What determines the assembly and stability of complex communities is a central question in ecology. Past work has suggested that mutualistic interactions are inherently destabilizing. However, this conclusion relies on the assumption that benefits from mutualisms never stop increasing. Furthermore, almost all theoretical work focuses on the internal (asymptotic) stability of communities assembled all at once. Here, we present a model with saturating benefits from mutualisms and sequentially assembled communities. We show that such communities are internally stable for any level of diversity and any combination of species interaction types. External stability, or resistance to invasion, is thus an important but overlooked measure of stability. We demonstrate that the balance of different interaction types governs community dynamics. A higher fraction of mutualistic interactions can increase the external stability and diversity of communities as well as species persistence, if mutualistic interactions tend to provide unique benefits. Ecological selection increases the prevalence of mutualisms, and limits on biodiversity emerge from species interactions. Our results help resolve long-standing debates on the stability, saturation and diversity of communities.

决定复杂群落的聚集和稳定性的因素是生态学的核心问题。过去的工作表明，互惠互动本质上是不稳定的。然而，这个结论依赖于互惠互利的好处永远不会停止增加的假设。此外，几乎所有的理论工作都集中在一次聚集的社区的内部（渐近）稳定性上。在这里，我们提出了一个模型，该模型从互惠互利和顺序组装的社区中获得了饱和收益。我们表明，对于任何多样性水平和物种相互作用类型的任何组合，此类群落都是内部稳定的。因此，外部稳定性或对入侵的抵抗力是衡量稳定性的重要但被忽视的衡量标准。我们证明了不同交互类型的平衡控制着社区动态。如果互惠互动倾向于提供独特的好处，那么更高比例的互惠互动可以增加群落的外部稳定性和多样性以及物种的持久性。生态选择增加了共生的普遍性，物种相互作用对生物多样性的限制。我们的结果有助于解决关于社区稳定性、饱和度和多样性的长期争论。