Impact of temperature on bite force and bite endurance in the leopard iguana (Diplolaemus leopardinus) in the Andes Mountains

How rising temperatures affect organisms is of primary concern to enable predictions of the impact of the anthropogenic climate change on biota (Dawson et al., 2011). The Intergovernmental Panel on Climate Change (IPCC) predicts an increase of 1.5–2.0°C in mean global temperature over the next 20 years, as well as an increase in the frequency of heat waves, droughts and the frequency of extreme temperature events (IPCC, 2018). Under this scenario, changes in phenology, abundance, distribution and extinction risk of natural populations are expected (Sinervo et al., 2010, 2018; Bestion et al., 2015; Llewelyn et al., 2018).

Ectotherms are especially vulnerable to climate change (Deutsch et al., 2008) because their performance depends on body temperature (T_b), which is largely dependent on environmental temperatures (Halliday and Blouin-Demers, 2017). Body temperature influences physiological performance including locomotion, bite force, reproduction, growth, courtship and activity patterns (Adolph and Porter, 1993; Anderson et al., 2008; Angilletta, 2009; Bestion et al., 2015; Kubisch et al., 2012). Typically, these traits are optimally efficient within a narrow range of T_b (Angilletta et al., 2002). Thus, the ability to maintain a relatively constant T_b, despite the thermally heterogeneous environment, has direct effects on survivorship and, therefore, fitness (Black et al., 2019).

Another strategy to cope with thermal heterogeneity includes the plasticity and adaptability of different traits (thermoregulatory behaviour, thermal sensitivity) to the changing abiotic circumstances (Angilletta, 2009; Black et al., 2019; Pigliucci, 2001). In particular, the potential of a genome to produce a range of phenotypes in response to distinct environmental conditions is known as phenotypic plasticity (Pigliucci, 2001). Thus, acclimation, consisting of reversible trait changes in response to environmental changes typically on a temporal scale of weeks or months, can be considered as a form of phenotypic plasticity (Schulte et al., 2011). The organism's ability to adjust thermally sensitive traits in a labile way, to the climate, allows predictions of population responses to global change (Llewelyn et al., 2018). However, this ability could vary across populations and species: some animals could maximize their performance at the temperature to which they are acclimated (‘beneficial acclimation’ hypothesis; Leroi et al., 1994; Wilson and Franklin, 2002), whereas other animals could perform better when they are acclimated to high temperatures than when they are acclimated to low temperatures (‘hotter is better’ hypothesis; Huey and Berrigan, 1996).

One of the performance traits that could be crucial for individual fitness is bite force (Anderson et al., 2008). Biting is used in prey selection (Erickson et al., 2004; Herrel and O'Reilly, 2006), prey handling (i.e. to capture, subdue and consume prey; Verwaijen et al., 2002), defense against predators (Anderson et al., 2008) and aggressive interactions (Huyghe et al., 2005; Lailvaux et al., 2004), particularly as a key weapon in agonistic fights among males to determine dominance (Husak et al., 2006). Males also bite females during copulation, seemingly to restrain them (Herrel et al., 1996; Husak et al., 2006), which may prolong copulation and increase success. Bite force is strongly influenced by body size and proportions such as head shape and size (Herrel et al., 2001a). Compared with other performance traits such as locomotion, bite force is less dependent on temperature because muscular generation of force is relatively insensitive to thermal change (Bennett, 1985). Previous studies in Trapelus pallida (Herrel et al., 2007) and Hemidactylus frenatus (Cameron et al., 2018) showed that 80% of maximal performance of bite force could be sustained across a broad range of temperatures, 22.5 to 37.5°C for T. pallida and 20 to 30°C for H. frenatus, suggesting that bite force is largely independent from temperature in certain ranges. However, studies in several species of lizard, such as Stellagama stellio, Uromastix aegyptia and U. acanthinura (Herrel et al., 1999), Gecko gecko (Anderson et al., 2008) and Bradypodion pumilum (Segall et al., 2013), have shown that T_b affects bite force. Particularly, in B. pumilum, a steep drop in bite performance at higher T_b was observed. Thermal acclimation of bite force has been scarcely studied in ectotherms, thus the present study explores a novel field within the thermal biology of lizards.

