Understanding how metacommunities respond to natural and anthropogenic disturbances is a key objective in ecology. In this study, we introduce a robust analytical framework to identify communities whose extirpation triggers stronger (hereafter keystone communities) or weaker (hereafter idle communities) cascading effects on extinction and colonization events that ultimately drive temporal changes in compositional patterns of the remaining communities. These cascading dynamics reflect the impact of extirpated communities on connectivity and subsequent dispersal dynamics. Since the framework uses spatial information on compositional similarities to infer changes that would unfold over time due to the extirpation of one or more communities, we describe it as a space-for-time approach. Through mechanistic simulation models that replicate removal experiments, we demonstrate that our framework accurately estimates ”keystoneness”, ranking local communities by their role in maintaining the metacommunity’s compositional patterns. As such, our models demonstrate that the relationship between patch characteristics and our keystoneness metric is closely linked to the structure and dynamics of their metacommunities. A key feature of our framework is its ability to generate community keystoneness estimates that are statistically independent of local diversity, providing a valuable tool for assessing the relevance and conservation value of local communities. This is particularly important in cases where high local diversity reflects an influx of individuals into demographic sinks, a common consequence of human activities near natural areas. To showcase the unique insights of this framework, we examined and contrasted the effects of artificial light at night on the diversity and keystoneness of a moth metacommunity sampled over two decades. We conclude with a discussion of the framework’s potential applications and underlying assumptions, emphasizing its relevance for addressing both conceptual and applied ecological questions, particularly its potential to assess the conservation value of local communities under ecological stress.