Angilletta and colleagues [1] address the novel and exciting question of whether or not urban environments alter tolerance of extreme temperatures in ectotherms. They test the hypothesis that ants from urban environments tolerate heat better and cold worse than do ants from rural environments because urban areas are hotter than their rural surroundings. Here, we report comments raised during a journal club discussion of this paper, hopefully promoting further discussion, one of the goals of PLoS ONE.

First, environmental temperatures for ant microclimates were presented only for urban and not rural areas. A significant difference between rural and urban ant microclimates is the basis for the prediction tested, supplying its environmental context. Furthermore, temperature may not be the only difference between rural and urban environments. For example, higher pollution levels might affect heat tolerance in urban ants [2,3]. We consider more rigorous documentation of environmental temperature a pre-requisite for this type of study.

Second, where presented, ant microclimates are inappropriately estimated. While we do not disagree that trail surface temperature may approximate ant body temperature, i-Button datalogger temperature probably does not. Physical properties such as size and reflectance, both of which influence equilibrium temperature [4], differ greatly between ants and i-Buttons. Trail surface temperature would be better measured using a fine, shaded thermocouple. Moreover, since leaf-cutter ants are primarily nocturnal, as reported by Angilletta and colleagues, they were unlikely to have been active at the recorded daily trail temperature maxima of 45°C [see also 5]. Sharp daily fluctuations in temperature at ~11h and ~15h suggest that the i-Button was exposed to direct sunlight during this period, when ants were probably in nests or shade. We agree with the authors that visual observations of this species’ behaviour are required to determine thermal maxima actually experienced by individual ants.

Third, recent thermal histories probably differed between rural and urban ant groups. Without accounting for acclimatization, acclimation and/or hardening effects, one simply cannot conclude that city ants possess higher heat tolerance than rural ants. Indeed, previous work investigating upper critical thermal limits of ants has shown significant responses to temperature acclimation, inducing ~4.7°C variation [6]. While the authors acknowledge a detectable difference ‘in phenotypes’ of ants from São Paulo, animals in both groups should have been in the same physiological state at the start of temperature tolerance experiments. For example, heat tolerance generally declines with age within an insect species (e.g. [7,8]). The difference between urban and rural ants may simply reflect a difference in population age-structure. Although the authors acknowledge that they are unable to differentiate between phenotypic plasticity and/or adaptation, readers are left unable to ascertain whether or not the reported differences in heat tolerance are an effect of experimental conditions. We are of the opinion that the key question addressed in the study would be more appropriately addressed using either common-garden (e.g. [9]) or reciprocal transplant experiments (e.g. [10]). The latter are, however, difficult and time-consuming. Nonetheless, a test of the predictions made by the authors would be better evaluated using ants from both sites acclimated to a common temperature.

Factors such as hydration state and/or behavioural adjustments might also underlie the reported differences in thermal tolerance. For example, animals should be allowed to reach a full hydration state just before the thermal tolerance assays. Instead, Angilletta et al. [1] allowed ants to drink during the heat tolerance assay, and corrected for body size in an attempt to control for inter-individual differences in hydration. An ANCOVA with each individual’s body water content as a covariate would have been more appropriate, as body mass does not necessarily represent hydration state.

Fourth, differences between observers might have introduced unexplained variation into the tolerance assays. In our experience, practise reduces among-individual variation in observed thermal limit endpoints for insects. Observers should either reach consensus on scoring of responses during pre-trial experiments, or observer identity should be incorporated into statistical models as a fixed factor.

Finally, dispersal rates, patterns of movement and gene flow among populations all need to be accounted for before evolutionary adaptation can be inferred. Ants from the two sampling locations may have represented one super-colony, or geographically and genetically distinct populations. Without knowledge of these factors, one cannot infer this species’ potential for physiological change in the face of urbanisation. Although the study by Angilletta et al. [1] constitutes one of the first exciting glimpses into the effects of urbanization on environmental physiology of ectotherms, their conclusions would be better supported by a study designed to address the factors we raise above. We hope that our critique will aid the design and interpretation of future investigations of insect urban physiology.

Susana Clusella-Trullas, John S. Terblanche, Keafon Jumbam, Meagan Stevens and Sue Jackson

Centre for Invasion Biology
Department Botany and Zoology
Stellenbosch University
Private Bag X1
Matieland, 7602
South Africa

REFERENCES
1. Angilletta MJ, Wilson RS, Niehaus AC, Sears MW, Navas CA, Ribeiro PL (2007) Urban physiology: city ants possess high heat tolerance. PLoS ONE 2(2): e258. doi:10.1371/journal.pone.0000258.
2. Højer R, Bayley M, Damgaard CF, Holmstrup (2001) Stress synergy drought and a common environmental contaminant: studies with the collembolan Folsomia candida. Global Change Biology 7: 485-494.
3. Nadeau D, Corneau S, Plante I, Morrow G, Tanguay RM (2001) Evaluation for Hsp70 as a biomarker of effect of pollutants on the earthworm Lumbricus terrestris. Cell Stress Chaperones 6: 153-163.
4. Gates DM (1980) Biophysical Ecology. Springer, New York.
5. Cerdá X, Retana J, Cros S (1998) Critical thermal limits in Mediterranean ant species: trade-off between mortality risk and foraging performance. Functional Ecology 12: 45-55.
6. Kay CAR, Withford WG (1978) Critical thermal limits of desert honey ants: possible ecological implications. Physiological Zoology 51: 206-213.
7. Bowler K (1967) Changes in temperature tolerance with adult age in Tenebrio molitor. Entomologia Experimentalis et Applicata 10: 16-22.
8. Klok JC, Chown SL (2001) Critical thermal limits, temperature tolerance and water balance of a sub-Antartic kelp fly, Paractora dreuxi (Diptera: Helcomyzidae). Journal of Insect Physiology 47: 95-109.
9. Hoffmann AA, Shirriffs J, Scott M (2005) Relative importance of plastic vs genetic factors in adaptive differentiation: geographical variation for stress resistance in Drosophila melanogaster from eastern Australia. Functional Ecology 19: 222-227.
10. Ellers J, Boggs CL (2004) Functional ecological implications of intraspecific differences in wing melanization in Colias butterflies. Biological Journal of the Linnean Society 82: 79-87.