Diplolaemus leopardinus, known as the leopard iguana, is endemic to high elevations in the Andes Mountains and lower elevations in the Patagonian steppe environments of Mendoza Province, Argentina. These lizards belong to the Leiosauridae family (Frost et al., 2001) and, like other leiosaurids, they are oviparous, insectivorous and occasionally saurophagus (Cei, 1986; Donoso-Barros, 1966; Van Devender, 1977). Indeed, as in other allied genera, Diplolaemus species are also known as ‘gruñidores’ (grumblers), because both females and males produce a hissing sound during aggressive displays and they often vocalize and bite when disturbed (Laspiur et al., 2007; Donoso-Barros, 1966). The leopard iguana is a saxicolous (rock-dwelling), territorial lizard that is a sit-and-wait forager and uses self-made burrows typically under basking rocks. They are medium-sized lizards [mean snout–vent length (SVL): 104.5 mm], without evident sexual dimorphism or dichromatism, and are characterized by having a triangular head with strong wide jaws and marked development of the mandibular muscles (Ibargüengoytía et al., 2004; Victoriano et al., 2010; Fig. 1).

Given the ecological relevance of bite performance in D. leopardinus (diet, territorial and defensive behavior), we tested whether thermal variables influence bite force and bite duration, taking into account the expected direct effect of head morphology (Anderson et al., 2008; Herrel et al., 1999, 2001a; Verwaijen et al., 2002). Previous studies reported that bite force is influenced positively by head length and width, because a larger head may support both more muscle to generate bite force and longer bones for greater leverage (Herrel et al., 2001a), whereas bite endurance is influenced by lower jaw length (Gomes et al., 2020 preprint). We evaluated the thermal sensitivity of bite force and bite endurance before and after two thermal acclimation treatments under the hypothesis that variation in temperature produces differences in bite performance. For the lizards analyzed before acclimation treatments, we predicted that maximum bite performance would be achieved at a T_b close to the species' preferred temperatures, as shown for other physiological traits (i.e. locomotor performance; Angilletta et al., 2002), following the predictions of the ‘thermal coadaptation’ hypothesis (Angilletta et al., 2006; Huey and Bennett, 1987). For the lizards analyzed after acclimation treatments, we predicted (i) a plastic response in the thermal sensitivity of bite force, because leopard iguanas inhabit environments with great seasonal thermal variation, where maintenance of plasticity, despite its costs, would have a positive impact on fitness (Angilletta, 2009; Huey and Kingsolver, 1989; Pigliucci, 2001); and that (ii) those lizards acclimated to high temperatures should have better performance across all temperatures than those acclimated to low temperatures, following the ‘hotter is better’ hypothesis (Huey and Kingsolver, 1989; Huey and Berrigan, 1996).

The study area was located in the Angostura mountain valley (32.983°S, 69.333°W, WGS84, 2405 masl), in west-central Mendoza Province, Argentina. This area is surrounded by mountain peaks more than 3000–4000 masl, and it is characterized by a cold arid climate with a mean annual temperature of 6.3°C and mean annual precipitation of 359 mm (Sileo et al., 2015). The study area registers maximum temperatures in summer of 29°C and minimum temperatures in winter of −15.5°C. In this type of environment, lizards remain inactive from autumn until early spring.

Sampling was conducted 17–19 December 2018, in late spring. We captured by hand or noose 21 leopard iguanas (10 males and 11 females) during daylight, 09:00 to 18:00 h. We georeferenced each capture location using a handheld GPS (Garmin ^® eTrex 20, Garmin Ltd, Olathe, KA, USA) with datum WGS84 to enable the return of each individual to their capture site after completion of the experiments. The T_b of active lizards (N=15, 8 males and 7 females) was recorded within 15 s of capture by inserting a TP-K01 thermocouple ∼5 mm into the cloaca. Micro-environmental temperatures were also recorded at the capture site in order to identify the main heat source used. Substrate temperatures (T_s) were measured by placing the thermocouple (TP-K01) onto the substrate surface, whereas air temperature (T_a) was measured 2 cm above the ground. Thermocouples were connected to a TES 1312A digital thermometer (TES ^® Electrical Electronic Corp., Taipei, Taiwan, range: −50 to 1300°C, resolution: 0.1/1°C).

