Spatial assessment of soil organic carbon and physicochemical properties in a horticultural orchard at arid zone of India using geostatistical approaches
Spatial assessment of soil organic carbon and physicochemical properties in a horticultural orchard at arid zone of India using geostatistical approaches
复制标题
DOI：
10.1007/s10661-016-5522-x
复制
发表时间：
2016-08
影响因子：
3
通讯作者：
A. Singh;P. Santra;Mahesh Kumar;N. Panwar;P. Meghwal
中科院分区：
文献类型：
--
作者：
A. Singh;P. Santra;Mahesh Kumar;N. Panwar;P. Meghwal
土壤有机碳（SOC）是农业生产系统长期维持的重要指标。除了维持生产力，有机碳在气候变化的背景下也起着至关重要的作用。考虑到这些潜力，本研究通过地质统计学方法评估了位于印度干旱区由几个干旱果园组成的果园有机碳含量的空间变化。为此，从占地14.33公顷的果园的175个地点收集了地表和地下土壤样本，并收集了地理坐标。测定土壤样品的有机碳含量和土壤理化性质，并进行地统计分析，用于制图。在0 ~ 30 cm土层，果园平均有机碳储量为14.48 Mg ha - 1，范围从红桃9.01 Mg ha - 1到红桃19.52 Mg ha - 1。土壤有机碳含量的空间变异范围约为100 m， pH和电导率等土壤理化性质也表现出相似的空间变化趋势。这表明，为了获得更好的精度，未来SOC制图程序的最小采样距离可以保持在100 m以下。普通克里格技术对未采样点土壤有机碳含量的预测结果满意，均方根残差（RMSR）为0.35 ~ 0.37。由于土壤有机碳含量与pH、EC之间存在显著的相关性，因此协同克里格法在土壤有机碳含量的空间预测上略优于普通克里格法（RMSR = 0.26 ~ 0.28）。以90%的置信区间给出了SOC估计的不确定度。通过普通克里格或共同克里格方法估算土壤有机碳储量的空间不确定性也比非空间估算方法（如算术平均方法）低。在果园不同的果块人工林中，有黑木的果块土壤有机碳密度最高。
Soil organic carbon (SOC) is a major indicator of long-term sustenance of agricultural production system. Apart from sustaining productivity, SOC plays a crucial role in context of climate change. Keeping in mind these potentials, spatial variation of SOC contents of a fruit orchard comprising several arid fruit plantations located at arid region of India is assessed in this study through geostatistical approaches. For this purpose, surface and subsurface soil samples from 175 locations from a fruit orchard spreading over 14.33 ha area were collected along with geographical coordinates. SOC content and soil physicochemical properties of collected soil samples were determined followed by geostatistical analysis for mapping purposes. Average SOC stock density of the orchard was 14.48 Mg ha−1for 0- to 30-cm soil layer ranging from 9.01 Mg ha−1inCarissa carandasto 19.52 Mg ha−1inProsopis cinerariablock. Range of spatial variation of SOC content was found about 100 m, while two other soil physicochemical properties, e.g., pH and electrical conductivity (EC) also showed similar spatial trend. This indicated that minimum sampling distance for future SOC mapping programme may be kept lower than 100 m for better accuracy. Ordinary kriging technique satisfactorily predicted SOC contents (in percent) at unsampled locations with root-mean-squared residual (RMSR) of 0.35–0.37. Co-kriging approach was found slightly superior (RMSR = 0.26–0.28) than ordinary kriging for spatial prediction of SOC contents because of significant correlations of SOC contents with pH and EC. Uncertainty of SOC estimation was also presented in terms of 90 % confidence interval. Spatial estimates of SOC stock through ordinary kriging or co-kriging approach were also found with low uncertainty of estimation than non-spatial estimates, e.g., arithmetic averaging approach. Among different fruit block plantations of the orchard, the block withProsopis cineraria(‘khejri’) has higher SOC stock density than others.