies for estimating Soil Organic Carbon (SOC) are poorly developed at the field and landscape scales. SOC can be measured at the field scale using direct soil sampling and modelled at the landscape scale using representative biophysical data. But the effort and cost required to collect and process biophysical data (including soil samples) makes it unclear if SOC can be accurately quantified at broad scales. The highly variable temporal and spatial nature of SOC requires robust geospatial approaches in developing sampling methodologies for biophysical data collection and modelling. At this time, robust geospatial approaches are absent from efforts to quantify SOC in agricultural landscapes.

Soil carbon levels are typically quantified using Soil Organic Carbon (SOC) measurements. SOC is a measurable fraction of Soil Organic Matter (SOM) (a commonly measured soil property). While SOM comprises a small percentage of most soil mass (between 2-10%), it plays an important role in the biophysical function of soils. SOC refers to only the organic carbon component of SOM.

SOC is highly spatially and temporally variable. This is related to soil properties along with historical and present-day land use, land cover, and land management. At the same time, SOC levels fluctuate throughout the year, which is related to temperature, precipitation, and land management actions. The variability of SOC requires the development of sampling methodologies which reflect spatial and temporal characteristics of SOC. Large numbers of soil samples would be required to accurately quantify SOC even at the field scale.

SOC can be easily quantified in most research and commercial soil laboratories. However, preparing and processing soil samples for SOC measurements is labor-intensive and costly. Sample processing costs pose another serious challenge to accurately quantifying SOC.

The IPCC recognizes that intensive soil sampling is not possible at the landscape scale. Therefore, carbon models have been developed to enable SOC estimates at the landscape scale. Models are divided into Tiers 1, 2, and 3. Tier 1 models rely on generalized, default biophysical values provided by the IPCC and therefore have a high degree of uncertainty. Tier 2 models replace default values with some landscape-specific biophysical data and land use, land cover, and/or land management information. Tier 2 models reduce some uncertainty in comparison to Tier 1 approaches. Tier 3 models have much lower uncertainty when compared to the other tiers. Tier 3 models require landscape-specific biophysical and landscape data including:

soil type
land use history
current land use and land cover

land management

Thanks to the significant uncertainty of Tier 1 and Tier 2 models, they should not be used for anything other than estimates at project planning stages. Tier 3 models may have the potential to accurately estimate SOC at the field and landscape scales. But very few Tier 3 models have been successfully implemented and verified at this time.

Scientific and grey literature focused on carbon sequestration focuses on SOC at the field scale and largely neglects the use of geospatial methodologies. This is a major shortcoming in the literature since geospatial methodologies are essential to scaling up field-scale biophysical data to broader landscape scales. Basic operations such as field boundary delimitation, field area measurements, and land use and land cover stratification types can be achieved using a combination of handheld GPS units, GIS software, and high- resolution aerial photos. At the same time, it should be stressed the pre-existing geospatial soil datasets lack the resolution, accuracy, and soil profile information required for accurate Tier 3 modelling.