We measured operative temperatures (T_e, sensu Bakken, 1992) in the field using physical models (Bakken and Angilletta, 2014; Hertz et al., 1993) consisting of a hollow pipe made of three layers of grey propylene with both ends sealed with cork. We placed a temperature logger (Thermochron i-Button DS1921G) in each pipe. Models were calibrated by comparing temperatures of different sizes and colours simultaneously with the cloacal T_b of an adult of D. leopardinus exposed to sunshine and shade prior to deployment. The best physical model type was selected according to the best-fit correlation between the model and the T_b of D. leopardinus and corresponded to a dull grey PVC pipe, 120 mm length×27 mm diameter (Spearman rank correlation, T_b versus model_HEATING, R=0.99, t=13,300.31, N=181, P<0.001, CI=0.99–1.00; T_b versus model_COOLING, R=0.99, t=13,596.7, N=183, P<0.001, CI=0.99–1.00). We recorded T_e during December of 2018. We deployed 20 physical models in each representative microsite used by the species (exposed rocks, bare soil, beneath shrubs, under rocks) to evaluate the thermal quality of the habitat (Hertz et al., 1993). We programmed the temperature loggers to record every 15 min. We used only the temperature data recorded during the animals' daily period of activity (between 09:00 and 20:00 h) for the analyses.

We brought lizards to the laboratory in individual cloth bags. Lizards were housed in the bioterium located at the Instituto Argentino de Investigaciones de las Zonas Áridas (IADIZA, CONICET). Lizards were maintained in standard conditions in an environmental chamber with controlled temperature (24.0±1°C) and photoperiod (12 h:12 h, light:dark). This temperature was a midpoint between the temperatures of both acclimation treatments. Leopard iguanas were individually located in terraria (32×45×27 cm), provided with substratum and rocks collected at the capture site, and had free access to a shelter and basking sites. Water was supplied ad libitum, and food (crickets, mealworms and Blaptica sp. cockroaches) was provided every other day ad libitum. Lizards received UV light (Sylvania ^® Reptistar T8 lamps, Osram Sylvania, Wilmington, MA, USA) 8 h per day. All procedures were performed with the permission of the provincial authority (res. nos. 81/2018 and 96/2018, Dirección de Recursos Naturales Renovables, Mendoza Government) and complied with Argentinian National Law no. 14346 on animal care. The protocols used were approved by the Institutional Committee for the Care and Use of Laboratory Animals (CICUAL) of the Universidad Nacional de Cuyo (res. nos. 143/2018).

We used 21 lizards (10 males and 11 females) in trials to measure preferred body temperature (T_p; sensu Pough and Gans, 1982; Hertz et al., 1993) in a terrarium constructed of fiber-board (60×120×30 cm) partitioned into three lanes and partly covered by a ceiling on one end. As previous research suggested that both heliothermic and thigmothermic behaviours have an important role in the thermoregulation of D. leopardinus (Laspiur, 2016), we constructed a thermal gradient composed of two heat sources. A 75 W halogen lamp (luminescence equivalent to a 100 W incandescent lamp) was suspended from the ceiling at one end of each lane to generate a heliothermic temperature gradient. The gradient was enhanced by lining one-third of the ceiling and wall surfaces proximate to the lamp with aluminum foil. In addition, to generate the thigmothermic temperature gradient, four thermostat-controlled electric heat coils were connected to four mica sheets (each 30×60 cm), laid on the floor of the terrarium across all lanes, and covered with 5 cm of sand. The first sheet located at the cold end of the terrarium had a 20°C threshold, the second sheet 30°C, the third 40°C and the fourth 50°C. The two heat sources combined to generate a thermal gradient ranging from 23°C to 50°C. We affixed an ultrafine (44 ga) Type T thermocouple to the belly of each lizard using insulation material and surgical tape. The thermocouples were connected to an eight-channel datalogger (Measurement Computing 1.2 kHz Data Acquisition Device, OMEGATC-08±0.5°C, Stamford, CT, USA). After 30 min, we recorded T_b at 1 s intervals for 140 min.

