THE EFFECTS OF MULTIPLE FACTORS ON VIABILITY SELECTION IN HYLA GRATIOSA TADPOLES

Two full sib families of Hyla gratiosa larvae were compared in growth rate and survival under twelve ecological conditions in field enclosures. The twelve conditions represented the independent absence or presence of two predators (nymphal dragonflies and larval salamanders) and a range of three initial tadpole densities (8, 16, 32 larvae per enclosure). This 2 × 2 × 3 design for variance analysis was replicated five times in a natural pond. The presence of either predator reduced survival levels by 24%. There was no consistent effect of tadpole density on survival. However, at low tadpole density (8 per enclosure), the presence of salamanders did not affect tadpole survival levels; effect of the salamanders was restricted to the higher densities (16 and 32 per enclosure). The combined effect on overall survival of the two predators was additive. One sibship (denoted A) consistently displayed a higher survival level than the other (denoted B). However, the level of differential survival, measured as the survival of B relative to A, varied among predator combinations. Survival differences among treatments and sibships were related to body size differences. The changing levels of differential survival between sibships did not reflect a changing level of differential body size but, rather, an ecologically mediated change in the relationship between the level of body size variation and the subsequent level of survival variation. Ecological factors such as conspecific density variation or predator abundance do not act as isolated selective pressures but, rather, interact in their effects on mortality rates. These interactions cause the value of a trait related to a fitness component to vary with ecological condition. These results suggest that a thorough understanding of how selection really does act in natural populations requires a thorough understanding of the relevant ecological factors, a point all too often unappreciated.