Complex responses to environmental temperature fluctuations in populations of Lemna minor

Background/Question/Methods

An understanding of the consequences of global climate change requires the ability to predict how environmental variability will alter populations. While early work assumed that environmental variability merely introduces “noise” to ecological systems, there are theoretical and empirical reasons to believe that populations may respond to variability in complex ways. The most manageable approach to addressing this issue empirically is to quantify the responses of populations to constant environmental conditions, and then use this information to make testable predictions about how populations respond to variable environments. To explore the utility of this approach, we parameterized a simple model of growth as a function of constant temperature conditions in laboratory populations of the aquatic plant Lemna minor. This model was then used to make predictions about how L. minor populations should respond to different regimes of constant and fluctuating temperatures. To account for the possibility of delayed responses to temperature fluctuations, we constructed a model of population growth dependent on a weighted average of the responses to past conditions. These two models were tested by collecting time-series and total growth data of L. minor populations under constant and various fluctuating temperature treatments.

Results/Conclusions

We fitted a constant-temperature growth curve for L. minor across a range of constant conditions. The result was concave up at low temperatures and concave down at high temperatures. Given Jensen’s inequality, we predicted that population growth rates would be depressed by fluctuations in low temperature ranges and elevated by fluctuations in high temperature ranges. We tested this prediction by measuring the growth of L. minor populations under different temperature regimes (constant, sine-wave, and pink-noise fluctuations) with mean temperatures of either 13^oC or 28^oC. Consistent with our predictions, we found a significant reduction in overall population growth rate under fluctuating versus constant conditions with mean temperatures of 28^oC (p=0.0042). However, fluctuations around 13^oC led to a reduction in growth relative to constant environments (p=0.0404), contradicting our simple prediction. We tested the possibility of a delayed effect of past conditions by collecting time-series data from populations grown under different temperature regimes, with shifts between 25^oC, 35^oC, and 15^oC. We found a clear signal of a lag response, but the direction of change was inconsistent with a simple lag-model. These results suggest that more complicated shock-response models are needed to characterize how populations respond to environmental variability.