After completion of the measurement of thermal preference, we recorded the sex of each individual, and measured their body mass (M_b, 100-g Pesola ^® micro-line spring scale, ±1 g), SVL and tail length (digital calliper, 0.01 mm accuracy). Males were identified by an enlargement at the base of the tail caused by the hemipenes, and if the enlargement was not prominent (i.e. subadults), the hemipenes were everted to confirm sex.

After 1 week of care in the laboratory, 16 individuals were each randomly assigned to one of two acclimation chambers for a 2-week trial. Previous studies have indicated that a 2-week acclimation period is sufficient to elicit phenotypic changes in herpetofauna (Bacigalupe et al., 2018; Barria and Bacigalupe, 2017; Wilhoft and Anderson, 1960). Eight individuals (3 males and 5 females) were housed in the 20°C-acclimation trial (±2°C), which represented the mean operative temperature registered in the field during December 2018, and eight individuals (4 males and 4 females) were housed in the 30°C-acclimation trial (±2°C), which equals the mean preferred temperature of the species, according to our present study. During each trial, we recorded the daily minimum and maximum temperature using a digital thermometer.

Bite trials were performed on 20 individuals of D. leopardinus (9 males and 11 females). We measured bite endurance and bite force with a piezo-electric isometric force transducer (Kistler^® Type 9203±500N, Kistler Group, Winterthur, Switzerland) fitted with stainless steel plates, connected to an A/D converter that was connected to a computer. The device measures bite force three times per second for 100 s. To avoid damaging the lizard's teeth, the end of each steel plate was covered with leather.

We measured bite endurance and bite force before thermal acclimation treatments (within the first 5 days after capture) and bite force was measured also after acclimation (2 weeks at 20°C or 30°C). For bite endurance and force trials individuals were induced to bite the apparatus three times at each of five cloacal T_b (20, 24, 28, 32 and 34°C) with 20 min rest after each bite and 1 day after each temperature trial. The order of the trials was randomized, and lizards were placed in an environmental chamber at the temperature trial for at least 1 h prior to bite measurements, and cloacal T_b was measured before each bite. We analyzed only front bite position.

To estimate endurance, we used the longest-lasting bite in any one of the three tests at each temperature. We measured endurance as the total time lizards sustained bite force greater than zero (Fig. 2), carefully observing during the trials when lizards were biting the device. For maximum bite force at each temperature, we used the greatest magnitude in any one of the three tests.

SVL and M_b were recorded to determine body condition index (BCI). This index was calculated following the methodology proposed by Peig and Green (2009), which consists of calculating the scale mass index (used as an index of body condition) as BCI=M_i (SVL₀/SVL_i)^bSMA, where M_i and SVL_i are the M_b and SVL of the individual, SVL₀ is the arithmetic mean SVL of the population, and b_SMA is the standardized major axis slope from the regression of ln(mass) on ln(SVL) for the population (Peig and Green, 2009, 2010). The b_SMA exponent was calculated using the package lmodel2 (https://CRAN.R-project.org/package=lmodel2) in R (https://www.r-project.org/). To evaluate the influence of morphology in bite force and endurance before acclimation treatments, we measured head length (from the tip of the snout to the posterior edge of the parietal bone), head width (the widest part of the head) and jaw length (distance between the jaw articulation to the tip of the lower jaw). All linear measurements were taken with a digital caliper (0.01 mm accuracy).

The dependence between T_b and micro-environmental temperatures (T_s and T_a) was analyzed using a repeated-measures ANOVA, whereas linear regression was used for the relationship of T_b on T_s and T_a. In order to determine the efficiency of thermoregulation and the preferred body temperature, we used the methodology proposed by Hertz et al. (1993). We calculated the mean preferred body temperature (T_pmean) and the mean T_set (the interquartile range of T_p) from the T_p obtained from each lizard (N=21). The accuracy of T_b was calculated as the mean of the absolute values of the deviations of T_b from the upper or lower bound of T_set (individual deviation, d_b). Operative temperature, T_e, was obtained from the mean of the temperatures registered by all the models (N=20) during the capture period. To calculate the thermal quality of habitat (d_e) we obtained the mean of the absolute value of the deviations of T_e from the upper or lower bound of T_set, large values indicate lower thermal quality (Hertz et al., 1993). Finally, we calculated the index of effectiveness of thermoregulation (E) proposed by Hertz et al. (1993), defined as E=1−(mean d_b/mean d_e), which approaches zero when animals do not thermoregulate, approaches one when lizards behave as efficient thermoregulators, and negative values of E occur when lizards avoid thermally high-quality habitats with T_e near or within the range of T_pmean (Christian and Weavers, 1996). We compared differences between sex in T_b, T_p, morphometric variables and bite performance using a t-test comparison for parametric data and a Mann–Whitney rank sum test when the assumptions of normality or homogeneity of variance were not fulfilled. We evaluated correlation between morphometric variables using the cor.test function in R. We regressed bite endurance and force on head size (head length, head width and jaw length) using the lm function of the package stats v3.6.2 in R.

