Biologists increasingly report anthropogenic disruptions of both organisms and ecosystems, suggesting that these processes are a fundamental, qualitative component of the Anthropocene. Nonetheless, the notion of disruption has not yet been theorized in biology. To progress in that regard, we work on a special case. Relatively minor temperature changes impact plant-pollinator synchrony, disrupting mutualistic interaction networks. Understanding this phenomenon requires a specific rationale since models describing them use both historical and systemic reasoning. Specifically, history justifies that the system is initially in a very narrow part of the possibility space where it is viable, and the disruption randomizes this configuration. Building on this rationale, we develop a formal framework inspired by Boltzmann's entropy. With empirical networks, we show historical trends depending on latitude. Then we propose an initial definition of disruption in ecology. When a specific historical outcome contributes to a system's viability, disruption randomizes this outcome, decreasing viability.