Impact of Agricultural Land Use on Soil Organic Carbon Sequestration at Sub-catchment Scale

D.D.A.E. Hemamali1, U.W.A. Vitharana2*, B.L.W.K. Balasooriya3, C.P. Attanayake2, W.S. Dandeniya2 and S.I. Nimanthi2 1Postgraduate Institute of Agriculture, University of Peradeniya, Peradeniya, Sri Lanka 2Department of Soil Science, Faculty of Agriculture, University of Peradeniya, Peradeniya, Sri Lanka 3Department of Biotechnology, Faculty of Agriculture and Plantation Management, Wayamba University of Sri Lanka, Kuliyapitiya, Sri Lanka


INTRODUCTION
Soils are the largest global terrestrial pools of organic C which stores about two times the carbon stored in above-ground biomass (Batjes and Sombroek, 1997). Soil organic carbon has received worldwide attention due to the important role played in the global C cycle and its potential feedbacks on the global warming. Carbon sequestration that describes the process of capturing and long-term storage of atmospheric carbon dioxide helps to mitigate global warming and avoid climate change.
The "4 per mille" is a global concept which intends to increase the soil organic matter and carbon sequestration through agricultural practices (Minasny et al., 2017). Therefore, estimation of the spatial patterns and stocks of SOC is essential for understanding the potential of soils to sequester C, and for quantifying the capacity of soils to serve as a SOC sink or source in changing environments (Mulat et al., 2018).
Anthropogenic interferences can alter the carbon cycle causing negative feedbacks on SOC stock. It is believed that land use change is the second main cause of carbon emission after fuel consumption (Watson et al., 2000).
The land use change and subsequent management are considered as key activities those alter the potential of soils to sequestrate C. However, the impact of agricultural land use on SOC sequestration have rarely been investigated in Sri Lanka. Yet, quantitative information of land use impacts on SOC sequestration is essential for making accurate spatial estimates of SOC stocks. Therefore, the present study was conducted to explore the changes of SOC stocks under long-term agricultural land uses compared to forest land to determine the interrelationships between SOC stock and topographic variables at subcatchment scale in upcountry of Sri Lanka.

MATERIALS AND METHODS Study area and site description
The study was carried out in a sub-catchment which comprised of agricultural lands and forest area in Seetha Eliya in the Wet Zone of Sri Lanka (central coordinates 6°56'13.87"N, 80°47'32.41"E). Contour digital maps (1:10000) purchased from the survey department were interpolated using local kriging interpolation technique to construct a Digital Elevation Model (DEM) of the study area. The procedure developed by Jenson and Dominique (1988) was used to delineate the sub-catchment area (Figure 1). The study area receives a mean annual rainfall of about 2500 mm and soil in this area belongs to the Red Yellow Podzolic great soil group which falls under the soil order Ultisols and sub order Udults according to soil taxonomy. Nuwara Eliya series is the dominant soil series found in this area (Mapa et al., 2009).

Soil sampling and laboratory analysis
Conditioned Latin hypercube sampling approach was used to pick 109 samples within the sub-catchment area. These soil samples were taken at 42, 44 and 23 locations in the vegetable (90 ha), tea (120 ha) and forest (140 ha) land uses, respectively. This sampling approach was use to select samples location while representing the spatial variability of potential covariates namely elevation, slope and land use ( Figure 1). A portable GPS receiver (GarminOREGON 650) was used to navigate the sampling locations. Soil samples were collected representing the surface (0-30 cm) and subsurface (30-60cm) soils at each location. Disturbed soil samples were collected at each depth using the gauge auger and the core sampler was used to obtain undisturbed soil core samples. Air dried samples were sieved using a 2 mm sieve to obtain the fine earth fraction. Then, a subsample of the fine earth fraction was ground using a mortar and pestle to obtain a powdered sample (particle diameter <0.125mm). Total organic carbon contents of powdered samples were determined using the modified Walkley and Black method (Nelson and Sommer, 1996). The fine earth fraction was used to determine the pH (1:2.5 soil to water suspension) and soil texture was determined using the pipette method (Gee and Or, 2002). Undisturbed samples were also taken with a core sampler for bulk density determination.

Calculation of bulk density, SOC stock and topographic attributes
The bulk density ( , Mg/m 3 ) and the SOC stock (SOCS, Mg/ha), stream power index (SPI) and wetness (WI) of a depth interval was computed using the following equations.
where, oven-dried mass of the core soil sample and its volume are given by W1 (Mg) and A (m 3 ) respectively. The thickness of the soil layer (m) and the percentage of organic carbon are given by T and C, respectively.
The wetness index reflects the tendency of surface flow water to accumulate at any point in catchment (Moore et al., 1988). It is defined as a ratio between the slope gradient in degrees ( ) and specific catchment area (As).
The stream power index is a measure of the erosive power of flowing water and calculated as (Wilson and Gallant, 2000); where Sl is measured in degrees.
Digital elevation model was used to obtain elevation values at sample locations. Further DEM was used to generate the slope map of the area by using the method developed by Wilson and Gallant (2000). ArcGIS 10.3 was used for topographic analysis.

