Most lizards maintain quite constant body temperatures by behavioural means. Seasonal variations of environmental factors, such as temperature, sunlight exposure and wind intensity, influence lizard thermoregulatory abilities. Understanding how seasonal environmental shifts influence lizards’ thermoregulation helps us to know how they deal behaviourally with environmental changes, in general. We examined seasonal shifts (spring vs. summer) in behavioural thermoregulation in Podarcis lilfordi from Binicodrell islet (Menorca, Spain). Operative temperatures varied between microhabitats and seasons, being lower in spring than in summer, regardless of sunlight exposure. Lizard body temperatures were also lower in spring than in summer. Lizards used sunny microhabitats more frequently in spring and shaded areas in summer. Habitat thermal quality was similar during both seasons, but lizards thermoregulated less accurately in spring than in summer. Thermoregulatory effectiveness was low in spring (0.28) and moderate in summer (0.76). In comparison with previously published results, our findings showed the marked seasonal variation in the effectiveness of thermoregulation amongst island populations, which should be considered in future comparative studies.

temperature, behavioural thermoregulation, seasonality, lacertids, islands, Balearic lizard, environmental changes

As ectotherms, lizards’ performance and fitness depend on their body temperatures (Angilletta 2009). Lizards often maintain their body temperatures within a relatively narrow range using behavioural thermoregulation, which involves various mechanisms, e.g. adjusting activity periods, changing body posture to maximise heat gain or loss or moving between hot and cool microhabitats (Carrascal et al. 1992; Adolph and Porter 1993; Bauwens et al. 1996; Sears et al. 2016). The extent of behavioural thermoregulation depends on its costs and benefits (Huey and Slatkin 1976; Sears and Angilletta 2015). Costs of thermoregulation mainly entail wasted energy and time, predation risk and losing opportunities for other activities. Meanwhile, benefits are clear: maximising physiological performance and, accordingly, individual fitness. The balance between these costs and benefits will determine the extent to which a lizard thermoregulates in a given habitat (Huey and Slatkin 1976; Sears and Angilletta 2015).

Energetic costs of thermoregulation increase as environmental temperatures depart from the thermal optima for the lizard’s physiological functions (Huey and Slatkin 1976; Sears and Angilletta 2015). Accordingly, spatio-temporal variation in biotic and abiotic environmental factors influences the degree of lizard thermoregulation. In temperate lizards, their microhabitat selection and thermoregulatory effectiveness is greatly influenced by the seasonal variation in environmental conditions (Van Damme et al. 1987; Díaz et al. 2006; Ortega and Pérez-Mellado 2016). For example, Mediterranean lacertids usually thermoregulate more effectively in the summer, when environmental temperatures are higher, than in the spring (Ortega and Martín-Vallejo 2018).

This is also the case in the Balearic lizard, Podarcis lilfordi. A previous study, examining the seasonal shift in thermal ecology of two populations, showed different seasonal effects on thermoregulation (Ortega et al. 2014). The magnitude of seasonal changes in thermoregulatory effort was higher in the population from the smaller islet. Probably its lower plant cover, smaller heterogeneity of microhabitats and smaller size (higher exposure to sea winds) offered fewer options for lizards to thermoregulate than on the larger islet. However, studying only two populations precluded the testing of this hypothesis.

To test the hypothesis that seasonal changes in thermal ecology depend on islet size and traits, studies of several insular populations of the same species are needed. Here, we examine the thermal ecology of a third P. lilfordi population, living on Binicodrell, an islet located south of Menorca (Spain). This population occupies the small islet (0.5 ha), which is located less than 30 km away from larger islets with previously studied populations (Fig. 1). The smaller the islet, the more it is subjected to sea winds (Nitis et al. 2005) and the lower habitat and, consequently, climatic heterogeneity (Algar and Mahler 2016). This would entail a negative relationship between island size and the seasonal variation of the climatic niche available for lizards (Algar and Mahler 2016; Pafilis et al. 2016). Thus, other factors being equal, within the same lizard species, we would expect a negative relationship between the area of the islet and the magnitude of seasonal thermoregulatory variation. We predicted that (i) lizards from Binicodrell thermoregulate accurately and effectively and (ii) the magnitude of seasonal variation in the accuracy and effectiveness of thermoregulation will be higher than in lizards from the previously studied populations.

Figure 1. 

Map of the three studied populations of the Balearic lizard: Podarcis lilfordi codrellensis from Binicodrell islet, assessed in the present study and P. lilfordi lilfordi from Aire islet and P. lilfordi brauni from Colom islet, studied in Ortega et al. (2014).

