Abstract: The samples of surface soil, deep soil, and vegetables were collected from the liver cancer’s high and low incidence areas in Zhujiang River Delta to study the relationships between soil Mn forms and vegetables’ Mn enrichment. The results showed that the soil Mn in study area was mainly derived from parent materials, and rarely come from human activities. The average soil Mn content in liver cancer’s high incidence area was 577.65 mg·kg^-1, being significantly lower than that of liver cancer’s low incidence area (718.04 mg·kg^-1) and whole country (710 mg·kg^-1). The Mn forms in high incidence area were mainly of residual Mn and FeMn oxide, and less of water soluble Mn and exchangeable Mn, with the sum of the latter two’s distribution coefficients being not higher than 4%. In low incidence area, the distribution pattern of soil Mn forms was similar to that in high incidence area, but the absolute contents of the Mn forms were significantly higher. Soil total Mn and soil pH had significant effects on soil Mn forms. There existed significant positive correlations between soil total Mn and the Mn forms of Fe-Mn bound, humic acid bound, carbonate bound, and residual, and negative correlations between soil pH and soil water soluble and organic bound Mn forms. Among the test five kinds of vegetables, Youmai lettuce and Chinese cabbage in liver cancer’s high incidence area had a significantly lower Mn content than in low incidence area, while the other three had less difference. The Mn enrichment in test vegetables was positively correlated with to the content of soil available Mn (sum of water soluble Mn and exchangeable Mn), but had no correlations with the contents of soil total Mn and other Mn forms.

Key words: Phyllostachys praecox, organ, silicon, nutrient element, interreaction.