Vertical distribution of soil organic carbon and nitrogen under warm-season native grasses relative to croplands in west-central Indiana, USA

Abstract

Establishment of grasslands can be an effective means of sequestering soil organic carbon (SOC) and reducing atmospheric CO₂ that is believed to contribute to global warming. This study evaluated the vertical distribution and overall sequestration of SOC and total nitrogen (N) under warm-season native grasses (WSNGs) planted 6–8 years earlier relative to a corn (Zea mays L.)–soybean (Glycine max L.) crop sequence, and switchgrass (Panicum virgatum) relative to tall mixed grasses of big bluestem (Andropogon gerardi), indiangrass (Sorghastrum nutans), and little bluestem (Andropogon scoparius). Paired soil samples from 0–15, 15–30, 30–60 and 60–100 cm depth increments were taken from WSNGs and adjoining croplands at 10 locations, and from switchgrass and adjoining tall mixed grasses at four locations in three major soil types of alfisols, mollisols, and entisols in Montgomery County, Indiana. Significant differences in SOC and N concentrations of WSNGs and croplands were limited to the surface 30 cm. On average, SOC concentrations in the surface 15 cm depth were higher in WSNGs than croplands (average: 22.4 and 19.8 g kg⁻¹ C, respectively) but significant differences were observed in just 4 of 10 locations. Similarly, surface soil SOC concentrations were not different for switchgrass (22.1 g kg⁻¹) relative to tall mixed grasses (21.4 g kg⁻¹). Soil N concentrations never differed significantly among land use treatments. On average, SOC mass calculated to 1.0 m depth was 9.4% higher under WSNGs than cropland (P < 0.058), and 8.1% higher in switchgrass relative to tall mixed grass (P < 0.054), but soil N mass was the same for both WSNGs and cropland. Vertical distribution under WSNGs of SOC mass was 26, 21, 28, and 25%, and of total N mass was 31, 25, 28 and 16%, in the 0–15, 15–30, 30–60, and 60–100 cm depth intervals, respectively. Even though we acknowledge the potential influence of soil variability or prior landscape processes on our results at some locations, we estimated that WSNGs sequestered an average 2.1 Mg C ha⁻¹ yr⁻¹ more than the corn–soybean sequence.

Introduction

Recent concerns about global warming due to atmospheric CO₂ accumulation have encouraged the achievement of a better understanding of the role of agriculture in mitigating CO₂ emissions. Restoration of grasslands is believed to be a cheap, efficient, and environmentally friendly method to reduce the rate of increase of atmospheric CO₂ concentrations, virtually stop soil erosion, increase soil nutrient and water retention, and improve soil and environmental quality. Although reseeding tilled soils with perennial grasses have been shown to provide greater litter and root biomass for C storage than annual cereal crops (Mapfumo et al., 2002, Mensah et al., 2003), widely different effects and rates of C sequestration by grasses have been reported (Staben et al., 1997, Robbles and Burke, 1998). This is because the quantity and rate of SOC stored under grasses is affected by such factors as soil type (Gebhart et al., 1994, Ma et al., 2000a), soil nutrient status (Ma et al., 2000c), grass types, cultivars or grass combinations (Frank et al., 2004, Tufekcioglu et al., 2003, Ma et al., 2000c), and length of time after establishment (Ma et al., 2000a, Ma et al., 2000c, Zan et al., 2001, Garten and Wullschleger, 2000, Gebhart et al., 1994).

The period of time following the establishment of grasses (age of grass) is important to the rate of SOC sequestration. Ma et al. (2000a) reported that although switchgrass influenced C mineralization, microbial biomass, turnover, and soil respiration in the short term, total SOC concentration measured at 0–15 and 15–30 cm depth intervals under switchgrass was not different 2–3 years after establishment. However, SOC was 45 and 28% higher in these depth intervals 10 years after establishment relative to adjacent fallow soil (Ma et al., 2000c). Staben et al. (1997) found that C mineralization potentials and SOC pools were significantly higher for wheatgrass (Agropyron spp.) but that total organic C and microbial biomass C were not significantly different for wheatgrass relative to wheat-fallow rotation, 4–7 years after grass establishment. Similarly, Garten and Wullschleger observed a significantly higher coarse root C for switchgrass relative to tall fescue, corn and pasture of mixed grasses, but SOC under switchgrass was not significantly different from these soil covers in the surface 40 cm soil depth, 5 years after establishment.

The effects of grass species on C accumulation also vary considerably. For example, Lal et al. (1998) showed that tall fescue (Festuca arundinacea) and smooth bromegrass (Bromus inermis) increased the soil C pool by 17.2% relative to a corn–soybean rotation. Garten and Wullschleger (2000) reported that 19–31% of soil C pool originated from switchgrass, 5 years after establishment. In contrast, Frank et al. (2004) reported that switchgrass accumulated more C than continuous wheat (Triticum aestivum L.) crop but less than native pasture. Varietal differences in SOC or N sequestration may also occur but, thus far, Sladden et al. (1991) showed that total N content of above-ground harvested biomass of eight switchgrass varieties were not significantly different. In the latter study soil N contents under these switchgrass varieties were not reported.

