Soil is a complex environment made of multiple microhabitats in which a wide variety of microorganisms co-exist and interact to form dynamic communities. While the abiotic factors that regulate the structure of these communities are now quite well documented, our knowledge of how bacteria interact with each other within these communities is still insufficient. Literature reveals so far contradictory results and is mainly focused on antagonistic interactions. To start filling this gap, we isolated 35 different bacterial isolates from grains of soil assuming that, at this scale, these bacteria would have been likely interacting in their natural habitat. We tested pairwise interactions between all isolates from each grain and scored positive and negative interactions. We compared the effects of simultaneous versus delayed co-inoculations, allowing or not to a strain to modify first its environment. One hundred fifty-seven interactions, either positive or negative, were recorded among the 525 possible one’s. Members of the Bacillus subtilis, Pseudomonas and Streptomyces genera were responsible for most inhibitions, while positive interactions occurred between isolates of the Bacillales order and only in delayed inoculation conditions. Antagonist isolates had broad spectral abilities to acquire nutrients from organic and inorganic matter, while inhibited isolates tended to have little potentials. Despite an overall domination of antagonistic interactions (87%), a third of the isolates were able to stimulate or rescue the growth of other isolates, suggesting that cooperation between bacteria may be underestimated.

土壤是由多种微生境组成的复杂环境，其中多种微生物共存并相互作用以形成动态群落。尽管调节这些群落结构的非生物因素已得到充分证明，但我们对细菌在这些群落中如何相互作用的知识仍然不足。文献揭示了迄今为止矛盾的结果，并且主要集中在对抗性相互作用上。为了填补这一空白，我们从土壤颗粒中分离出35种不同的细菌分离株，假设在这种规模下，这些细菌可能已经在其自然栖息地中相互作用。我们测试了每个谷物的所有分离株之间的成对相互作用，并对它们的正负相互作用进行了评分。我们比较了同时接种和延迟接种的效果，允许还是不允许首先修改其环境。在525个可能的互动中，有157个互动（正面或负面）被记录下来。成员枯草芽孢杆菌，假单胞菌属和链霉菌属负责大多数的禁忌，而积极的相互作用的菌株之间发生芽孢杆菌秩序，只有在延迟接种条件。拮抗菌株具有广泛的光谱能力，可从有机和无机物中获取营养，而被抑制的菌株则潜力不大。尽管总体上拮抗作用占主导地位（87％），但三分之一的分离物能够刺激或拯救其他分离物的生长，表明细菌之间的合作可能被低估了。