Research Of Soil Spatial Heterogeneity And Soil Sampling Method In Karst Area

Soil sampling is the fundamental of Soil Research, and it is also a major source of error in researches about soil. To improve the accuracy of soil sampling is the biggest obstacle to soil science development in the tendency that Soil Science develop towards more precise, although there have been a lot procedures of soil sampling. So to study and update of soil sampling method has become to the barrier which must be overcome with the development of Soil Science. In this thesis, five plots in three type of plot-karst rocky desertification de-farming slop, karst forest plot and dolomite grassplot was researched as the research object. The repeated result of analog sampling by familiar soil sampling method was used to compare with each other in this thesis based on the research of spatial heterogeneity of soil in these plots. The results and conclusions was summarized as follow:1. The difference between each microhabitat SOC of limestone plot in karstarea was significant. It is advisable to classify the soil of limestone plotaccording to the microhabitat.2. Effects of the quantity of sampling points on the accuracy rate and percentof accuracy is much more than different sampling methods.3. The regression result of sample’s accuracy rate and accuracy with thequantity of sampling points was as follow.Sampling method Sample’s accuracy rate Sample’s accuracyRegression model Coefficient of determination Regression model Coefficient of determination“S” type y=0.4653e4.9053 x R2=0.8573 y=0.8058e4.5397 x R2=0.7957Plum method y=0.5433e4.7505 x R2=0.8508 y=0.666e4.9282 x R2=0.8952Random method y=0.5438e4.9426 x R2=0.9498 y=0.7058e5.0536 x R2=0.8902Grid method y=0.4016e5.3666 x R2=0.9101 y=0.6443e5.1035 x R2=0.8728 The microhabitat method y=0.5104e4.5613 x R2=0.6714 y=0.8731e4.6995 x R2=0.9039 4. From the view of sample’s representativeness and reproducibility, themicrohabitat method and the “S” type method are the two methods whichneed the least sampling points. 5. Considering about representativeness and reproducibility of sample andthe method’s maneuverability, the microhabitat method is the mostsuitable method for the plot which has a small area and in which themicrohabitats distribute intensively and obviously when sampling in karstarea. For those plot which has a large area or in which the microhabitatsdon’t distribute intensively and obviously. 6. To achieve 80% sample accuracy rate and 70% sample accuracy, at least23 sampling points is needed when the microhabitat method was used forsoil sampling in 400m2 karst plot. 24 sampling points is needed when the“S” type method was used.