Fertilization practices and presence or absence of associated leguminous plants are other factors that affect the widely different rates of C sequestration reported for grasses. In a grass versus cropland study of conservation reserve programs (CRP), Gebhart et al. (1994) found that unfertilized perennial grasses added 1.1 Mg C ha⁻¹ yr⁻¹ to the upper 1 m of Midwestern soils over a 5-year period, but both rhizosphere deposition and fine root turnover in switchgrass may add up to 3 Mg ha⁻¹ yr⁻¹ (Bransby et al., 1998). In a more recent study in the Midwest, Al-Kaisi et al. (2005) in Iowa reported a 1.2 Mg C ha⁻¹ yr⁻¹ in the 0–15 cm soil depth for switchgrass, 10 years after establishment. Similarly, Mensah et al. (2003) reported that after 5–12 years of establishment a restoration grassland that consisted of mixed-species of wheat grass, blue gramagrass (Tripsacum dactyloides), and alfalfa (Medicago sativa) gained about 0.6–0.8 Mg C ha⁻¹ yr⁻¹ in the top 15 cm soil depth relative to wheat–wheat–canola–pea and wheat–wheat–fallow rotations in east central Saskatchewan, Canada. In Mandan, North Dakota, Frank et al. (2004) reported that SOC measured to 0.9 m depth increased at the rate of 10.1 Mg C ha⁻¹ yr⁻¹, 3 years after establishment of switchgrass with fertilizer application.

While several studies in the Great Plains region and elsewhere have shown that conversion of croplands to grasses can increase SOC, relatively few studies are available to better understand soil N cycling among grass types, or when croplands are converted to grasses, despite the acknowledged interdependency of SOC and N. The dynamics and accumulation of SOC and N may depend on vegetation composition. Legumes can increase SOC and N accumulation rate, C₃ grasses decrease these rates, while C₄ grasses may increase SOC but not N accumulation (Knops and Tilman, 2000). In the surface soil, Chen and Stark (1999) found total N to be significantly greater under big sagebrush (Artemisia tridentata) than wheatgrass. However, Svejcar and Sheley (2001) reported little or no consistent difference in total N when native perennial vegetation was replaced by annuals. In a recent study, Al-Kaisi et al. (2005) reported higher total N under switchgrass than corn–soybean–alfalfa rotation in the 0–5 and 15–30 cm depth intervals, 10 years after switchgrass establishment. McLaughlin and Kszos (2005) estimated the buildup of SOC under switchgrass could result in up to 100 kg N ha⁻¹ yr⁻¹ over a 10-year growing cycle. In contrast, Paustian et al. (1990) reported a net deficit of soil N in grass leys relative to barley due to a higher rate of mineralization under grass leys. Similarly, Kucharik et al. (2003) reported a relatively lower N under a 24-year restored prairie ecosystem relative to monoculture corn. Paustian et al. (1990) attributed higher N mineralization rates in grasses to the presence of a substantially larger root mass throughout the growing period, and a subsequently more significant rhizosphere influence of N mineralization leading to higher N demand of the grass.

Given that more than two-thirds of the annual grassland biomass production can be allocated to below-ground structures (Korner, 2002), accumulation of organic matter in deep soil layers can make an important contribution to C sequestration in most grassland ecosystems (Korner, 2002, Liebig et al., 2005). Rumpel et al. (2002) showed that 50% of SOC was stored within the deep horizons with a higher residence time than for SOC stored in upper layers where microbial activity is high. Ma et al., 2000b, Ma et al., 2000c suggested that rooting system of switchgrass extended up to 3 m down the soil profile; as much as about 10% of total root biomass of switchgrass have been reported in the 60–90 cm soil depth interval (Bransby et al., 1998, Liebig et al., 2005). However, little information is available on SOC and N sequestration in deep soil horizons because the majority of studies on SOC sequestration are limited to storage in the 0–30 cm soil layer (Lemaire et al., 2005). Lemaire et al. (2005) and Murphy et al. (2003) have argued that uncertainties surrounding SOC depth and the absence of an integrated view of the C and N dynamics posed a serious limitation to our knowledge of the carbon–nitrogen cycles in grassland ecosystems.

In the Midwest USA, farmers are encouraged to established WSNGs partly for wildlife habitat restoration, soil conservation, and the potential of harvesting certain WSNGs for fuel production (Bransby et al., 1998). However, in more recent times following the Kyoto Protocols on reduction of greenhouse gases, future possibilities of financial incentives resulting from carbon credit trading, whereby farmers trade SOC conserved in their fields for financial credits from greenhouse gas emitting entities are an additional incentive if the SOC sequestration potential of WSNGs can be established. Because other conservation practices such as reduced tillage or continuous no-till practices also help to conserve SOC, there is considerable uncertainty about which management practices conserve more SOC than the other and to what depth this can be expected. In Indiana and elsewhere in the Midwest, data to systematically quantify SOC sequestered by pure- and mixed-stands perennial grasses especially at deeper depths (>60 cm) relative to croplands that are managed under reduced or no-till are limited. The objectives of this study were to evaluate the vertical distribution and total SOC and N accumulation to a 1.0 m soil depth for (1) WSNGs relative to croplands, and (2) switchgrass relative to mixed grasses.