Phenotypic plasticity in heterogeneous environments can provide tight environment-phenotype matching. However, the pre-requisite is a reliable environmental cue(s) that enables organisms to use current environmental information to induce the development of a phenotype with high fitness in a forthcoming environment. Here we quantify predictability in the timing of precipitation and temperature change to examine how this is associated with seasonal polyphenism in tropical Mycalesina butterflies. Seasonal precipitation in the tropics typically results in distinct selective environments, the wet- and dry seasons, and changes in temperature can be a major environmental cue. We sampled communities of Mycalesina butterflies from two seasonal and one aseasonal location. Quantifying environmental predictability using wavelet analysis and Colwell’s indices confirmed a strong periodicity of precipitation over a 12-month period at both seasonal locations compared to the aseasonal one. However, temperature seasonality and periodicity differed between the two seasonal locations. We further show that: (1) most females from both seasonal locations synchronise their reproduction with the seasons by breeding in the wet season but arresting reproduction in the dry season. In contrast, all species breed throughout the year in the aseasonal location, and (2) species from the seasonal locations, but not those from the aseasonal location, exhibited polyphenism in wing pattern traits (eyespot size). We conclude that seasonal precipitation and its predictability are primary factors shaping the evolution of polyphenism in Mycalesina butterflies, and populations or species secondarily evolve local adaptations for cue use that depend on the local variation in the environment.

The data represented in this manuscript were obtained both from the climate database (CRU database) and traits were quantified from the field-collected samples. From the climate database, we downloaded data from 2001-2019 on Mean Precipitation and Mean Temperature for three locations that were sampled in this study. This time series climate data was used for quantifying the predictability of the precipitation and temperature using wavelet analysis. From the field-collected samples, we measured the eyespot size of several species represented from all three locations and also dissected females to detect the reproductive activity of the female based on the presence/absence of spermatophores and eggs.