However, the soil biota in the three soils did differ in their effects on leachate phytotoxicity in the sterilization experiment . The soil biota in the evergreen broad-leaved forest soil eliminated phytotoxicity entirely, whereas the soil biota in the deciduous broad-leaved forest soil only reduced phytotoxicity, and leachate treatment in the roadside soil promoted growth in the presence of soil biota. Recently, Thorpe et al. found that root exudates from the invasive weed Centaurea maculosa were allelopathic to native species in its invaded range but not in its native range. This lack of allelopathic effect in the native community may have been due to biological degradation of allelochemicals by native soil microbes, rather than, or in addition to, adaptation by native plants. It is surprising that we did not detect either of the two primary allelochemicals in either rhizosphere or non-rhizosphere soils infested with E. adenophorum. Quantification and stability analyses of allelochemicals explained the reason for the disparity between previous studies of leachate phytotoxicity by laboratory bioassay and our experiments in soil. 9-Oxo-10,11-dehydroageraphorone and 9b-Hydroxyageraphorone decomposed by c. 3�C20% within 24 h in soil but not in sand, while the relative abundance of another compound, Di-n-octyl phthalate, greatly increased concurrently, suggesting that soil microbes may transform 9Oxo-10,11-dehydroageraphorone or 9b-Hydroxyageraphorone into Di-n-octyl phthalate. Zhang demonstrated that Di-noctyl phthalate occurs in the root exudates of E. adenophorum, but the biological activity of this compound is not clear. Soil microbes and E. adenophorum may contain the same or similar enzymes that catalyze a series of chemical reactions to make this change occur. Actually, there are many factors that could influence the results of the soil experiment and the sterilization experiment, such as the composition of the leachate and the environmental conditions. Zhang found that besides the two main allelochemicals, there are some other organic compounds in E. adenophorum leachate, which are possibly act alone or in combination to influence the growth of native plant species more or less even if they are not allelopathic. In different phenological stage, the composition and concentration of all kinds of chemicals including allelochemicals in E. adenophorum leaves are different, thus may cause a little difference in the leachates even if the extraction methods are the same. This can subsequently result in the different results in non-sterile soils between the soil experiment and the sterilization experiment. Also, the infiltration and spread efficiency can vary among field soil types or even the same type of soil in different seasons of the year. For example, the leachate which was poured into the soil may be diluted by soil water in more moist soils. Furthermore, the phytotoxicity of any chemical may vary with extraction methodology, analytical techniques and the vagaries of experimental application procedures. So it is not curious to find some instability of the data shown here. Our findings from both protective against manipulative greenhouse experiments and chemical analyses contribute to the growing body of evidence for the role of soil biota in plant interactions. Such complex effects depend on the presence of all interacting parties the invader, native plants, and soil microbes.