Geographical bias in physiological data limits predictions of global change impacts

Abstract

Climate affects all aspects of biology. Physiological traits play a key role in mediating these effects, because they define the fundamental niche of each organism. Climate change is likely to shift environmental conditions away from physiological optima. The consequences for species are significant: they must alter their physiology through plasticity or adaptation, move, or decline to extinction. The ability to understand and predict such organismal responses to global change is, however, only as good as the geographical coverage of the data on which these predictions are based. Geographical biases in the state of physiological knowledge have been identified, but it has not been determined if these geographical biases are likely to limit our capacity to predict the outcomes of global change. We show that current knowledge of physiological traits is representative of only a limited range of the climates in which terrestrial animals will be required to operate, because data for animals from only a limited range of global climates have been incorporated in existing compilations. We conclude that geographical bias in existing datasets limits our capacity to predict organismal responses in the vast areas of the planet that are unstudied, and that this geographical bias is a much greater source of uncertainty than the difference between the current climate and the projected future climate. Addressing this bias is urgent to understand where impacts will be most profound, and where the need for intervention is most pressing. A free Plain Language Summary can be found within the Supporting Information of this article.

The biogeography of warming tolerance in lizards

“…As recently demonstrated for Iberian lizards, for example, such variation can range up to 3°C (Herrando‐Pérez et al, 2019; Herrando‐Pérez et al, 2020), a factor among many others (e.g. sample size, geographical bias, plasticity) that can cause our predictions to either over‐ or underestimate true conditions (Duffy et al, 2021; Gunderson & Stillman, 2015; Seebacher et al, 2015; White et al, 2021). Thus, our results should not be generalised to all populations of a species, since thermal physiology and microclimatic conditions vary across geographical ranges.…”

Linking physiology to ecosystem function: how vulnerable are different functional groups to climate change?

“…We found that trends in functional group vulnerability within each broad-scale region mirrored trends in vulnerability at the global scale (increasing vulnerability with increasing trophic level). However, there was little variation between tertiary consumer and secondary consumer warming margins in the low-latitude region, potentially due to geographic sampling biases, where wet-tropical regions and developing nations tend to be under sampled compared to temperate regions (White et al, 2021) (Figure 1; Supplementary Figure 1). In the global analysis species warming margins increased with absolute latitude, suggesting that species that inhabit low-latitude regions are more vulnerable to climate change.…”

A comprehensive database of amphibian heat tolerance

“…While compiling a comprehensive sample of the literature is challenging, one can reach comprehensiveness by systematically reviewing both grey and traditional academic literature from numerous databases 29 . Second, the geographical coverage of most physiological databases is relatively poor 24 , 30 , 31 . One way to minimise this bias is to conduct the literature search in non-English languages because a non-negligible amount of non-English-language literature still exists in countries where English is not widely spoken 30 , 32 .…”