| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
USDA-ARS, 1420 Experiment Station Road, Watkinsville, GA 30677-2373
* Corresponding author (afranz{at}uga.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: LSD, least significant difference • SOC, soil organic carbon • TN, total nitrogen
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Estimates of potential C and N sequestration at depths below the traditional plow layer (i.e., 0.3-m depth) are few, but are needed to improve our understanding of management influences on greenhouse gas mitigation and nutrient cycling. A difference in SOC between adjacent cropland and a 5-yr-old conservation reserve grassland in the Great Plains USA was detected in various increments to 0.4-m soil depth, but not in depth increments from 0.4 to 3 m below the surface (Gebhart et al., 1994). Summed to 1-m depth, SOC content was not different between cropland and conservation reserve. In a shortgrass-steppe with 56 yr of grazing in Colorado, SOC was not different between unharvested and lightly grazed rangeland at any soil depth increment to 0.9-m depth (Reeder et al., 2004). In contrast, SOC was greater in four of seven depth increments to 0.9-m depth under heavily grazed compared with unharvested rangeland. At the end of 12 yr of grazing on a previously ungrazed mixed-grass rangeland in Wyoming, SOC and TN were greater with light and heavy stocking than an ungrazed exclosure at a depth of 0 to 0.3 m, but statistically similar between treatments at a depth of 0 to 0.6 m (Schuman et al., 1999). These results suggest that SOC and TN sequestration are soil depth-, management-, and site-specific.
The effect of forage utilization regime on C and N accumulation in surface residue and the surface 6 cm of soil was highly significant during the first 4 yr of bermudagrass management in Georgia (Franzluebbers and Stuedemann, 2001; Franzluebbers et al., 2001). Rates of C and N sequestration in this near surface zone under grazed systems were more than double those under ungrazed systems. Whether these changes in SOC and TN near the surface were also occurring deeper in the profile was not determined. Since the perennial root system of bermudagrass can reach to a depth of 1 m, management of aboveground forage could affect belowground distribution of roots and subsequent accumulation of SOC and TN. On a Piedmont soil in Georgia, bermudagrass root distribution in the soil profile (0–1.5 m) was 87 ± 4% (mean ± standard deviation) in the surface 0.3 m and 12 ± 4% in the 0.3- to 0.9-m depth (Carreker et al., 1977).
Most studies of profile SOC and TN have found few differences among management systems when summed to approximately 1 m, a typical rooting zone in many ecosystems (Gebhart et al., 1994; Schuman et al., 1999). Despite potential surface accumulation of SOC and TN with a particular management, consideration of the entire soil profile might reveal fewer differences due to (i) reversal of effects with depth or (ii) relatively low concentration and high natural variability of SOC and TN with depth resulting in little power to detect differences should they occur. Detection of differences due to management might be improved with sampling designs that take these variations into account.
One variable found to affect spatial distribution of SOC and TN in pastures is animal behavior. Under tall fescue (Festuca arundinacea Schreb.) in both Iowa (West et al., 1989) and Georgia (Franzluebbers et al., 2000), SOC and TN were greater within 10 m of shade and water sources than zones farther away in <1-ha pastures. Accumulation of SOC and TN near shade and water sources was attributed to more frequent fecal deposition and greater plant growth.
Our objectives were to (i) quantify profile contents of SOC and TN at the beginning and end of 5 yr of ‘Coastal’ bermudagrass management on typical soils in the Southern Piedmont USA, (ii) evaluate whether stratification of pastures into zones delineated by animal behavior might improve detection of differences in SOC and TN with time due to management, and (iii) quantify variability in SOC and TN with depth.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Experimental Design
The experimental design was a randomized complete block with treatments in a split-plot arrangement in each of three blocks, which were delineated by landscape features (i.e., slight, moderate, and severe erosion classes). Main plots were nutrient source (n = 3) and split-plots were forage utilization (n = 4) for a total of 36 experimental units. Grazed plots (i.e., paddocks) were 0.69 ± 0.03 ha. Spatial design of paddocks minimized runoff contamination and facilitated handling of cattle (Bos taurus) through a central roadway. Each paddock contained a 3 x 4 m shade, mineral feeder, and water trough placed in a line 15-m long at the highest elevation. Unharvested and hayed exclosures (100 m2) were placed side-by-side in paired low- and high-grazing pressure paddocks of each nutrient source.