Statistical analysis
All statistical calculations were performed with SPSS 20.0 software (SPSS Inc., USA), indicating the relationship between SOC stocks and topographic factors. The data were statistically analyzed by one-way analysis of variance (ANOVA), by considering the sampling sites as replicates (random effects) within a land use and land uses as treatments (fixed effects). The mean separation of the land uses was evaluated by Tukey's post hoc test at 95% confidence interval.

RESULTS AND DISCUSSIONS
Effect of land use on soil physical and chemical properties Table 1 shows summary statistics of measured properties in surface and subsurface soil in three land uses.

Effect of topographic factors on SOC stocks at sub-catchment scale
The observed Pearson's correlation coefficients between the SOC stocks and the topographic variables namely the elevation, slope, SPI and the WI are presented in Table 2. None of the correlations were significant at p<0.05, indicating that the variability of SOC stock is not affected by the topographic factors.

Impact of land use on SOC stocks
The variability of surface and subsurface soil organic carbon stocks in relation to land use are presented in Table 3.
The highest average surface SOC stock was observed in the forest soil (71.8 Mg/ha), which was followed by vegetable (61.77 Mg/ha) and tea (50.7 Mg/ha). Addition of organic matter in the form of leaf litter, longer residue time in soil due to less soil disturbances, subsequent accumulation of higher organic carbon and stabilization within aggregates could be attributed to largest SOC stocks in forest soils (Haynes, 2005). However, the surface SOC stock in forest soil did not show any significant difference compared to that of the vegetable soils. This was partly due to the higher heterogeneity of SOC stock in forest lands (CV=36.8%) in comparison to the vegetable fields (CV=24.6%). Surface application of organic matter such as poultry manure to vegetable lands is a common practice in the upcountry wet zone. This could have largely attributed for the accumulation of organic carbon in vegetable soils though the soil texture reveled a considerable soil erosion. Greater C storage in comparatively lower clay containing vegetable soils further highlight the consistent external input of C to the soil in vegetable lands. Tea grown soils showed the lowest SOC stock at the surface soil (p<0.05). Tea grown soils are subjected to soil erosion than that of forest and also application of organic manure is not a common management practice in tea cultivation. The application of organic manure is rarely practice in tea cultivation. Further disturbance caused by anthropogenic activities such as removal of leaves, pruning, weed control could reduce capacity of C sequestration in tea lands. These reasons could have led to lowest SOC stock observed in tea soils in the surface layer.
Forest (69.2 Mg/ha) soil showed a significantly larger subsurface SOC stock (p<0.05) than those of tea (65.6 Mg/ha) and vegetable (57.9 Mg/ha). The mean SOC stock gradually decreased with soil depth in vegetable and forest land use. Interestingly a larger subsurface SOC stock was observed in tea land use. This further justified the depletion of SOC stock in the surface soil of tea. The subsurface store of SOC could be attributed to translocation of organic matter from surface soil and transport of clay-bound organic matter, plant roots and root exudates, dissolved organic matter and bioturbation (Rumpel and Kogel-Knabner., 2011). The soils under forest (69.4 Mg/ha) had significantly highest SOC stock in 0-60 cm depth than vegetable (59.8 Mg/ha) and tea (57.8 Mg/ha) land uses (Figure 2). Soil organic carbon stocks in vegetable and tea lands were 13.8% and 16.8% lower as compared to soils under forest. These findings are in agreement with those of Geraei et al. (2016), who observed 36% depletion in SOC after the conversion of forest to agriculture lands and attributed the loss to rapid mineralization of SOC resulting from intensive tillage. Conversion of forest lands to agriculture has been claimed to damage the soil structure and enhance mineralization of organic matter. Damaged soil structure exposes organic matter which is no longer physically protected by the aggregates from microbial decomposition (Haynes, 2005;Islam and Weil, 2000).

CONCLUSION
This study concludes that the topographical factors (elevation, slope, SPI and WI) have no significant influence on variation of SOC stocks in the studied sub catchment area. The land use type is a key determinant of the spatial heterogeneity of SOC and forests have the greater ability of sequestering C compared to tea and vegetable land uses. Conversion of forest land in to agriculture may exert a potential threat of reducing SOC stocks by 13.8% and 16.8% for vegetable and tea cultivation, respectively.

AKNOWLEDGEMENT
Financial assistance from Sri Lanka Council for Agricultural Research Policy (Grant No NARP/16/UP/AG/02) is acknowledged for the financial support.