Tobacco Leaf's Chemical Quality–based Spatial Variability And Differential Management Of Soil Nutrients For Tobacco Field

Generally, chemical quality, physical properties and sensory indicators were used to assess the quality of tobacco, but the chemical composition had no influence by people's subjective judgment. the quality of the same soil in different depths or on the same plane was not entirely homogeneous, the physical properties of soil at all points of the value was also not the same. Finding out soil properties of tobacco fields'especially nutrient spatial variability could manage soil nutrient and reasonable fertilization. In turn, variability of soil nutrients caused tobacco growth. Therefore, studying on spatial variability of soil nutrients could be reasonably developed and improved nutrient utilization. In this experiment, GPS was used to take soil sample in Nan yang Technology Garden, the interval was 40*40m. the nutriment of leaf and soil was combined with the chemical quality in order to selected related degree between tobacco lead and soil nutrients, GIS and geostatistics method were used to analyze the spatial patterns of soil nutrients, and the pattern of soil nutrients was mapped. The results showed as follows:(1) The contents of carbon compounds of leaf from the study areas was slightly higher than the suitable range; the contents of nitrogen compounds was in the appropriate average range; potassium, chloride and their ratio were basically suitable, the relationship of acid and base was also suitable(2) The chemical constituents in tobacco leaf samples were analyzed, and four leaf chemical components factors were extracted, accounting for total variance of ten kinds of chemical components was 89.839%, and the feasibility of using factor analysis was confirmed by a stone map. These four factors were classified as integrated factor, carbon compounds factor, nitrogen compounds and combustion factor, and at last, the overall score of the chemical composition of each sample point leaves were given.(3) The average contents of soil available N, P and K were 114.270mg/kg, 10.737mg/kg and 135.780mg/kg, and were basically suitable; the average contents of total phosphorus 0.663g/kg, was lower than suitable range, while the average contents of total K 27.574 g / kg was higher than suitable range; the contents of soil organic matter was 14.798g/kg. the variation of coefficient of soil organic matter, available N, total K, available K and effective copper were also between 11.866% -23.883%; but that of Available phosphorus, total phosphorus, available Fe, Mn and Zn were between 33.484% -60.895%. By frequency analysis of the soil nutrient, the nutrients were divided into different levels except organic matter and effective copper. This result showed that the soil showed a greater spatial variation, it is necessary to differentiate its management.(4) The relation degree between the soil properties and chemical composition of tobacco leaf samples were analysised by used gray correlation analysis method. The results showed: the correlation coefficient of the soil potassium, phosphorus, organic matter, total K, available Zn and Cu was respectively 0.665,0.663,0.632,0.624,0.616 and 0.607.(5) The data of soil total K, available Zn was treated by Cox-box and logarithmic transformation, all of them accorded the normal distribution. statistical analysis of soil nutrients was taken by using the GS + statistical soft, the results showed that the value of nugget effect of soil available P and effective copper was less than 25%, and it showed a strong spatial correlation and indicated that phosphorus and copper in the structural factors of the total valid variance larger proportion; the value of Co / (Co + C) of soil available K, organic matter, total K and Zn of ratio was during 25% and 75%, it showed moderate spatial correlation and indicated that the spatial variation was decided by random factors and structural factors .(6) On Arcgis platform, the distribution pattern of soil nutrients of the study area was accurately described. The study area was made partitions by using hierarchical clustering method, and the result of variance analysis showed that each area of the soil nutrient divided into three management units was the best, but combining agricultural production with input-output ratio, it was considered that divided into two management units was the best method.