Sources of Variation in Laboratory Measurements of Water Relations of Reptilian Eggs and Embryos

Abstract

We performed an experiment on the water relations of eggs and embryos of common snapping turtles (Chelydra serpentina) to evaluate several sources for the random variation that is inherent in laboratory investigations using environmental chambers to provide a uniform environment or to apply specific treatments. Eggs were incubated on wet and dry substrata in an experiment that was replicated within and between environmental chambers set at the same apparent temperature (26°C). Clutch of origin was the identifiable factor accounting for most of the random variation in data for length of incubation and for mass and carapace length of hatchlings, while effects of box and environmental chamber were minor. Only in the case of data for net change in mass of eggs over the course of incubation did box and environmental chamber contribute importantly to random variation. Nonetheless, treatment effects overwhelmed the several sources for random variation, with eggs incubating on wet media (-150 kPa) absorbing more water than eggs on dry media (-950 kPa) and with embryos developing longer and attaining larger size before hatching in wet environments than in dry ones. Experiments focusing on length of incubation and on attributes of hatchlings will benefit more from considering clutch of origin than from replicating by box or environmental chamber. However, studies considering patterns of net water exchange may need to replicate fully by clutch, box, and environmental chamber to achieve the necessary sensitivity in statistical tests and to avoid critical biases resulting from pseudoreplication, especially in those instances in which treatment effects are relatively small.