Time Changes Everything: Analysing Temporal Patterns of Evaporative Water Loss
Abstract Higher air temperatures and drier conditions may create stronger water vapour pressure and increase rates of cutaneous water loss, while elevated body temperatures may in turn directly speed up metabolic rates that lead to higher respiratory water loss. Therefore, water budgets are an important organismal trait for understanding their responses to climate change. The most common method of water loss estimation combines respiratory and cutaneous pathways by measuring body weight loss over a defined period of time. Currently, obtained values are often summed or averaged for population or species comparisons. We warn about potential statistical problems using average or summed values of water loss due to emerging temporal patterns. In this study we used a model dataset of lizards and to investigate temporal patterns in water loss datasets. We found that temporal patterns strongly vary across datasets and often deviate from the summed/average profile. Also, the duration of the experiment needs to remain long enough to detect the temporal patterns and produce representative results, while averages at different end-points of the experiment will also vary with temporal patterns. We propose that a simple statistical approach including hour of the experiment as non-linear explanatory variable in GAMM is used to investigate and adequately account for temporal patterns, which will ensure comparability of studies using meta-analyses in the future. Found signal of temporal variation in water loss also suggests that it holds significant biological relevance, potentially mostly connected to behavioural but also physiological adjustments and needs research attention in the future.