The influence of T_b on bite endurance and bite force was analyzed within the before-acclimation and the after-acclimation treatments using a generalized mixed linear model (GLMM), with bite endurance or force as the response variable, T_b (explanatory variable), morphometries and body mass (covariates) as fixed effects, and individual identity as a random variable. We first tested whether individual identity significantly influenced performance across the entire T_b gradient using a χ² test. A Gaussian or gamma probability distribution was used depending on the nature of the data. For bite force in the before-acclimation treatment, we used the glmmadmb function of the glmmADMB package, with a gamma distribution, because data were not parametric (http://glmmadmb.r-forge.r-project.org/; Fournier et al., 2012). We also analyzed the effect of temperature on bite force using a Friedman repeated-measures ANOVA on ranks, and Tukey test as a posteriori test. For bite endurance and for force in the after-acclimation treatment, we used the lme function of the nlme package, with a Gaussian distribution (Bates et al., 2015). Model selection was made using a likelihood ratio test to compare models with different fixed-effect structures throughout the ANOVA function (Zuur et al., 2009). All statistical analyses were performed in R.

Activity cloacal T_b and T_p did not show significant associations with M_b (linear regression: T_b, F_1,15=4.22, P=0.06; T_p, F_1,21=0.01, P=0.92), SVL (linear regression: T_b, F_{1, 15}=3.37, P=0.09; T_p, F_1,21=0.44, P=0.51) or BCI (linear regression: T_b, F_1,15=0.61, P=0.45; T_p, F_1,21=2.92, P=0.1). Moreover, sexual differences in either T_b or T_p were not found (t-test: T_b, t_2,15=−0.28, P=0.78; T_p, t_2,21=−1.07, P=0.30).

Diplolaemus leopardinus showed a mean T_b of 26.69±1.18°C (Table 1, Fig. 3), which was higher than the T_a and did not differ from T_s (repeated-measures ANOVA: F_2,15=49.93, P<0.001, Tukey test: T_b×T_a, q=12.92, P<0.001; T_b×T_s, q=1.49, P=0.55). Body temperature depended significantly on both T_s and T_a (linear regression: T_s, F_1,15=22.38, P<0.001, R²=0.65; T_a, F_1,15=11.39, P<0.005, R²=0.5).

The T_pmean was 30.85±0.26°C, whereas the mean lower T_set was 29.80±1.27°C and the mean upper T_set was 31.88±1.40°C (Table 1, Fig. 3). According to the distribution of temperatures recorded in the field, 43% of the records of T_b were included in the T_set (Fig. 3), and 57% were below the T_set. Mean T_e was 19.42±0.49°C, with minimum temperatures of 9°C and maximum temperatures of 44°C during the activity time. The mean d_b was 4.26±0.83 and the mean d_e was 10.31±0.28. E was 0.57.

Males and females did not differ in SVL (t-test: t_2,20=0.49, P=0.63), M_b (t-test: t_2,20=0.82, P=0.42), BCI (t-test: t_2,20=−0.32, P=0.75), head width (Mann–Whitney U-test: U_2,20=45, P=0.76), head length (Mann–Whitney U-test: U_2,20=30, P=0.15), jaw length (t-test: t_2,20=0.52, P=0.61), maximum bite force (t-test: t_2,98=−0.6, P=0.56) or bite endurance (t-test: t_2,20=1.02, P=0.32). Body condition did not influence maximum bite endurance (linear regression: F_1,17=1.31, P=0.27, R²=0.07) or maximum bite force (linear regression: F_1,18=1.15, P=0.30, R²=0.06).