Males and females maintained similar T_b (Table 2). In addition, the sunlight situation and the wind speed had a significant effect on T_b, but not the substrate. The only studied factors affecting T_b were the season, with higher temperatures in summer and the T_a, with a positive relationship (Table 1, Table 2). Body temperatures varied seasonally when correcting by T_a variation (F_{1, 80} = 63.663, p < 0.0001). Linear regression slopes between T_b and environmental temperatures were similar for both seasons (T_b – T_a: F_{1, 75} = 0.010, p = 0.919; T_b – T_s: F_{1, 75} = 1.599, p = 0.210; Fig. 2).

Figure 2. 

Left: linear regressions between body temperature (T_b) and air temperature (T_a). Right: linear regressions between T_b and substrate temperature (T_s) of the studied population of Podarcis lilfordi codrellensis at Binicodrell island (Menorca, Spain). Slopes of both regressions are similar for spring and summer. 95% CI are depicted by grey lines.

Operative temperatures were lower in spring than in summer for all sunlight exposures (Table 3). Lizards’ microhabitat use varied amongst different sunlight exposures in each season. In spring, they used places under the full sun more frequently (G = 6.930; p = 0.008), whereas in summer, they used shaded areas more frequently (G = 3.865; p = 0.049). The use of microhabitats under filtered sun was similar for both seasons (G = 0.452; p = 0.501; Fig. 3).

Figure 3. 

Frequency of observations (%) of Podarcis lilfordi codrellensis lizards under different sunlight exposures for spring and summer. Frequency of use of sunny microhabitats was significantly higher in spring than in summer and the opposite took place for shaded microhabitats, while the use of filtered microhabitats was similar in both seasons.

In spring, 90% of T_bs were below the PTR and only 10% fell within the PTR. In summer, 28.3% of T_bs were within the PTR and 71.7% were above it. The index of accuracy of thermoregulation was significantly higher in the spring than in the summer (U = 0.00, p < 0.0001; Table 1). The index of habitat thermal quality was similar within seasons (U = 4427.00, p = 0.161; Table 1). Podarcis lilfordi codrellensis showed a poor thermoregulatory effectiveness in spring, significantly lower than in summer (U = 0.00, p < 0.0001; Table 1).

Seasonal shifts in the extent of behavioural thermoregulation of P. lilfordi codrellensis at Binicodrel islet were notably broad. In the nearby islets of Aire and Colom, body temperatures of P. lilfordi lilfordi and P. lilfordi brauni, respectively, show a much lower degree of seasonal variation. Body temperatures found in Binicodrell were lower in spring than those from Aire and Colom and were similar in summer (Ortega et al. 2014). Linear relationships between body and air temperatures varied seasonally in lizards of Aire and Colom islets (Ortega et al. 2014), whereas we found similar slopes for lizards of Binicodrell.

Microhabitat use also changed seasonally. In spring, lizards behaved similarly regarding the use of sun patches at Binicodrell when compared to the other two studied islets (Ortega and Pérez-Mellado 2017), preferring sunny areas, that offered temperatures closer to the optimum of sprint speed (≈ 35 °C; Bauwens et al. 1995). In summer, Binicodrell lizards used shaded places (≈ 30 °C) significantly more than in spring, while they used sunny patches (≈ 45 °C) less. The use of microhabitats under filtered sun was similar in spring and summer. By using microhabitats differently, lizards can achieve different thermoregulation accuracy and effectiveness and seasonal changes are a key factor in conditioning lacertid’s thermoregulatory effort (Ortega and Martín-Vallejo 2018).

Thermoregulatory accuracy of this population is different from those previously studied. In spring, it ranged from 1.9 °C in Aire to 1.1 °C in Colom (Ortega et al. 2014), whereas the studied lizards from Binicodrell showed a mean of 3.7 °C of deviations of body temperatures from the PTR. Additionally, during summer they deviated more from the PTR: 0.4 °C in Aire, 0.6 °C in Colom (Ortega et al. 2014) and 1.2 °C in Binicodrell. Thus, assuming the PTR and its seasonal acclimatisation are similar in all populations (Gvoždík 2012), the body temperatures of the studied lizards deviated by more than double from their thermal preferences vs. those of the nearby isles.