Nutrient application was targeted to supply 200 kg N ha–1 yr–1 from one of three sources: (i) inorganic fertilizer as NH4NO3 broadcast in split applications in May and July, (ii) crimson clover (Trifolium incarnatum L.) cover crop plus supplemental inorganic fertilizer with half of the N assumed fixed by clover biomass and the other half as NH4NO3 broadcast in July, and (iii) chicken (Gallus gallus) broiler litter broadcast in split applications in May and July (Table 1).
|
Sampling and Analyses
Soil was sampled in April 1994 and February 1999. Unharvested and hayed exclosures were not established until July 1994, and therefore, were not part of the sampling scheme at initiation. In April 1994, subsampling points within grazed paddocks were on a 30-m grid. Due to the nonuniform dimensions of paddocks, subsampling points within a paddock varied from four to nine, averaging 7 ± 1. Point samples collected in 1994 (n = 122) were analyzed separately for C and N concentration and then data pooled before statistical analysis into approximate zones of 0- to 30-, 30- to 70-, and 70- to 120-m distances from shade to mimic zones used for collection in 1999. Pooling resulted in 2.3 ± 0.9 point subsamples per zone. In February 1999, subsampling locations within grazed paddocks were arranged along three semicircles, that is, at 5, 31 ± 3, and 72 ± 10 m from shade/water. Along each of these semicircles, three cores were randomly collected and composited. Distance from shade/water varied somewhat in each paddock, because of the different shapes of the paddocks and the intent to create three equally sized zones in each paddock. Soil cores (4.1-cm diam) were extracted with a hydraulic probe mounted on a tractor and sectioned into depth units at 0.15, 0.3, 0.6, 0.9, 1.2, and 1.5 m. Surface residue was scraped to the side before sampling. Soil samples with roots were air-dried and ground to <2 mm in a mechanical grinder in 1994. In 1999, soil with roots was oven-dried (55°C, 72 h) and initially crushed to pass an 8-mm screen and subsequently a portion of the sample ground to <2 mm in a mechanical grinder before total C and N determinations.
Soil was analyzed for total C and N concentration with dry combustion at 1350°C (Leco CNS-2000, St. Joseph, MI).1 Soil standards were used for calibration. It was assumed that total C was equivalent to SOC because soil pH was 
6.5 at all soil depths. Samples were analyzed shortly after collection each year.
To verify some unexpected changes in SOC and TN during the 5-yr period, where analyses were determined separately in 1994 and in 1999, we analyzed all samples from both years simultaneously using near-infrared spectroscopy (Dalal and Henry, 1986). This alternative analysis technique was chosen based on the ease of determination, low cost for the large number of samples, and availability of dry combustion data for comparison. Data of SOC and TN were presented only derived from dry combustion, while those from near-infrared spectroscopy were used for verification only.
To assess nutrient source and forage utilization effects, data from multiple samples within an experimental unit were averaged and not considered a source of variation using the general linear models procedure (SAS Institute, 1990). Only to assess the impact of animal behavior on potential accumulation of SOC and N, was a separate analysis of variance conducted using zone (i.e., 0–30, 30–70, and 70–120 m from shade/water) and its interactions with nutrient source and forage utilization as sources of variation. Cumulative SOC and TN contents with depth (i.e., integrated from surface to specified depth) were calculated assuming bulk density of 1.3 Mg m–3 at 0 to 0.15 m, 1.4 Mg m–3 at 0.15 to 0.3 m, and 1.5 Mg m–3 at depths below 0.3 m. Some variation in bulk density among treatments, replications, and years of sampling may have occurred, but we did not expect it to be significant relative to changes in SOC and TN concentrations. Assumed bulk density values were based on observations from a nearby pasture study (Franzluebbers et al., 2000). During the first 4 yr of this study, soil bulk density in the surface 6 cm was not affected by forage utilization (Franzluebbers et al., 2001). Changes in SOC and TN contents during 5 yr were calculated from the difference between values in 1994 and 1999. Significant changes in SOC and TN (i.e., different from zero) were determined from the least-square-means output. Analysis of variance was conducted separately by depth (i.e., both incremental and cumulative) for SOC and TN concentration and content determined in 1999 (i.e., end of 5 yr) and for differences in SOC and TN content between years (i.e., net change) according to the split-plot design with three blocks. All effects were considered significant at P 0.1. Although this was a lenient probability level, we did not want to overlook potentially important trends. Actual Pr > F values for effects were also reported for many effects in tables.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
When averaged across nutrient sources, effects of forage utilization were significant only for the 0- to 0.15-m depth for SOC concentration (Table 2) and for the 0- to 0.15-, 0.6- to 0.9-, and 0.9- to 1.2-m depths for TN concentration (Table 3). At a depth of 0 to 0.15 m, SOC and TN concentration were lower under haying than under unharvested and both grazing pressures (i.e., concentrations under haying were 74 ± 5% of those under other management systems). The lower SOC and TN concentrations with haying were also reported from this same study from shallow (0–6 cm) surface samples collected on a yearly basis (Franzluebbers and Stuedemann, 2001; Franzluebbers et al., 2001). Haying removed aboveground forage from the field, whereas unharvested and grazed pastures allowed return of either whole or partially digested forage to the soil where it could be converted to soil organic matter by soil microorganisms.
