Vegetation patterns in arid and semi-arid ecosystems as a self-organized response to resource scarcity is a well-documented issue. Their formation is often attributed to the symmetry-breaking type of instability. In this contribution, we focus on a regime far from any symmetry-breaking instability and consider a bistable regime involving uniformly vegetated covers and a bare state. We show that vegetation populations exhibit non-random two-phase structures where high biomass density regions are separated by sparsely covered areas or even bare soil. These structures are referred to as phase separation vegetation covers. We provide observations of this phenomenon in Gabon, Angola, Argentina, and Mexico. The inhomogeneities in environmental conditions are crucial to explain the origin of phase separation vegetation covers. We derive a simple equation from ecologically relevant models to explain various field observations. The bifurcation diagrams obtained from this model allow us to prove that inhomogeneity in the aridity parameter is a source of resilience for vegetation covers, avoiding collapsing towards a bare state. We characterize the natural observations and the equilibria from the model by using Fourier transform technique, spatial autocorrelation analysis, and size distribution of patches analysis.

干旱和半干旱生态系统中的植被模式作为对资源稀缺的自组织反应是一个有据可查的问题。它们的形成通常归因于对称性破坏类型的不稳定性。在这篇文章中，我们关注一个远离任何对称破坏不稳定性的机制，并考虑一个涉及均匀植被覆盖和裸露状态的双稳态机制。我们表明，植被种群表现出非随机的两相结构，其中高生物量密度区域被稀疏覆盖的区域甚至裸露的土壤隔开。这些结构被称为相分离植被覆盖。我们在加蓬、安哥拉、阿根廷和墨西哥提供对这种现象的观察。环境条件的不均匀性对于解释相分离植被覆盖的起源至关重要。我们从生态相关模型推导出一个简单的方程来解释各种实地观察。从该模型获得的分岔图使我们能够证明干旱参数的不均匀性是植被覆盖的弹性来源，避免了向裸露状态塌陷。我们通过使用傅里叶变换技术、空间自相关分析，以及补丁分析的大小分布。