We study a dynamic version of multi-agent path finding problem (called D-MAPF) where existing agents may leave and new agents may join the team at different times. We introduce a new method to solve D-MAPF based on conflict-resolution. The idea is, when a set of new agents joins the team and there are conflicts, instead of replanning for the whole team, to replan only for a minimal subset of agents whose plans conflict with each other. We utilize answer set programming as part of our method for planning, replanning and identifying minimal set of conflicts.