Patterns of reproduction in African bats can be compared in three laxon-based groups: fruit bats (Megachiroptera), fretailed bats (Microchiroptera: Molassidae) and the nonmolossid Microchiroptera. In the fruit bats and nonmolossid Microchiroptera there is a trend from either seasonal or aseasonal polyestry, with prolonged or continuous spermatogenesis in the tropics, towards seasonal monestry and seasonal spermatogenesis at more temperate latitudes. Reproductive delays (sperm storage, delayed implantation and delayed development) are rare at tropical latitudes, but are the norm in the nonmolassid Microchiroptera away from the tropics. The molassids are mostly polyestrous at tropical and temperate latitudes, although the duration of the reproductive season decreases with increasing latitude. The molassids appear to have escaped the constraints that affect reproduction of the other Microchiroptera. We propose that this may be due to their flight capabilities and foraging behavior, which give them access to year-round food, and to the thermal characteristics of their roosts. We suggest that the ancestral reproductive pattern of the Chiroptera was probably aseasonal or seasonal polyestry, as seen in extant tropical species, and therefore that reproductive cycles have evolved from the polyestrous to the monestrous condition. Short periods of reproductive delay occur in some species of tropical bats; we suggest that these reproductive delays originally were not adaptations to temperate latitudes but rather to the long dry season, which is characteristic of African tropical latitudes. With the move away from the tropics, selective pressures, acting on the timing of lactation and spermatogenesis, would have ensured that these processes continued to occur in the warm wet season, and that the length of the reproductive delay increased. This model accommodates the probable evolutionary origin of bats and links the evolution and development of reproductive delays to the differences in climate that occur with changes in latitude. There is evidence that mate choice and sperm competition may be important to modern bats, but we believe that they need not be invoked as causal factors in the evolution of reproductive delays, which can be adequately explained using purely energetic arguments.
