Both plant-soil feedbacks (PSF) and plant competition drive plant community assembly, but their interactive effects have rarely been investigated, and the role of community composition in modulating these interactions is unknown. We conducted a fully reciprocal experiment with two plant species (Polygonum criopolitanum and Carex thunbergii) and their associated soil biota communities to untangle the relative importance of PSF and competition. The two species were grown either in monoculture or in mixed plantings, with various densities of each plant's neighbor, and either in the presence or absence of the soil biota associated with either species. When grown individually, each plant species showed a neutral PSF in the presence of its own soil biota, while feedbacks in competition were predominantly negative. P. criopolitanum produced more biomass in the presence of soil biota than in autoclaved soil, but the opposite was observed for C. thunbergii. In competition, both plant species were suppressed by its presence of their own soil biota, and neither had a competitive advantage, resulting in a significant negative pairwise PSF in most cases. C. thunbergii also showed stronger per capita effects than did P. criopolitanum. The predication on species coexistence largely depended on the presence of soil biota, the soil origin, and the plant species ratio. When planted in equal proportion, P. criopolitanum and C. thunbergii only coexisted in the presence of soil biota of P. criopolitanum, while C. thunbergii usually excluded P. criopolitanum under all other conditions. When the density of the neighboring plant increased, either species was excluded in the presence of their own soil biota. However, in the absence of a soil biota, there was a priority effect, giving an advantage to the species that was planted first. Our study revealed that the strength and direction of these interactions were altered by interspecific competition at the community level, and both the identity and the density of the competitors must be taken into account in interpreting plant community assembly processes.

植物-土壤反馈 (PSF) 和植物竞争都推动植物群落组装，但很少研究它们的相互作用，并且群落组成在调节这些相互作用中的作用尚不清楚。我们对两种植物（Polygonum criopolitanum和Carex thunbergii）进行了完全交互实验) 及其相关的土壤生物群落，以解开 PSF 和竞争的相对重要性。这两个物种在单一栽培或混合种植中种植，每种植物的邻居密度各不相同，并且存在或不存在与任一物种相关的土壤生物群。当单独生长时，每个植物物种在其自身土壤生物群的存在下表现出中性 PSF，而竞争中的反馈主要是负面的。P. criopolitanum在土壤生物群的存在下比在高压灭菌的土壤中产生更多的生物量，但在C. thunbergii 中观察到相反的情况. 在竞争中，两种植物物种都因其自身土壤生物群的存在而受到抑制，并且都没有竞争优势，导致在大多数情况下成对 PSF 显着为负。C. thunbergii也表现出比P. criopolitanum更强的人均效应。物种共存的预测很大程度上取决于土壤生物群的存在、土壤来源和植物物种比例。当等比例种植时，P. criopolitanum和C. thunbergii仅在P. criopolitanum土壤生物群存在的情况下共存，而C. thunbergii通常排除P. criopolitanum在所有其他条件下。当邻近植物的密度增加时，任何一个物种都被排除在它们自己的土壤生物群之外。然而，在没有土壤生物群的情况下，存在优先效应，使先种植的物种具有优势。我们的研究表明，这些相互作用的强度和方向被群落层面的种间竞争改变了，在解释植物群落组装过程时，必须同时考虑竞争者的身份和密度。