At a depth of 0.6 to 1.2 m, TN concentration was greater under unharvested than under grazed treatments when averaged across nutrient sources (Table 3). Deep forage rooting may have been enhanced when forage was allowed to accumulate aboveground vegetation. It may be that the slow decomposition of roots at these depths contributed to TN accumulation under unharvested management. In general SOC and TN responded similarly to management, but the detection of differences in TN with soil depth suggests that fertilization with an adequate level of N (i.e., 200 kg ha–1 yr–1) may have improved the nutritional status for soil microbial activity at lower depths, such that N moving through the profile could have stimulated mineralization of C and subsequent sequestration of organic N into stable organic compounds. Although differences in C/N ratio among treatments were rare at any depth within a sampling event, C/N ratio was reduced in 1999 compared with 1994 at all depths, except at 1.2 to 1.5 m (data not shown). The C/N ratio was 17.7 ± 1.5 in 1994 and 14.6 ± 3.1 in 1999 (among depths to 1.2 m). These data suggest that the input of fertilizer N reduced C/N ratio by increasing organic N in the profile.
Retention of forage on pasture either as unharvested conservation reserve or through season-long cattle grazing was important for increasing SOC and TN concentrations compared with forage removal (Tables 2 and 3). Removal of forage as hay not only reduced the input of C and N from residues at the soil surface, but also may have limited rooting depth and activity below the immediate soil surface due to senescent-inducing signals from regularly clipped aboveground vegetation (Morgan and Brown, 1983).
Lack of difference in profile SOC and TN between unharvested and grazed systems contrasted with greater SOC and TN under grazed than unharvested management at a depth of 0 to 6 cm (Franzluebbers and Stuedemann, 2001; Franzluebbers et al., 2001), suggesting that grazing had only superficial effects on SOC and TN. From five sites in the Great Plains USA, SOC content was greater under unharvested forage than under cropland at depth increments to 0.4 m, but not below this depth (Gebhart et al., 1994).
Soil Organic C and Total N Changes with Time Due to Management
Having determined initial profile SOC and TN allowed us to calculate the change in SOC and TN with time, although initial values were only available for low and high grazing pressure treatments under each of the nutrient source regimes. The single measurement of SOC and TN concentration at the end of 5 yr relied on randomly placed treatments to overcome the natural variability in these soil properties to determine relative management effects. Initial SOC and TN concentration are needed to directly estimate SOC and TN sequestration with time. In addition, temporal variations in SOC and TN can be more explicitly separated from spatial variations. Using LSD values at 0- to 0.3-, 0- to 0.9-, and 0- to 1.5-m depths (Mg ha–1), random variation of the absolute difference in SOC and TN (Table 4) between two times was 77 ± 24% of the relative difference in SOC (Table 2) and TN (Table 3) among treatments at one time. Therefore, not only was the difference in SOC and TN with time necessary to obtain rates of SOC and TN sequestration, it also reduced the LSD to be able to detect smaller differences among treatments.
|
Averaged across nutrient sources, there was no effect of grazing pressure on net change in SOC and TN with time at any depth (Table 4). However, significant sequestration of SOC and TN occurred under nearly all management systems to at least 0.3 m and sometimes to as much as 1.5-m depth. At a depth of 0 to 0.3 m, sequestration of SOC was 6.2 ± 2.0 Mg ha–1 (among nutrient source and forage utilization regimes) and sequestration of TN was 0.64 ± 0.17 Mg ha–1. Below the 0.3-m depth, differences in SOC between 1994 and 1999 among the 24 combinations of nutrient source-forage utilization-depth increment were less than zero in six cases and not different from zero in all other cases (data not shown). All of the declines in SOC concentration between 1994 and 1999 that were detected with dry combustion were corroborated with values that were equally or more negative using near-infrared spectroscopy. For TN, three cases were less than zero and two cases were greater than zero. Therefore with increasing soil depth below the surface 0.3 m, there was some, but overall, little tendency for loss of SOC and TN with time.
