【作者】 吴若玉；

【导师】 张世熔；

【作者基本信息】 四川农业大学 ， 土壤学， 2007， 硕士

【摘要】 本文在收集区域自然条件、社会经济发展资料和化验分析野外采集的88个土样所得数据的基础上，基于GPS和GIS技术平台，用普通克里格、泛克里格和析取克里格方法研究了金堂县环溪河流域由小到大的A、B、C3个丘陵区土壤全钾和速效钾含量的空间分布特征及其影响因素。主要结论如下：（1） A、B、C3个区域土壤全钾含量随区域的扩大而增加，速效钾含量随区域的扩大先减小后增大，但二者在3个不同大小区域间的含量差异皆不显著。A、B、C3个区域土壤全钾平均含量分别是17.96±3.48 g kg^-1、18.12±3.46 g kg^-1、19.47±2.90 g kg^-1，土壤速效钾平均含量依次为131.43±63.03 mg kg^-1、125.74±52.48 mg kg^-1和135.63±48.43 mg kg^-1。（2）通过误差分析得出，土壤全钾含量在A区和C区以析取克里格高斯模型拟合效果最好，B区以普通克里格高斯模型拟合效果最好；土壤速效钾含量在A、B、C区均以普通克里格球状模型拟合效果最好。用所得的最优模拟模型分别对3个区域土壤钾素含量进行模拟，从模拟结果看，3个区域土壤全钾和速效钾含量均呈块状或条带状分布。其中，A区土壤全钾和速效钾含量均呈现以西部和东南部丘陵中下部为高值中心，向东北部低值区逐渐减少的趋势；在B区，土壤全钾含量的高值区出现在西北部，速效钾含量的两个高值区出现在东南部和西部，二者低值区均分布于东北部；在C区，土壤全钾含量呈现以西北部和中部为高值中心向两侧逐渐减少的趋势，两个低值区集中分布于东部和西南部，速效钾含量呈片状交错分布，高值区出现在东南部，低值区分布于东北部。分级面积统计表明，随着研究区面积的扩大，土壤全钾含量的低值区所占比例逐渐减小；速效钾含量随区域的扩大向120～160 mg kg^-1含量级集中。（3）成土母质、土壤类型和微地形条件等自然因素对土壤全钾和速效钾含量的影响在不同大小区域总体上差异显著，个别区域差异不显著。其中，在不同成土母质下，土壤全钾含量在A区差异不显著(F_2，45=2.001，P=0.147)，B区达到显著水平(F_2，45=3.828^*，P=0.037)，而在C区达到极显著水平(F_2，45=8.012^**，P=0.002)；速效钾含量在A区和C区均存在显著差异，F值分别为3.775^*和3.843^*，B区存在极显著差异(F_2，45=7.785^**，P=0.003)。在不同土壤类型下，土壤全钾含量在A区和C区均存在极显著差异，F值分别为4.698^**和5.142^**，B区不显著；速效钾含量差异在A区和B区均达到极显著水平，F值分别为4.314^**和5.284^**，C区不显著。在微地形条件下，丘陵顶部土壤全钾和速效钾含量显著低于丘陵中部、下部和冲沟，而后三者之间差异未达到显著水平。（4）土地利用方式、轮作方式、施肥状况和秸秆还田水平等人为因素对土壤全钾和速效钾含量的影响程度各不相同。从总体上看，各因素下土壤全钾含量差异不显著；速效钾含量差异较为显著。其中，在不同土地利用方式下，A、B、C3个区域土壤全钾含量差异均不显著；速效钾在A区达到极显著水平，B区达到显著水平(F_3，21=4.813^*，P=0.011)，C区不显著(F_3，20=1.277，P=0.308)。在不同轮作方式下，土壤全钾和速效钾含量差异均未达到显著水平，但速效钾含量变化较全钾明显。在不同施肥状况下，土壤全钾含量差异不显著，速效钾存在极显著差异(F_3，84=4.275^**，P=0.008)。在不同秸秆还田水平下，土壤全钾和速效钾含量的差异均不显著。（5）在土壤性质方面，相关分析结果表明，土壤全钾与pH和有机质之间均无显著的相关性，r值分别为0.154和0.026；速效钾与pH相关关系不显著（r=0.167，n=88），而与有机质呈极显著正相关关系(r=0.363^**，n=88)。更多还原

