Objectives Conducting analysis of spatial variation and impacting factors of trace elements and understanding the interior mechanism of the phenomenon are needed for the effective management of soil fertility, maintaining of human health and protection of soil ecological enviroment.

Methods Using grid placement method, 283 soil samples (0–20 cm) were collected in Poyang Lake plain, Jiangxi Province. 20 samples with abnormal values beyond triple folds of standard deviation were screened off. The pH, texture, N, P, K and 6 trace element（B, Mn, Mo, Se, I, F) contents were determined. Using geostatistic theory, variance analysis, correlation analysis and regression analysis methods to analysis the relationship among the trace element contents, spatial distribution variation and the impacting factors were analyzed.

Results The average contents of B and Mo are between abundant and upper limit. The Mn is 330.24 mg/kg, which is deficient. The Se and I are at moderate level. And the F is 474.59 mg/kg at marginal level. All the 6 elements have moderate variation, but the variation degree of each element is different. The optimal interpolation models of B, Mn and Mo are exponential, while Se, I and F are spherical. The nugget effect of I is less than 25%, while the other five elements are between 27.09% and 41.77%. The spatial distribution pattern of the six trace elements are comparatively different. Except that the content of Se isn’ t deficient in the whole area, the other five elements all are deficient. B, Mn, F elements are affected by soil type, topsoil texture significantly. The correlation between Mn, Mo and slope, pH reached extremely significant levels. Mn, Se, I and the total P showed significant correlation. There are no correlation between the slope direction and the distance to the river with the six elements. Other factors only show significant or extremely significant correlation with a few elements. The topsoil texture has the greatest influence to B and I among their main factors, pH contributed the most to the content change of Mn and Mo among the main factors. The content of Se reach significant or extremely significant correlation with soil type, total P, slope and elevation, with the total P has the highest impact. The correlation between the content of F and soil type, topsoil texture and total N are significantly, and total N was the main influencing factor.

Conclusions 1) The total contents of soil trace elements B and Mo are relatively high of the study area, the I and Se are at the moderate level, and the Mn is relatively deficient overall. The F is at the marginal level. All the six elements show moderate variation. 2) The optimal interpolation models of soil trace elements B, Mn and Mo are exponential, while those of Se, I and F are spherical. Except for the consuming spatial autocorrelation of element I, the other five elements show moderate level. 3) Among the all impacting factors of trace elements, there is no significant correlation between the slope direction, the distance to the river with the content of each element. The element B and I are most affected by topsoil texture, pH influenced Mn and Mo mostly, Se is mostly impacted by total P, while F is mainly affected by total N.