1. The widespread observation that heat tolerance is less variable than cold tolerance (‘cold-tolerance asymmetry’) leads to the prediction that species exposed to temperatures near their thermal maxima should have reduced evolutionary potential for adapting to climate warming. However, the prediction is largely supported by species-level global studies based on single estimates of both physiological metrics per taxon.
2. We ask if cold-tolerance asymmetry holds for Iberian lizards after accounting for intraspecific variation in critical thermal maxima (CT_max) and minima (CT_min). To do so, we quantified CT_max and CT_min for 58 populations of 15 Iberian lizard species (299 individuals). Then, we randomly selected one population from each study species (population sample = 15 CT_max and CT_min values), tested for variance homoscedasticity across species, and repeated the test for thousands of population samples as if we had undertaken the same study thousands of times, each time sampling one different population per species.
3. The ratio of variances in CT_max to CT_min across species varied up to 16-fold depending on the populations chosen. Variance ratios show how much CT_max departs from the cross-species mean compared to CT_min, with a unitary ratio indicating equal variance of both thermal limits. Sampling one population per species was six times more likely to result in the observation of greater CT_max variance (‘heat-tolerance asymmetry’) than cold-tolerance asymmetry. The null hypothesis of equal variance was twice as likely for cases of cold-tolerance asymmetry than for the opposite scenario.
4. Range-wide, population-level studies that quantify heat and cold tolerance of individual species are urgently needed to ascertain the global prevalence of cold-tolerance asymmetry. While broad latitudinal clines of cold tolerance have been strongly supported, heat tolerance might respond to smaller-scale climatic and habitat factors hence go unnoticed in global studies. Studies investigating physiological responses to climate change should incorporate the extent to which thermal traits are characteristic of individuals, populations and/or species.

Author contributions: Monasterio, Beukema and Gómes lead field (lizard sampling) and lab (estimation of thermal limits and measurement of body weights) work, and Monasterio and Araújo designed experiments. Herrando-Pérez conceived the idea of the two research manuscripts (Functional Ecology, Journal of Animal Ecology) and submitted the data to Dryad.

Funding: MBA partly funded through CGL2011-26852 project of the Spanish Ministry of Economy and Competitiveness. Field and experimental work funded by IC&DT 1/SAESCTN/ALENT-07-0224-FEDER-001755 project led by MBA.

Acknowledgements: We thank the Biological Station of “El Ventorrillo” for hosting the field team and for providing the thermal experimental facilities essential for this research. We also thank Tim Leerschool, Filipe Serrano and Matthijs Hollanders for their support in the field.

Collection permits: Samples, experiments and use of experimental animals supported for Portuguese populations by permits 360 to 362/2014/CAPT and 550 to 552/2014/CAPT (Instituto da Conservação da Natureza e das Florestas), and for Spanish populations by permits (autonomous communities in parenthesis hereafter) SGYB/EF/FJRH Re-9H/13 & SGYB/AF/DBP Re-79y131/14 (Andalucía), 2014-500201724/2014/02343 & INAGA/5000201/24/2013/04434 (Aragón), DGMEN/SEN/avp_13_025_aut & DGMEN/SEN/avp_14_020_aut (Castilla La Mancha), P/CYL/101/2013 & EP/CYL/106/2014 (Castilla y León), CN0023/14/ACA1587(14) (Extremadura), 2566/RX131316//clave031/2013 & 2241/RX123724//clave018/2014 (Galicia), and 10/033298.9/13 & 10/013907.9/14 (Madrid).