Morphometric variables (head length, head width and jaw length) were correlated (Pearson's correlation: head length and head width: t_1,21=3.93, P<0.001, r=0.70; head length and jaw length: t_1,21=4.71, P<0.001, r=0.74; head width and jaw length: t_1,21=2.25, P<0.05, r=0.5). Maximum bite endurance did not relate to jaw length, head length or head width (Table 2). Maximum bite force was related positively to head length and jaw length, but did not relate to head width (Table 2).

None of the explanatory variables were significantly related to bite endurance; the best model was the null model with fixed effects (ANOVA likelihood ratio=6.17; P=0.19; Fig. 4A). Table 3 shows the full model with all explanatory variables.

Bite force was negatively influenced by body temperature (Fig. 4B) and by the interaction between body mass and HL. Force was positively influenced by BM and HL (Table 3). In addition, D. leopardinus showed differences in bite force among the T_b values analyzed (Friedman repeated-measures ANOVA on ranks: χ²_4,20=29.08, P<0.001), with a strong bite magnitude at the lowest body temperatures (Tukey post hoc test: 20°C–34°C, q=5.2; 20°C–28°C, q=3; 24°C–34°C, q=3.5; 32°C–34°C, q=3.1; all P<0.05; Fig. 4B).

We found differences in bite force between treatments (Fig. 5): individuals acclimated to 30°C showed greater bite force compared with those acclimated to 20°C (Table 4, Fig. 5).

In the 20°C acclimation treatment, D. leopardinus showed differences in bite force among the different T_b analyzed (one-way repeated-measures ANOVA: F_4,9=5.03, P<0.05). The highest T_b studied showed the lowest bite magnitude compared with the rest of body temperatures (Tukey test: 20°C–34°C, q=5.34; 24°C–34°C, q=4.61; 28°C–34°C, q=5.34; all P<0.05). In the 30°C acclimation treatment, there were no differences in bite force among the different T_b analyzed (Friedman repeated-measures ANOVA on ranks: χ²_4,8=1.72, P=0.78).

The leopard iguana inhabits the Altoandina and Patagonica phytogeographic provinces, which are characterized by low mean temperatures, high solar radiation and high daily thermal variation (Corte and Espizua, 1981). In this type of environment, thermoregulatory efficiency is crucial to reduce the costs of thermoregulation such as the energy spent in locomotion or the risks of predation when basking (DeWitt, 1967). Diplolaemus leopardinus exhibited a low mean T_b (26.69±1.18°C) compared with other Andean lizards from northern latitudes, such as Phymaturus antofagastensis (31.6±2.6°C; Cruz et al., 2009), P. punae (29.45±0.48; Ibargüengoytía et al., 2008), Liolaemus yanalcu (34.16±2.21°C), L. irregularis (36.1±1.78°C), L. multicolor (36.52±1.33°C) and L. albiceps (36.17±1.92°C; Valdecantos et al., 2013), P. palluma (31.67±0.25°C; Vicenzi et al., 2017), and P. extridilus (32.32±3.13°C) and L. parvus (33.74±2.46°C; Gómez Alés et al., 2018). But, the mean T_b was similar to that of other leiosaurids such as Pristidactylus achalensis (mean 25.65°C; Pilustrelli et al., 1987), P. scapulatus (27.00±1.8°C; Villavicencio et al., 2006), P. torquatus (22.7°C) and P. volcanensis (27.80°C; Labra, 1995), and P. valeriae (26.10°C; Sufán-Catalán and Ñúñez, 1993). The dependence of body temperature on T_s shows that D. leopardinus is mainly thigmothermic and individuals thermoregulate by shifting among rocks exposed to direct solar radiation, areas shaded by vegetation or rocks, and their shelter. This lizard behaves as a moderate thermoregulator (E=0.57), and despite the low thermal quality of their environment (d_e=10.31±1.27), 43% of the lizards were able to attain T_b values in the field that are between the T_set values registered in the thermal gradient in the laboratory.

