Disturbance is a key factor in most natural environments and, globally, disturbance regimes are changing, driven by increased anthropogenic influences, including climate change. There is, however, still a lack of understanding about how disturbance interacts with species dispersal capacity to shape marine assemblage structure. We examined the impact of ice scour disturbance history (2009–2016) on the nearshore seafloor in a highly disturbed region of the Western Antarctic Peninsula by contrasting the response of two groups with different dispersal capacities: one consisting of high-dispersal species (mobile with pelagic larvae) and one of low-dispersal species (sessile with benthic larvae). Piecewise Structural Equation Models were constructed to test multi-factorial predictions of the underlying mechanisms, based on hypothesised responses to disturbance for the two groups. At least two or three disturbance factors, acting at different spatial scales, drove assemblage composition. A comparison between both high- and low-dispersal models demonstrated that these mechanisms are dispersal dependent. Disturbance should not be treated as a single metric, but should incorporate remote and direct disturbance events with consideration of taxa-dispersal and disturbance legacy. These modelling approaches can provide insights into how disturbance shapes assemblages in other disturbance regimes, such as fire-prone forests and trawl fisheries.

干扰是大多数自然环境中的一个关键因素，并且在全球范围内，干扰机制正在变化，这是由人为因素（包括气候变化）增加引起的。但是，对于扰动如何与物种扩散能力相互作用以塑造海洋组合结构仍然缺乏了解。我们通过对比两组具有不同分散能力的响应，研究了南极半岛西部高度受干扰地区的冰擦扰动历史（2009–2016）对近海海底的影响：一组由高分散物种组成（一组移动上层幼虫）和低分散性物种之一（与底栖幼虫无柄）。构建了分段结构方程模型以测试潜在机理的多因素预测，基于假设的两组反应。至少两个或三个扰动因子以不同的空间尺度起作用，推动了组件的组合。高分散模型和低分散模型之间的比较表明，这些机制是依赖于分散的。干扰不应被视为一个单一指标，而应考虑到分类群扩散和干扰传统，将远程和直接干扰事件结合在一起。这些建模方法可以提供洞察力，以了解在其他易受干扰的地区（如易火森林和拖网渔业）中的干扰形状如何组合。高分散模型和低分散模型之间的比较表明，这些机制是依赖于分散的。干扰不应被视为一个单一指标，而应考虑到分类群扩散和干扰传统，将远程和直接干扰事件结合在一起。这些建模方法可以提供洞见，以了解扰动形状如何在其他扰动范围内组合，例如易火森林和拖网渔业。高分散模型和低分散模型之间的比较表明，这些机制是依赖于分散的。干扰不应被视为一个单一指标，而应考虑到分类群扩散和干扰传统，将远程和直接干扰事件结合在一起。这些建模方法可以提供洞见，以了解扰动形状如何在其他扰动范围内组合，例如易火森林和拖网渔业。