Deceptive Pollination Under Individual Modeling - Based on Reinforcement Learning

Pollination is a fundamental ecological process in ecosystems that provides more efficient reproduction for most flowering plants, and the lack of pollination would lead to a significant reduction in the number of plants in an ecosystem. In a complex ecological network, the activities of individuals in an insect population are often accompanied by the transfer and interaction of information, and pollination behavior is also carried out with the activities of individual insects. Deceptive pollination is undoubtedly one of the most specific of the many pollination methods, and generally occurs when deceptive flowering plants that do not provide substantial resources attract insect visitors by mimicking sensory cues. In response to the reduced biological costs associated with this deception, insects often begin their own learning to avoid visiting deceptive flowering plants. The aim of this study is to establish a pollinator-deceptive plant network model, to study and analyze the effects of interactions between individuals in a population on their behavioral decisions and collective evolution, and to build a model using ABM and reinforcement learning methods to reveal the population's evolutionary trends in response to individual behaviors. The results show that pollinators and deceptive plants can reach a reasonable sustainable level to maintain the balance of the ecosystem.