Leopard iguanas were not sexually dimorphic in bite endurance or in bite force. This is in contrast with several species with dimorphism in bite force caused principally by sex differences in body size, head shape and reproductive behaviour (Huyghe et al., 2009; McBrayer and Anderson, 2007). Males have greater M_b and bite force than females in several species such as Elgaria coerulea (McBrayer and Anderson, 2007), Sauromalus ater (Lappin and Husak, 2005), Sceloporus clarkii, S. undulatus, Urosaurus ornatus and Cophosaurus texanum (Meyers and Irschick, 2015), and Lacerta oxycephala and Podarcis melisellensis (Verwaijen et al., 2002). In D. leopardinus, the maximum bite force depended on head length and jaw length, as predicted. A larger head results in greater absolute volume of jaw muscle, conferring a bite performance advantage as observed in other lizards in which head size was a good predictor of bite force (Anderson et al., 2008; Herrel et al., 1999, 2001a,b, 2002; Verwaijen et al., 2002; Gomes et al., 2020 preprint).

Bite endurance was completely independent of temperature within the range of T_b measured (Fig. 4A); based on these results, we can hypothesize that the muscle fibers used for sustaining tension are less sensitive to temperature than those involved in the generation of force. Whereas great force requires fast-twitch muscle fibers, which are more easily exhausted, endurance needs slow-twitch muscle fibers, which are less easily fatigued albeit incapable of great force (Herrel et al., 1999). Previous studies, such as Gomes et al. (2020 preprint), have evaluated this novel trait in Podarcis bocagei as the time at which the bite force curve crossed at half the maximal value. Those authors found that, in contrast to bite force, bite endurance is not influenced by sex, and is probably related to muscle architecture, which shows few differences between males and females in Podarcis species (Gomes et al., 2020 preprint). Accordingly, we did not find differences between sexes in endurance; however, we measured bite duration differently than Gomes et al. (2020 preprint), because we considered the total time that lizards maintained a sustained full-mouth grip (Fig. 2). Diplolaemus leopardinus showed a bite pattern in which shorter peaks of bite magnitude alternated with periods of sustained bite. This bite pattern is congruent with feeding behavior, where longer periods of prey holding could alternate with short mash chewing movements (N. Vicenzi, unpublished observations). Moreover, this bite pattern is consistent with mating behavior in leiosaurid lizards. During courtship and mating, the male subjugates the female by sustaining neck-bite hold for more than 1 h (A. Laspiur, unpublished data). The male assumes a dominant position on the female dorsum with a sustained neck-bite hold, then strongly grasps the skin of the neck so as to stretch the neck and lift the stiff body of the female, and bring the cloaca closer to that of the male to facilitate the introduction of the hemipenis (A. Laspiur, unpublished data). This type of behavior was observed in several other species of lizards, such as P. bocagei (Galán, 1997), Tropidurus spinulosus (Pelegrin, 2019) and Phymaturus cf. palluma (Eisenberg and Werning, 2012).

Mean maximum bite force, before acclimation treatments (recorded at 20°C T_b), dropped by more than half at 34°C T_b (Fig. 4B). In contrast, previous studies have shown that muscle force decreases significantly below 25°C (Herrel et al., 2007; Marsh and Bennett, 1986); in D. leopardinus, the better temperature for maximum bite force performance was 20°C, lower than the mean T_b of activity (26.69±1.18°C) and lower than the T_pmean of the species (30.85±0.26°C). Similar results were found in B. pumilum, which has an optimal bite temperature of 23.7±1.0°C and a T_pmean of 29.3±1.3°C (Segall et al., 2013). The better temperature of 20°C was similar to the minimum T_b of activity registered in the field (19.50°C) and the mean T_e obtained at the study site (19.42±0.49°C). These results suggest that if lizards are cold, they would still be able to feed and defend themselves, and they may prefer these low temperatures even though they have the potential to attain higher body temperatures by thermoregulation. The maximum bite force recorded at low T_b in the leopard iguana could be related to the ‘fight versus flight’ behavioural response observed in several species of lizard (Herrel et al., 2007; Hertz et al., 1982). According to the theory on this behavioral response, at higher T_b lizards escape quickly from potential predators, but at lower T_b they do not run but stand their ground and attack aggressively (Hertz et al., 1982). Greater bite force at lower T_b by leopard iguanas is thus consistent with this theory. Several authors suggest that behavioural functions can generate evolutionary changes in morphology, physiology and ecology (e.g. Mayr, 1963; Wyles et al., 1983), and the ‘fight versus flight’ behaviour of D. leopardinus reinforces this hypothesis.

