Agricultural water demand may dynamically vary as socioeconomic characteristics of farmers evolve. Reservoir operation under dynamic demand has a quite new origin. This paper employs an Agent-Based Model (ABM) in combination with the standard reservoir operating policy (SOP). Proposed framework (ABM-SOP) simulates a dynamic long-term reservoir operation considering the variations of agricultural water demand as socioeconomic behaviors of farmers evolves. In the proposed ABM, farmers are assumed to have memory and individual characteristics that make exclusive decisions based on their knowledge about the prevailing conditions. Farmers' water demand varies dynamically with their main socioeconomic characteristics addressed by their age, education, risk-taking ability, and dependency on agricultural income. Farmers respond to amount of available water by changing the cropping pattern and irrigation technology without any obligation from policymakers. Results of this study indicate that first, through the dynamic interaction between farmers and reservoir operators, water stress experienced by farmers is reduced by about 50%; they gain 20% more profit by making adapted decisions under different circumstances. Second, the most critical factor influencing the farmers' responses is risk-taking rate that might be managed to increase flexibility of farmers in their decisions. Third, a bottom-up approach applied in this research shows that selfish farmers could be smart enough to make proper decisions in response to fluctuation in available water as a result of self-organizing ability (i.e. ability of individual agents to organize their collective behaviors through interacting with each other rather than external intervention or education) of complex human-natural systems. Farmers would balance their risk-taking status based on their education, and as they receive information on reservoir storage and water availability for allocation. So, policymakers can improve the efficiency of their decisions by promoting the information exchange with all stakeholders and parties involved in the system operation.