Sequestration of SOC in the 0- to 0.3-m depth on a yearly basis was 1.2 ± 0.4 Mg ha–1 yr–1 (among the six combinations of nutrient source and forage utilization) (Table 4). From yearly samples of the surface 6 cm of soil in this same study, the estimate of SOC sequestration was 1.4 ± 0.2 Mg ha–1 yr–1 among the same group of treatments (Franzluebbers et al., 2001), indicating that most C sequestration occurred near the soil surface. The two estimates of SOC sequestration agreed reasonably well, although both treatment and random variation were greater in the current investigation based on deep sampling. The LSD(P=0.1) for SOC sequestration at a depth of 0 to 0.3 m in the current investigation was 1.1 Mg ha–1 yr–1 (Table 4), while the LSD(P=0.1) for SOC sequestration at a depth of 0 to 6 cm in the study based on yearly analysis was 0.3 Mg ha–1 yr–1. Therefore, the more intensive sampling protocol of surface soil allowed a more precise estimate.
Although estimates of net change in profile SOC generally declined with depth, the switch from significant effects near the soil surface to nonsignificant effects at lower depths appeared to be as much due to increased random variation as due to less actual change. For every 0.3 m of soil below the plow layer, the change in SOC and TN content would have had to become an average of 0.6 and 0.10 Mg ha–1 greater, respectively, to maintain significance at P = 0.1. Therefore, a hypothetical change in SOC of 4 Mg ha–1 to meet significance at P = 0.1 would have been declared not significant if diluted with soil from deeper in the profile, even if no other change below this depth occurred. To achieve significantly positive SOC and TN sequestration at a depth of 0 to 1.5 m, rates 
1.3 and 0.14 Mg ha–1 yr–1, respectively, were required. Although possible, such high rates of SOC and TN sequestration are rarely achieved in forage management systems (Follett et al., 2001).
The coefficient of variation in SOC concentration among individual soil cores increased from 37 ± 13% in the surface 0.3 m to 83 ± 26% below the 0.3-m depth (Fig. 1). By combining data from the 7 ± 1 subsamples collected within a paddock in 1994 into three zones (i.e., 0–30, 30–70, and 70–120 m from shade/water), the coefficient of variation was reduced to 10, 8, 18, 40, 35, and 17% at 0 to 0.15, 0.15 to 0.3, 0.3 to 0.6, 0.6 to 0.9, 0.9 to 1.2, and 1.2 to 1.5 m, respectively. To obtain more precise estimates of SOC for a given management system, it seems necessary to composite a number of subsamples for adequate representation. Compositing cores in the field to represent a management unit would require the same amount of labor in the field as collecting and analyzing individual cores, but would save time and operating expenses in the laboratory without sacrificing accuracy of determination.
|
Soil Organic C and Total N Changes with Time Due to Pasture Zone
Profile SOC and TN contents were affected by pasture zone, primarily due to effects that occurred near the surface (Fig. 2 and 3). In the zone nearest shade/water (i.e., 0–30 m), the change in profile SOC and TN was significant at all depths, except below 0.3 m for SOC under low grazing pressure. At a depth of 0 to 0.15 m, SOC and TN sequestration were 3.8 ± 3.2 and 0.41 ± 0.23 Mg ha–1 greater than in other zones (i.e., 30–70 and 70–120 m from shade/water). Cattle spend more time near shade and water sources, depositing more feces and spoiling more forage in these areas. This animal behavior was a likely reason for preferential accumulation of SOC near shade and water sources, a result similar to that reported in other grazing studies (West et al., 1989; Franzluebbers et al., 2000).
|
|
Although inorganic N did accumulate in the soil profile during these 5 yr of some management systems (8 ± 8 kg ha–1 yr–1) (Franzluebbers and Stuedemann, 2003), the estimated portion of profile TN increase from inorganic N was 1 ± 14% (among nutrient source-pasture zone combinations). Most of the TN increase, therefore, was due to organic N accumulation.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication April 19, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. J. Franzluebbers Comments on "No-Tillage and Soil-Profile Carbon Sequestration: An On-Farm Assessment" Soil Sci. Soc. Am. J., March 1, 2009; 73(2): 686 - 687. [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Stuedemann and A. J. Franzluebbers Cattle performance and production when grazing Bermudagrass at two forage mass levels in the southern Piedmont J Anim Sci, May 1, 2007; 85(5): 1340 - 1350. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
