In any ecosystem, chaotic situations may arise from equilibrium state for different reasons.To overcome these chaotic situations sometimes the system itself exhibits some mechanisms of selfadaptability. In this paper, we explore an eco-epidemiological model consisting of three aquatic groups: phytoplankton, zooplankton and marine free viruses. We assume that the phytoplankton population are infected by external free viruses and zooplankton get affected on consumption of infected phytoplankton; also the infected phytoplankton do not compete for resources with the susceptible one. In addition, we model a mechanism by which zooplankton recognize and avoid infected phytoplankton, at least when susceptible phytoplankton are present. The zooplankton extinction chance increases on increasing the force of infection or decreasing the intensity of avoidance. Further, when the viral infection triggers chaotic dynamics, high zooplankton avoidance intensity can stabilize again the system. Interestingly, for high avoidance intensity, nutrient enrichment has a destabilizing effect on the system dynamics, which is in line with the paradox of enrichment. Global sensitivity analysis helps to identify the most significant parameters that reduce the infected phytoplankton in the system. Finally, we