There are two ways for lizards to cope with seasonal environmental variation: (1) behavioural thermoregulation and (2) physiological thermal acclimatisation of preferred body temperatures (Little and Seebacher 2016) and a mix of both strategies would also be possible. Some lacertids change their preferred temperature range to adapt to seasonal thermal requirements, as is the case for Iberolacerta galani (Ortega et al. 2016c). Nevertheless, it seems not to be the case for P. lilfordi, whose preferred temperature range is similar during spring and summer. Hence, P. lilfordi seems to rely on behavioural strategies to adapt their thermoregulation to seasonal shifts.

Our present and past results (Ortega et al. 2014) illustrate the strong intraspecific variability in the thermoregulatory abilities of the Balearic lizard. In addition, the interaction between the seasonal changes in thermoregulation effectiveness and the population (Ortega et al. 2014) is reinforced with the results of P. lilfordi from Binocodrell. Lizards are predicted to increase thermoregulation accuracy with habitat heterogeneity (Sears et al. 2016). Heterogeneity of vegetation and substrates is greater in Colom islet (P. lilfordi brauni), intermediate in Aire islet (P. lilfordi lilfordi) and lower in Binicodrell (P. lilfordi codrellensis; see data in niche breadth and biotic capacity of the three islets in Garrido and Pérez-Mellado (2014)). Although thermal heterogeneity is probably closely related to the heterogeneity of substrates and vegetation, we do not have the empirical evidence that relates both variables. Nonetheless, lizards thermoregulate less effectively (and suffer from a greater seasonal shift in thermoregulation abilities) in the less heterogeneous habitat (Binicodrell). Previous studies in other Mediterranean lacertids found a lack of general thermoregulation patterns at small islets (Pafilis et al. 2016). In any case, applying new thermoregulation metrics can lead us to a better understanding of the influence of habitat thermal quality in thermoregulation effort of this and other lizard species and its flexibility between seasons and islets (Vickers and Schwarzkopf 2016).

Furthermore, thermoregulation differences found on the different populations of P. lilfordi might be connected with biological differences amongst the three subspecies. They differ in many biological traits (e.g. size, colour, diets and foraging behaviour) which could influence their thermal biology. Lizards from Aire (P. lilfordi lilfordi) are melanic and larger than the others, lizards from Colom (P. lilfordi brauni) are non-melanic (brownish-greenish colouration) and of intermediate size and lizards from Binicodrell (P. lilfordi codrellensis) are also non-melanic (brownish) and smaller than the others (Salvador 2014; Mencía et al. 2017). Their foraging behaviour and predation pressure are also highly variable (Pérez-Mellado and Corti 1993; Pérez-Cembranos et al. 2016; Mencía et al. 2017). Small behavioural differences, even in morphotypes of the same species sharing the same habitat, can lead to different body temperatures, as found in Podarcis melisellensis (Huyghe et al. 2007).

A reciprocal transplant experiment showed that Phrynosoma hernandesi lizards living at different elevations immediately adjusted to the use of light environment by means of phenotypic plasticity (Refsnider et al. 2018). Phenotypical plasticity can have an important role in the thermoregulation variability found in the Balearic lizard. This would provide lizards with the potential to behaviourally buffer the impacts of climate change, whenever cool microsites are available within their habitats (Refsnider et al. 2018). The fact that lizards adapt to seasonal changes in thermal conditions suggests, at least, some degree of phenotypic plasticity. We still do not know whether the high thermoregulation flexibility found in the Balearic lizard is due to habitat differences, local adaptation or phenotypic plasticity. It is probable that all three causes contribute to the observed flexibility in behavioural thermoregulation and disentangling their relative contributions will be necessary to understand how insular populations of lizards modulate their thermoregulation responses to environmental changes.

Lizards were captured under permit CEP 35/2013 of the Balearic Islands’ Government. We thank M. Garrido and A. Pérez-Cembranos for support during fieldwork. We thank Mary Trini Mencía and Joe McIntyre for linguistic revision. During fieldwork, funding was provided to Z. O. and A. M. by predoctoral grants from the University of Salamanca (FPI program) and partially supported by the research project CGL2012-39850-CO2-02 (Spanish Ministry of Science and Innovation). We also thank the financial support of Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brazil) and the Programa de Pós-graduação em Ecologia e Conservação (PPGEC) and Programa de Pós-graduação em Biologia Animal (PPGBA) of the UFMS. During analysis and writing, A. G. was funded by a CNPq Master’s Scholarship and Z. O. and A. M. were funded by PNDP/CAPES post-doctoral fellowships. All research was conducted in compliance with ethical standards and procedures of the University of Salamanca.