Reply from

In his recent TREE news & comment, Watkinson1 drew attention to the role of soil microorganisms in plant population dynamics. In particular, he reported on the dynamical framework for the inter-relations between the composition of plant and soil communities, proposed last year by Bever et al.2 It is worth adding allelopathic interactions to this picture. Allelopathy has been defined by Rice3 as ‘any direct or indirect harmful or beneficial effect by one plant (including microorganisms) on another through production of chemical compounds that escape into the environment’. If we apply this definition to Bever et al.’s framework for the feedback interaction between the soil community and two plant species, at least two further aspects come into play. The first concerns the potential mechanisms for positive and negative feedbacks. We can hypothesize direct interactions, such as mycorrhizal systems, for positive feedback (the fungal symbiont allows plant species to explore more soil resources, thus augmenting the autotrophic community) and pathogens for negative feedback. But we can also hypothesize indirect interaction: the plant produces allelochemicals that are metabolized by soil microorganisms4, leading to the release of compounds into the soil that might affect (positively or negatively) the plant species5. The second concerns the feedback model proposed by Bever et al.2 The authors did not depict any direct or indirect interaction between the two plant species in their model. Nevertheless, one could add connections between them because of the potential occurrence of direct allelopathic interactions among plant species. I agree with Watkinson’s conclusion that ‘the soil community is something that plant population biologists can no longer ignore’ (if indeed they do), but they should not ignore allelopathy either.