土壤养分变异及合理取样数的初步研究

A Preliminary Studies on the Soil Nutrient Variability and Rational Sampling Quantity

利用地统计学方法、地理信息技术对土壤养分变异及合理取样数做了初步研究。结果表明大部分土壤养分都具有良好的半方差结构及强烈的空间相关性。统计分析结果表明土壤养分变异越大,相应要达到一定精确度的取样数量也会增大,研究区域各种养分变异系数大小及其取样数量的排列顺序基本一致,为速效磷>全氮>碱解氮>有机质>速效钾。用纯随机取样法初步探讨了研究区域土壤养分合理取样数。速效磷取样以90%置信水平20%相对误差为宜,取样数量为4个;速效钾取样以10%相对误差为宜,取样数量为5个;碱解性氮取样以20%相对误差为宜,取样数量为1个;有机质取样以10%相对误差为宜,取样数量为5个;全氮取样以10%相对误差为宜,取样数量为14个。

Soil nutrient variability and rational sampling was preliminary studied using Geo-statistics and GIS. The results showed that most soil nutrients possessed semi-variance structure and space relevance. Statistics results showed that sampling quantity increased with soil nutrient variability increasing. The arrange order of Cv. of soil nutrients was accordance with rational sampling quantity; the order was available P, total N, available N, OM and available K. The rational sampling quantity could be determined by using Cocharan in this study area. The rational sampling quantity for available P would be 5 at 0.1 level and 20% relative error, K would be 5 at 0.1 level and 10% relative error, N would be 1 level at 0.1 and 20% relative error, OM would be 5 at 0.1 level and 10% relative error and total N would be 5 at 0.1 level and 10% relative error.