Adaptation to similar environments can lead to the evolution of similar phenotypes in phylogenetically independent lineages. However, the extent of parallel evolution often varies. Because such variations can be due to environmental heterogeneity among seemingly similar habitats, identification of the environmental factors that cause non-parallel patterns can provide valuable insight into the ecological factors associated with phenotypic diversification. Armor plate reduction in replicate freshwater populations of the threespine stickleback () represents a well-known example of parallel evolution. Many freshwater populations in multiple regions of the Northern Hemisphere have reduced plate numbers, but not all freshwater populations exhibit plate reduction. In this study, we characterized plate number variation in Japanese freshwater populations and investigated the association between plate number and several abiotic environmental factors. We found that most freshwater populations have not reduced plate numbers in Japan. Plate reduction tends to occur in habitats with warmer winter temperatures at lower latitudes in Japan. In contrast, low dissolved calcium levels or water turbidity had no significant effects on plate reduction, although these were reported to be associated with plate reduction in Europe. Although our data are consistent with the hypothesis that winter temperatures are associated with plate reduction, further studies on the relationship between temperatures and fitness using sticklebacks with varying plate numbers are necessary to confirm this hypothesis and understand the factors causing variations in the extent of parallel evolution.