Bite force proved to be a plastic phenotypic trait when individuals were exposed to short-term acclimation treatments (Fig. 5). Phenotypic plasticity is expected in species that inhabit highly variable or unpredictable environments, such as the one in this study, as long as reliable cues exist to prompt beneficial plastic responses and the costs of plasticity are relatively low (Angilletta, 2009; Ghalambor et al., 2007; Marais and Chown, 2008). Leopard iguanas showed plasticity in both degree and pattern of the thermal function between T_b and bite force when acclimated to a constant temperature (Fig. 5). In animals acclimated at 30°C, we observed an increase in performance at higher temperatures, as the ‘hotter is better’ hypothesis predicts, but also a different pattern as bite force was almost constant across all T_b measurements (Fig. 5). This means that lizards exhibit changes in the slope value of the thermal reaction function and in its pattern (Pigliucci, 2001). Greater bite force implies the ability to capture larger prey and to reduce the time needed to handle and subdue prey, improving foraging efficiency (Herrel et al., 2001b, 1996). At higher acclimation temperatures, an increase in metabolic rate and, consequently, an increase in food requirements are expected (Dillon et al., 2010). Thus the greater bite force observed in the 30°C treatment could have ecological relevance because it improves prey quality and reduces handling cost.

Usually, studies about the thermal sensitivity of performance traits are based on physiological variables considered to be highly temperature dependent, such as locomotion speed or endurance, and not on traits typically considered less temperature dependent, such as bite and tail force (but see Diele-Viegas and Rocha, 2018; Segall et al., 2013). In this study, we focused on bite endurance and force, traits considered to have low temperature dependence and hence less affected by the increase in temperatures in a global warming scenario. However, we found that for D. leopardinus, temperature exerts a significant influence on bite force, and moreover, that bite force is a trait that showed thermal plasticity in short-term acclimations. Our results suggest that, at least in this trait, lizards are able to maintain and even improve bite performance with the projected rising temperatures. However, we also must consider that, in the Central Andes, the predictions for future climate change include more than an increase in mean temperature; they also include an increase in the frequency of extreme events such as drought and heat waves (Barros et al., 2014). Our study of acclimation was carried out with controlled and constant temperatures and with water ad libitum. Thus, a fuller understanding of lizard responses to global warming would require studies incorporating temperature variation and other sources of stress such as drought.

In the leopard iguana, T_pmean is significantly higher than maximum body temperature for bite force in lizards before acclimation, probably related to the fact that different physiological processes are taking place, influencing the thermal coadaptation between T_p and optimum temperature (Angilletta, 2009). Such differences in thermal optimum of other traits has been observed in several species, such as Liolaemus sarmientoi, in which T_pmean (34.43°C) was similar to optimum temperatures for long runs (34°C), but much higher than for sprint runs (27°C; Fernández et al., 2011). As with bite force, sprint runs are mostly related to defensive behaviors (Cabezas-Cartes et al., 2014). Present results indicate that this population of D. leopardinus shows phenotypic plasticity in bite force in laboratory conditions; thus lizards may have the plasticity to acclimate to short-term changes in environmental temperatures, and they may be able to adapt to a more gradual change, at least in bite force. However, according to our results, it seems possible that at higher temperatures (as exemplified by lizards acclimated to 30°C), individuals may lose their thermal dependence in bite force in response to thermal changes (Fig. 4B). This loss could produce a decoupling of the ‘fight or flight’ behavioral strategy and a significant increase in maintenance cost for the individuals. To sustain a maximal performance across different body temperatures, in both thermally dependent traits, would represent a significant increase in maintenance costs for the individuals, which could compromise the energy balance with negative implications for the persistence of the species.

We thank Jesús Pizarro for field support, Dr Mariano Gómez Berisso, Ignacio Artola and Dr Facundo Cabezas-Cartes for the design and construction of the bite force equipment, and Cielo Linares, Valeria Corbalán, IADIZA and the Dirección de Recursos Naturales Renovables of Mendoza government for their support in conducting this research. Marion Segall and an anonymous reviewer made insightful comments that substantially improved the manuscript.