Emergence of diverse epidermal patterns via the integration of the Turing pattern model with the majority voting model

Abstract

The Turing pattern model is one type of reaction-diffusion (RD) model. The first identification of pattern formation by the Turing pattern model in an actual animal was made in the 1990s with the observation of patterns in the sea anemone. But can we assume that all epidermal patterns in animals can be explained by the Turing pattern model? Even for fish, there are some fish that are clearly not Turing patterns, differing significantly from the patterns that can be generated by RD models. For example, the body pattern of the ornamental carp Nishiki goi produced in Japan varies randomly from individual to individual, and it is difficult to predict the pattern of the offspring from that of the parent fish. A model in which these fish patterns are formed randomly is the majority voting model. From this, it can be inferred that the epidermal pattern of fish can be explained by either the Turing pattern model or the majority voting model. But how do fish use these two different models? It is hard to imagine that completely different epidermal formation mechanisms are used among species of the same family. For this reason, there may be a more basic model that can produce patterns for either model. In this study, the Turing pattern model and the majority voting model were represented by cellular automata, and then a new model integrating these two models was proposed. By adjusting the parameters, this integrated model was able to create patterns that are equivalent to both the Turing pattern model and the majority voting model. By setting the intermediate parameters values of these two models, it was possible to create a variety of patterns that were more diverse than those created by each single model. Although this model is simpler than previously proposed models, it was able to confirm that it can create a variety of patterns.