If we investigate the number and populations of species in an ecosystem, we can observe universal characteristic patterns, the species abundance patterns, such as species abundance distributions, species-area relationships, etc. How to clarify the mechanisms underlying those patterns has been one of the "unanswered questions in ecology in the last century (May, 1999)" even though the knowledge obtained from it would affect vast areas of ecology. Various "statistical descriptors" such as the exponential distribution, the log-normal distribution, the power-law, etc., have been applied to ecosystem communities, but the mechanisms to generate those patterns based on the realistic population dynamics have not been fully clarified. The neutral theory (Hubbell, 2001) is a candidate of the mechanism for an adherent community like tropical rain forests and coral reefs but it have left an essential question, why such a community appears neutral. It also have left the more complex systems a mystery. Such systems occur on multiple trophic levels and include various types of interspecies interactions, such as prey-predator relationships, mutualism, competition, etc., and their adaptation and evolution. In my talk, the latest physical theory to link the evolutionary population dynamics to the species abundance patterns is introduced.