Argumentation-based frameworks are used as a decision-making mechanism for software agents. This paper aims to investigate how a formal argumentation framework is affected when the underlying causal relationships of its theory is modified in counterfactual situations, the so-called "what if" scenarios. In contrast to previous approaches where causality relationships were derived from static probabilistic distributions, we address scenarios where causal models are intervened. Two novel contributions in the synergy between argumentation and causal theories are presented: 1) we characterize interventions and their consequences in causal argumentation frameworks; and 2) we introduce an account of the so-called sequential interventions that give a characterization of manipulations on time.