Soils

12.5.1 Agriculture

Because more than 50% of the Earth’s vegetated surface is dedicated to agriculture (e.g., cropland and grazing land), understanding the role of agricultural management on SOC stocks is critical (see Ch. 2: The North American Carbon Budget). Virtually all management choices (e.g., crop type, rotation, tillage, fertilization, irrigation, and residue management) will affect carbon inputs (e.g., crop residues and manure) and the decay rate or erosional loss of SOM (Paustian et al., 1997; Smith 2008). In most cases, SOC changes occur slowly and short-term (annual) changes are difficult to measure, but studies from long-term experiments, together with improved predictive models, provide a basis for guiding management and policies to improve SOC stocks (NAS 2010; Ogle et al., 2014; Paustian et al., 2016).

Causes of SOC loss include 1) reduced biomass carbon inputs; 2) enhanced erosion and leaching; and 3) increased decomposition rates due to tillage disturbance (Paustian et al., 2016). A meta-analysis for Canadian soils reported that, when native soil was converted to agricultural land, there was an average loss of 24% ± 6% of soil carbon (VandenBygaart et al., 2003). Globally, agricultural soils have lost, on average, 20% to 45% of their original top soil carbon (0 to 30 cm) but with much higher losses in cultivated organic soils and where extensive erosion has occurred (Don et al., 2011; Ogle et al., 2005). Following restoration of perennial forest and grassland vegetation on annual cropland (e.g., for soil restoration or retiring marginal lands from production), much of the lost soil carbon stocks eventually can be recovered. Conversion of annual cropland to perennial grassland in temperate environments increased soil carbon stocks, on average, by 13% to 16%, with greater relative increases occurring in more mesic climates (Ogle et al., 2005).

In recent decades, SOC stocks in agricultural soils in the United States and Canada have stabilized and in some cases begun to increase (Follett et al., 2011; U.S. EPA 2015) as new conversion of land to agricultural use has largely halted and adoption of soil conservation practices and crop yields have increased (Chambers et al., 2016; Johnson et al., 2006). Effects of agriculture on soil carbon stocks, along with effects of conservation measures, are reviewed and quantified in Angers and Eriksen-Hamel (2008), Hutchinson et al. (2007), Luo et al. (2010), Palm et al. (2014), Paustian et al. (2016), Powlson et al. (2014), and many others. Improved residue management, added forage in crop rotations or adoption of agroforestry, double-cropping, conservation reserve planting, increased use of perennials in rotation, and use of practices that increase plant growth such as effective fertilization are successful in increasing soil carbon (Hutchinson et al., 2007; Luo et al., 2010; Palm et al., 2014), especially if more than one practice is used. In Canada, the wide adoption of reduced tillage and summer fallow over many regions has resulted in soil carbon increases and reduced erosion (Agriculture and Agri-Food Canada 2016; Soil Conservation Council of Canada 2016).

An analysis of no-till only versus conventional till by Palm et al. (2014) found that carbon gains occurred in only half the paired comparisons and that increased residue retention had a greater effect on soil carbon than reduced tillage. Powlson et al. (2014) argue that adoption of no-till agriculture can improve crop production and reduce erosion in many cases, but it may not have significant effects on carbon sequestration. However, a meta-analysis by Kopittke et al. (2017) saw an overall small positive (+9%) effect of conversion to no-till from conventional till methods. Most analyses of tillage effects do not account for SOC erosion. Montgomery (2007) calculated a mean erosion rate difference between conventional agriculture and no-till agriculture of about 1 mm per year. Although this eroded soil causes a net movement of carbon from the site with associated negative effects on soil fertility and health, this movement might not represent a net loss of soil carbon globally and could represent a net sink, because the eroded carbon can be buried and therefore protected. Meanwhile, carbon accumulation can continue in the site from which the erosion originally occurred via the usual processes of additions and transformations of plant residues (Wang et al., 2017).

Estimates of the current SOC balance for U.S. agricultural lands suggest a small net sink on long-term cropland (6.4 Tg C per year) and on land recently converted to grassland (2.4 Tg C per year), while small net losses of SOC were estimated for long-term grassland (3.3 Tg C per year) and land recently converted to cropland (4.4 Tg C per year; U.S. EPA 2015). A similar picture appears for Canadian agricultural soils with an estimated net sink of about 3 Tg C per year (ECCC 2015). A full soil carbon inventory for Mexican agricultural soils is still in progress; however, with ongoing forest conversion to agricultural uses (see Section 12.4.2), there likely is a substantial loss of SOC due to agricultural activities.

Other chapters present more information on management of agricultural soils and its effects on carbon (see Ch. 5: Agriculture; Ch. 7: Tribal Lands; and Ch. 10: Grasslands).

12.5.2 Forestry

A wide variety of forest management practices affect around 204 Mha of timberlands in CONUS (see Ch. 9: Forests). Those practices typically involve a combination of harvesting, stand regeneration, and stand tending. The intensity of those practices and their resulting effects on soils depend on landowner management objectives.

To date, most research on forest harvest effects on soil carbon has suggested that mild to moderate intensity harvesting does not cause measurable changes in upland soils (Johnson and Curtis 2001), but that intensive harvesting and plantation management may cause reductions in mineral soil carbon (Buchholz et al., 2014; Johnson and Curtis 2001), especially if imposed on old-growth natural stands. A meta-analysis of studies measuring effects of forest harvest on soil carbon stocks by Nave et al. (2010) found that while forest floor carbon generally was reduced after harvest, mineral soil carbon was less affected, although certain soil orders were more susceptible to mineral soil carbon loss than others. Forest soil carbon stores have the ability to recover to preharvest stages, although recovery might take decades (Nave et al., 2010) to a century or more (Diochon et al., 2009); thus, rotation length plays a significant role in the degree of harvest impacts on soil carbon. Several chronosequence studies have observed reductions in mineral-bound carbon pools in successional stands decades after harvesting (Diochon et al., 2009; Lacroix et al., 2016; Petrenko and Friedland 2015). Because this timing of carbon loss corresponds to periods of high nutrient demands during biomass re-accumulation, the cause could be mining of SOM by plants and mycorrhizal fungi to alleviate nutrient limitation. Dean et al. (2017) argue from a modeling standpoint that there are more significant losses of soil carbon with forest harvest of primary forests when calculated over centuries, but this model result is not supported by empirical studies.

Afforestation and agroforestry (the practice of integrating woody vegetation with crop and/or animal production systems) have been cited as having potential for increasing soil carbon sequestration (IPCC 2000; Upson et al., 2016). Several meta-analyses conducted on afforestation effects on former croplands have produced a general consensus that soil carbon gains may take more than 30 years to be measurable (Barcena et al., 2014; Li et al., 2012; Nave et al., 2013) but can increase carbon stocks by 19% to 53% (Guo and Gifford 2002; Nave et al., 2013). However, while tree establishment in both grasslands and croplands showed greatly increased aboveground biomass carbon storage, meta-analysis of studies found that tree establishment on pastureland led to losses or no changes in soil carbon (Shi et al., 2013).