This study addresses the distributed task allocation problem for heterogeneous multi-drone systems operating in dynamic environments. The paper introduces an algorithm, Consensus Based on Mental Ability (CBMA), which integrates environmental factors and mental ability functions to address challenges in dynamic distributed task allocation. Additionally, the study explores three dynamic task allocation strategies to effectively adapt to the evolving demands of new tasks. These strategies are designed to improve robustness of task allocation, thereby optimizing the overall mission performance. Finally, the advantages of the CBMA algorithm and the applicability of the three allocation strategies to dynamic task scenarios are validated through case studies involving multi-drone mission executions.