LAUSR

Fluctuations in food availability occur in many ecosystems and pose challenges to the organisms that inhabit them (Berg et al. 2017). Because lizards have significantly lower field metabolic rates than endotherms, they are often considered “low energy specialists” that can cope with food shortages (Brown and Griffin 2005). However, active reptiles that maintain high body temperatures during periods of low prey availability can still incur significant energetic costs because metabolic rates scale with body temperature. To reduce these energetic costs, some reptile species select lower body temperatures and/or reduce their activity levels when food is scarce (Angilletta 2009). While the adaptations of lizards to seasonal food shortages are reasonably well documented (Berg et al. 2017), the thermoregulatory responses of lizards to short-term food shortages are less well studied (Brown and Griffin 2005). According to an energetic model of temperature regulation, the body temperature under which net energy gain is maximized is higher for high food availability than for low food availability; therefore, lizards ought to select lower temperatures when food is scarce to maximize their net energy gain (Huey 1982). However, this model only considers energetic costs. That is, selection of lower body temperatures may involve costs associated with a reduction in performance of other fitness-enhancing activities, such as locomotion. In some species, individuals may maintain high body temperatures during fasting in order to carry out important fitnessrelated activities such as defending territories or searching for mates. For example, male spiny lizards Sceloporus jarrovi maintained high body temperatures during periods of fasting, presumably because the maintenance of higher body temperatures facilitated important functions such as territory defence and predator escape behaviors (Schuler et al. 2011). Thus, the degree to which ectotherms depress body temperatures during periods of low prey availability will depend on both the energetic and non-energetic costs and benefits of thermoregulation (Huey 1982; Angilletta 2009). At present, it is unclear whether the thermoregulatory strategies of many lizard species are influenced by food availability (Brown and Griffin 2005; Angilletta 2009). In this study, we investigated the effect of fasting on thermoregulation by females of the nocturnal gecko Amalosia lesueurii. One week after the females had laid eggs, we fasted them for 5 days, fitted them with miniature data loggers, and then measured their preferred body temperatures in a cost-free thermal gradient (10–50 C). We then fed the lizards with live crickets ad libatum, and recorded their body temperatures within the same thermal gradient for 24h (see Supplementary Material). We found that females maintained lower body temperatures after fasting (mean 1⁄4 23.6 C, standard deviation [SD] 1⁄4 0.30) than they did following food consumption (mean 1⁄4 26.3 C, SD 1⁄4 0.64 paired t-test, t31 1⁄4 20.30, P<0.001, Figure 1). While fasting, lizards also maintained a narrower set-point temperature range (22.4–25.2 C) than they did after feeding (24.0–28.7 C). Lizards also attained higher maximum temperatures after feeding (mean1⁄4 33.4 C) than during fasting (mean1⁄4 28.2 C; t311⁄4 28.0, P<0.001), but mean minimum temperatures of lizards were similar (17.1 C vs. 17.4 C, t31 1⁄4 0.607, P1⁄40.55). Repeated measures ANOVA revealed that body temperature was influenced by food consumption (F1,31 1⁄4 376.03, P<0.001), time of day (F1,311⁄4 258.71, P<0.001), and hour of the day (F23,713 1⁄4 4.51, P<0.001). Body temperatures of lizards were higher during the day than at night, irrespective of their feeding status (Figure 1). Mean diurnal and nocturnal body temperatures, and associated set-point temperatures, were 27.4 C (25.2–29.7 C) and 25.3 C (22.9–27.6 C) for fed lizards respectively, and 24.3 C (23.1–25.9 C) and 23.0 C (22.0– 24.4 C) for fasting lizards, respectively. Based on the thermal performance of sprint speed for adult females (Dayananda et al. 2017), fasted lizards maintained diurnal temperatures that would allow them to sprint at 91% of their maximal sprinting capacity (Figure 1). Our results demonstrate that in a cost-free thermal gradient, female velvet geckos selected higher body temperatures after feeding