In recent years dedicated observations have uncovered star formation at extremely low rates in dwarf galaxies, tidal tails, ram-pressure stripped gas clouds, and the outskirts of galactic disks. At the same time, numerical simulations of galaxy evolution have advanced to higher spatial and mass resolutions, but have yet to account for the underfilling of the uppermost mass bins of stellar initial mass function (IMF) at low star-formation rates. In such situations, simulations may simply scale down the IMF, without realizing that this unrealistically results in fractions of massive stars, along with fractions of massive star feedback energy (e.g., radiation and SNII explosions). Not properly accounting for such parameters has consequences for the self-regulation of star formation, the energetics of galaxies, as well as for the evolution of chemical abundances. Here we present numerical simulations of dwarf galaxies with low star-formation rates allowing for two extreme cases of the IMF: a “filled” case with fractional massive stars vs. a truncated IMF, at which the IMF is built bottom-up until the gas reservoir allows the formation of a last single star at an uppermost mass. The aim of the study is to demonstrate the different effects on galaxy evolution with respect to self-regulation, feedback, and chemistry. The case of a stochastic sampled IMF is situated somewhere in between these extremes.