| Soil organic matter (SOM) is of primary importance for maintaining soil fertility, and is a key component of the global carbon cycle. Accurate estimation of the variation of SOM content is essential to soil fertilization, agricultural environment management and C mass balance in the global ecosystem. And agricultural management practices may significantly influence SOM properties, however, how spatio-temporal variation characteristics of SOM change with agricultural practices is poorly reported or understood. Although great efforts on understanding the changes in soil properties spatial variations have been made, it remains controversial on the estimates and trends of SOM changes in different places. Moreover, few studies have been dedicated to the spatio-temporal characteristics of SOM change in Daxian County, a typical topography area of parallel ridge-valley in Eastern Sichuan, China.In this work, the changes of SOM in Daxian County from 1982 to 2007 were estimated by different methods respectively, and to evaluate the relationship between environmental factors and SOM, or the relationships among soil fertility and SOM. This information was derived from the second national soil survey (SNSS) of China in early 1980s (1982) and the soil testing for formula fertility era in 2007, Furthermore, because of this data is the most comprehensive and detailed study on soil characteristics in DaXian County, It is necessary to understand the change characteristics during the last 25 years. The objective of this research was to evaluate the distribution of SOM in different times, analyzed the characteristics of the spatio-temporal variations, explored the causing factors, investigated several feasible suitable measures to increase the SOM according to the published papers, and paid attention to some essential procedures in the implement. The main objectives were to understand the spatio-temporal changes of SOM in the typical topography area and to evaluate the effect of the humane practices on the SOM changes. Results are as followed:1. Methods selectionAccording to the selected data from the Second National Soil Survey in China in the early 1980s (1982), a total of 77 soil profiles data, there is little published geo-information or coordinate information about them, by which restricted them used in the interpolating method directly; the Pedological Professional Knowledge based (PKB) method, interpolating method with administrated regional average SOM value (IMARA-SOM) were applied into calculating the distribution of SOM in 1982 of the area, compared with the standard method of original OM map printed in 1980s, results showed that PKB method maybe the best method for the prediction of the distribution of SOM in the 1980s.As the same time,3311 samples were selected from the database of soil testing for formula fertilizer, which combined with enough geo- and coordinate information. With the help of the traditional statistics by SPSS software, the result of studying on the spatial variability of SOM in 2007 showed that the content of SOM had a significant variation, especially among different soil and land use types. Moreover, Geo-statistics was applied in soil science to estimate the distribution of SOM, and to determine the spatial scales and forms of the soil variation, in order to give rise to the validity of sampling, it also can be used to study the process of the soil variability. Such as Ordinary Kriging (OK), Inverse Distance Weighted (IDW), Radial Basis Functions (RBF), Global Polynomial Interpolation (GPI), Local Polynomial Interpolation (LPI), et al interpolator methods, which used to indicating the variation of the SOM in the area of the research.These results compared with each other from Cross-validation (Cross-validation uses all of the data to estimate the prediction model, it removes each data location, one at a time, and predicts the associated data value.) and Validation (Validation creates a model for a subset of data and predicts values for the rest of the location.) under the parameters--Mean error (ME), Root mean square error (RMSE), Standard derivation (SD) and Accuracy ratio (AR) et al, indicating that prediction model measured errors. Results showed that IDW is the best interpolating method according to the data of selection in 2007 during the soil testing for formula fertilizer era, which has a good result with little prediction error, no assumptions required of the data and avoiding the shortage that all the data have an abnormal distribution type, which have been collided with methods with the assumption for normal distribution only directly.2. Characteristics analyzing in spatio-temporal variation of SOMResponses to different data in different times, each data had been calculated with classical statistical methods, above all, the results indicated that there was significant varies among these different times data, especially in the different soil and land utilization type.Just as the methods selection introduction, different time data in 1982 or 2007, have their best interpolating methods, PKB and IDW methods respectively. The results of the SOM spatio-temporal variation in Daxian County in past 25 years just under the overlaying analysis between the time of 1982 and 2007, hence come into a saptio-temporal variation map of SOM.And then, there conformed into several factors causing SOM variation under the statistics analyzed: since the second national soil survey beginning in 1982, the SOM of Daxian County was varied significantly; the SOM increasing was centrally distributed in zonal mountain area, decreasing was distributed dispersedly in the entire county. Combined with the district map of Daxian County, the whole research was divided into five subregions, such as Broad valley and shallow hilly area, Monoclinal deep or sub-shallow hilly area, Ridge-like mountain area, Berm-like mountain area and Zonal mountain area. For further research, regional spatio-temporal variation characteristics had been analyzed; proportion of SOM in each content level had been classified in 1980s was significantly varied, however, with a same decreasing trend and centrally range from 6.00 g/kg to negative 6.00 g/kg.3. Based on the intense spatio-temporal variation and notable decreasing trend of SOM in research, several measures were concluded:Come to the cultivated land, various agricultural practices such as residues retention, rotation, conservation tillage, fertilization and irrigation influence the agricultural SOM, what we can do just change the unsuitable practice actually. According to the non-agricultural land, the forests projecting and other bio-technological measures in water and soil conservation would be strengthened.The objectives of the paper are to estimate SOM spatial variation in typical topography area of parallel ridge-valley of eastern Sichuan, China and to study SOM spatio-temporal variation characteristics from 1982 to 2007, with 1982 as the base year. The results further pointed out that the PKB method is the suitable method for these data without detailed location information or Geo-information, which would be widely used while the method's improving. Results obtained from the study of SOM spatio-temporal variation would be not only useful for international comparative research, but also for more accurately estimating and monitoring of changes of SOM in typical topographic hilly area, southwestern China.
|
PDF Full Text Request