1. In natural grassland ecosystems, root-feeding nematodes and insects are the dominant below-ground herbivores. In coastal foredunes, the ectoparasitic nematode Tylenchorhynchus ventralis would be a major root herbivore if not strongly controlled by soil microorganisms. Here, we examined if the suppressive effects of the microbial enemies of T. ventralis act by local interactions such as predation, parasitism or antagonism, or local induction of plant defence, or by non-local interactions, such as systemic effects when microorganisms in one section of the plant roots can affect nematode control in another section of the root system. We show that abundance of T. ventralis in the root zone of the grass Ammophila arenaria is suppressed by local interactions. 2. We compared local vs. non-local control of nematodes by a natural community of soil microorganisms in a split-root experiment, where nematodes and microbes were inoculated to the same, or to opposite root compartments. 3. The split-root experiment revealed that microorganisms affected T. ventralis numbers only when present in the same root compartment. Therefore, the effects of microorganisms on T. ventralis are due to local interactions and not due to induction of a systemic defence mechanism in the plant host. 4. When inoculated together with microorganisms, the nematodes were heavily infected with unknown bacteria and with fungi that resembled the genus Catenaria, suggesting that microorganisms control nematodes through parasitism. However, local defence induction cannot be completely excluded. 5. Besides microbial enemies of nematodes, the root zone of A. arenaria also contains plant pathogens. Root biomass was reduced by nematode infection, but also by the combination of nematodes and microorganisms, most likely because the soil pathogens overwhelmed the effects of nematode control on plant production. 6. We conclude that there may be a trade-off between beneficial effects of soil microorganisms on the plant host due to nematode control vs. pathogenic effects of soil microorganisms on the plant host. We propose that such trade-offs require more attention when studying below-ground multitrophic interactions.