]1 The standard representation of an experiment conducted to understand the response of an organism to a climate change-associated environmental variable: means and standard error bars. A threshold response is presented here for clarity; but the same principle applies for other response types. The points represent replicate individual responses.
Biologists have all read papers or sat through seminars about the biological consequences of climate change and seen a figure like the one here (without the dots). A traditional variance-style test and standard error bars show that a climate change variable (e.g., pH; temperature) has a marked influence on the mean response of some plant or animal species (e.g., calcification; survivorship; growth; fecundity), often at a particularly vulnerable life history stage (e.g., fertilization; planktonic larval; calcifying adult). As the presenter grimly purveys the bad news, I often look at such figures and feel optimistic. A typical experiment requires at least ten replicates, in which case the standard error is roughly a third of the standard deviation. The standard deviation is about half the 95% prediction intervals, which give a sense for where most of the data lie. These ballpark estimates suggest that there is a good chance that the most extreme responses occur more than six times further from the mean than standard errors indicate. For the case illustrated here: despite being subjected to an unfamiliar environmental extreme (e.g., one predicted to occur by century’s end), three of the ten organisms cope well relative to the threshold, and two organisms actually do better than the mean response of the control. That is, even in the absence of a plausible ramping up of the environmental variable over multiple generations, some individuals are likely to respond reasonably well and potentially go on to successfully reproduce and contribute to local adaptation. These individuals are the current focus of climate change biology, because researchers have moved beyond simple manipulative experiments and are addressing questions about heritability (i.e., the degree to which plasticity and genetic variability are driving phenotypic responses). In which case, researchers also need to move beyond the mean and present all the data. When not adequately presented, it should be second nature for readers and viewers to visualise where the underlying data lie and ask questions about the extremes. At a time when science is constantly uncovering (and must purvey) bad news about the consequences of environmental change, it is especially important to draw attention to the new focus of climate change biology—adaptation—as well as to just how tough organisms can be.