Methods of body temperature assessment in Conolophus subcristatus, Conolophus pallidus (Galápagos land iguanas), and Amblyrhynchus cristatus X C. subcristatus hybrid

Since cardiovascular, respiratory, and metabolic systems of reptiles are affected by temperature, accurate measurements are of great importance in both captive husbandry and research. Ectothermic animals generally have core body temperatures close to ambient temperature but can differ from the immediate environment if they are using sunlight to thermoregulate. Many zoological facilities and exotic pet caregivers have begun using infrared temperature guns to assess ambient temperatures of reptile enclosures but there are currently few studies assessing the efficacy of these devices for measuring the body temperatures of reptiles. Conolophus subcristatus, Conolophus pallidus (Galápagos land iguanas), and Amblyrhynchus cristatus X C. subcristatus hybrid are robust land iguanas endemic to the Galápagos archipelago. By comparing the infrared body temperature measurements of land iguanas against virtual simultaneous collection of cloacal temperatures obtained using a thermocouple thermometer, we sought to assess the efficacy of this non-invasive method. We found that internal body temperature can be predicted with a high level of accuracy from three external body temperature sites, providing a good non-invasive method that avoids the capture of animals.


INTRODUCTION
Since cardiovascular, respiratory, and metabolic systems of reptiles are affected by temperature, accurate measurements are of great importance in both captive husbandry and research (Sato et al., 1995;Deen & Hutchinson, 2001;Seebacher & Franklin, 2005;Long, 2016). Body temperature assessment is also an important part of veterinary health examinations (Music & Strunk, 2016;Cusack et al., 2018). While ectothermic animals generally have core body temperatures close to ambient temperature, it has been shown that the animal's temperature can differ from the immediate environment (Raske et al., 2012). The size (mass) of the animal would also effect this as smaller animals heat up and cool down more quickly than larger ones (Gillooly et al., 2001;Seebacher & Franklin, 2005).
Many zoological facilities and exotic pet caregivers have begun using infrared temperature guns to assess ambient temperatures of reptile enclosures (Rizzo, 2015), but there are only a few studies assing the efficacy of these devices for measuring the body temperatures of amphibians and reptiles (Hare, Whitworth & Cree, 2007;Rowley & Alford, 2007;Halliday & Blouin-Demers, 2017). By comparing the infrared body temperature measurements of land iguanas against virtual simultaneous collection of cloacal temperatures obtained using a thermocouple thermometer, we sought to assess the efficacy of this non-invasive method in a large terrestrial reptile.
Conolophus pallidus and C. subcristatus are land iguanas endemic to the Galápagos archipelago. The former species is restricted to Santa Fe Island and both species are classified as vulnerable by the IUCN Red List of Threatened Species (World Conservation Monitoring Centre, 1996a, 1996b. The hybrid has only been documented from South Plazas Island (Rassmann, Trillmich & Tautz, 1997). As part of a population health assessment authorized by the Galápagos National Park (GNP), wild iguanas from three islands (North Seymour, South Plazas, and Santa Fe) were captured in July 2018. Veterinary health examinations, that included sampling blood, ectoparasites, and feces, were performed on each animal in accordance with the ethics and animal handling protocols of Galápagos Science Center (GSC) and the GNP.

MATERIALS AND METHODS
This study was conducted in the Galápagos archipelago of Ecuador in July, 2018 as part of a population health assessment authorized by the GNP Service (Permit # PC-70-18 to G.A. Lewbart) and approved by the Universidad San Francisco de Quito ethics and animal handling protocol.
A total of 21 adult C. pallidus, 30 C. subcristatus, and one Amblyryhnchus cristatus X C. subcristatus hybrid weighing between 1.2 and 7.2 kg (mean = 3.9 kg) were captured for health assessments. All data were collected over a period of 4 days, with the ambient environmental temperature ranging from 20.5 to 26 C at the time of examination. Body temperature measurements with a Nubee Ò infrared temperature gun (Model NUB838OH) were collected from the right axillary space (T AXI ), dorsum (T DOR ), and right femoral space (T FEM ) (Fig. 1). The axillary space is the area just caudal to the insertion of the upper arm to the body. The dorsum is an area directly above the axillary location at the most dorsal part of the animal. The femoral space is the area just cranial to the insertion of the rear limb to the body. The device was held approximately seven cm from each location for 5-10 s (until the temperature stabilized). This device has a distant to spot ratio of 8:1. Thus, the area being measured is approximately one cm. Cloacal temperatures (T INT ) were measured using an EBRO Ò Compact J/K/T/E thermocouple thermometer (Model EW-91219-40; Cole-Parmer, Vernon Hills, IL 60061, USA) inserted approximately 10 cm into the vent. The thermocouple thermometer was held in place until a stable temperature reading could be recorded (approximately 15 s).