【Abstract】 Geostatistics combined with GIS was applied to analyze spatial distribution andinfluence factors of soil total potassium （STK） and soil available potassium （SAK） for 88soil samples in the Huanxi River Valley in Jintang that which was divided into areas A, Band C of different acreages. The main results were as follows:（1） Among the three areas, the contents of STK and SAK were not of significantdifference, and with study area extended, STK content increased while SAK contentdecreased originally and then increased. Distribution types of STK and SAK in areas A, Band C were all normal distribution. The average contents of STK were 17.96±3.48 gkg^-1,18.12±3.46 g kg^-1 and 19.47±2.90 g kg^-1. SAK contents were 63.03 mg kg^-1, 74±52.48mg kg^-1 and 135.63±48.43 mg kg^-1.（2） The spatial distribution of STK and SAK contents showed different scale effectthat they had different semivariograms parameters, error analysis, and spatial distributionin the three areas. According to ArcGIS9.0 software plat, the paper adopted gaussian,spherical and exponential models in three interpolation methods of OK, UK and DK, andstudied spatial variability of STK and SAK. Then cross-validation and model testing wereused for error checkout. The results showed that in areas A and C, in term of STK, gaussianmodel of DK was the optimal interpolation model and in area B, gaussian model of OKwas the optimal interpolation model. In term of SAK, spherical model of DK was theoptimal interpolation model. From optimal interpolation models in the three areas, for STK,the high value district in area A was nubby distribution in the west and middle, and the lowvalue district was in the northeast. High value district in area B was in the northwest andSTK content reduced to the south and east. Low value district in area C was nubbydistribution in the northeast and southwest. For SAK, high value district in area A wasnubby distribution in the west and middle and lower of hills of the southeast and low valuedistrict was in northeast. Low value district in area B was zonal distribution in the east.High value district in area C was in the southeast and low value district was in the northeast. Statistical results of classification areas show that with research area extended,the proportion of low value district of STK became smaller and the proportion of highvalue district became bigger, while SAK content was concentrated between 120 mg kg^-1and 160 mg kg^-1.（3） The influence on contents of STK and SAK caused by differert natural factorswere different. STK content of different parent materials was not significant in area A (F_2,45=2.001, P=0.147), but was significant in area B (F_2,45=3.828^*, P=0.037), was extremelysignificant in area C (F_{2, 45}=8.012^**, P=0.002). SAK content of different parent materialswas significant in both areas A and C according to F value 3.775^*and 3.843^*, and wasextremely significant in area B (F_2,45=7.785^**, P=0.003). STK content of different soiltypes was extremely significant in both areas A and C according to F value 4.698^**and5.142^**, but was not significant in area B. SAK content of different soil types wasextremely significant in both areas A and B according to F value 4.314^**and 5.284^**, butwas not significant in area C. STK and SAK content in the top of hill were lower than theother parts of hill and gully in micro-terrain situation, but was not significant among latter.（4） The influence on content of STK and SAK that caused by human actions weredifferent between each other. Totally, STK content was not significant, but it was relativelysignificant in SAK. It showed as follows: STK was not significant of different land usemanagements in all of the three areas, but SAK was extremely significant in area A, andwas significant in area B (F_{3, 21}=4.813^*, P=0.011), but was not significant in area C (F_3,21=1.277, P=0.308). STK and SAK content were not significant of different rotationsystems in all of the three areas, but a little change can be observed in SAK. STK contentwas not significant of different fertile managements, but SAK content was extremelysignificant (F_{3, 84}=4.275^**, P=0.008). STK and SAK content were both not significant ofdifferent vegetation recycle managements.（5） The influence on content of STK and SAK caused by pH and soil organic matter（SOM） were different. According to correlation analysis, STK showed no correlation to pH（r=0.154, n=88） and SOM （r=0.026, n=88）. SAK showed no correlation to pH （r=0.167,n=88）, but showed extremely significant correlation to SOM (r=0.363^**, n=88).更多还原