Statistical analysis
We used repeated-measures ANOVA with Huynd-Feldt (HF) method to correct for departure from sphericity (HF: ε = 0.7896778, Mauchly's Tests for sphericity: W = 0.54929, P 0.001) to test for temperature differences between the four methods. We then performed pairwise comparisons using paired t-test with Bonferroni-Holm correction. Given non-independence of measurements, we investigated the best predictors of internal (a.k.a. cloacal) body temperature (T INT ) in two complementary ways. Running simple linear models, we first looked at simple pairwise correlations using as regressors the three infrared external temperatures: Femoral (T FEM ), Dorsal (T DOR ), and Axillary (T AXI ). Then, we chose the external body temperature with the highest correlation coefficient as the best single predictor of internal temperature. We also extracted the first principal component of the three external temperature measures to be used as a single compound predictor. Statistical analyses were performed in R version 3.5.1 (R Development Core Team, 2018).

RESULTS
Average body temperatures measured at each site are displayed in Table 1. Figure 2 graphs all four temperature points for each animal. The average temperature differed significantly between the four methods (repeated-measures ANOVA: F 2.4,120.9 = 9.8397, P < 0.001); however, the effect size was small (g 2 = 0.01224529). Pairwise comparisons showed significant differences between T DOR and T AXI (t = 3.01, DF = 51, P = 0.012) and all three other pairwise comparisons (P < 0.01 for all cases). The non-significant differences included T INT -T DOR (t = 1.36, DF = 51, P = 0.359) and T DOR -T FEM (t = 0.68, DF = 51, P = 0.498). Simple pairwise correlation between temperature variables showed that all four measures were highly correlated (Fig. 2) highly and significantly with each of the three methods of recording external body temperature: T FEM (r = 0.967; t = 26.941, DF = 50, P < 0.001), T DOR (r = 0.928; t = 17.63, DF = 50, P < 0.001), T AXI (r = 0.918; t = 16.08, P < 0.001). The best single predictor (i.e., with the highest correlation coefficient) of internal body temperature (T INT ), running a simple linear model was femoral body temperature (T FEM ) (r 2 = 0.936, F 1,50 = 725.8, P < 0.001): where S E are standard errors of the estimated coefficients.
The predictive model of internal body temperature (T INT ), running a simple linear model using the compound variable (Principal component, PC1: r 2 = 0.948, F 1,50 = 903.8, P < 0.001) from all external body temperatures was (Fig. 3): where S E are standard errors of the estimated coefficients.

DISCUSSION
Thermal biology is a rich and dynamic area of biology that studies the way animals react and adapt to their environment by regulating body temperature. Camacho & Rusch (2017) provide a thorough and comprehensive review of the subject and the various methodologies that have been used to determine body temperature in lizards. Dozens of studies exist in reptiles over a range of topics that include preferred body temperature, voluntary thermal selection, and absolute thermal tolerance (Tattersall, 2016;Camacho & Rusch, 2017). Many of these studies use standard thermometers or thermocouples, but some use infrared cameras (Luna & Font, 2013) and others use infrared guns similar to the one used in our study (Hare, Whitworth & Cree, 2007;Rowley & Alford, 2007;Carretero, 2012;Halliday & Blouin-Demers, 2017). Carretero (2012) found that in small lacertid lizards of the genus Podacris, infrared readings were consistently higher than cloacal readings. Our study differs from the previous four works in the important area of animal size. The land iguanas in our study were between 10 times and 100 times larger than the small lizards, amphibians, and snakes in these studies. In our study the differences between the three methods of recording body temperature were small and all three external body temperatures slightly underestimated internal body temperature. Dorsal temperature was the closest to internal temperature, axillary was the lowest, and femoral was in the middle.
The best model predictor shows that internal body temperature can be predicted with a high level of accuracy from the three external body temperatures, providing a good non-invasive method that avoids the capture of animals. Femoral temperature, in spite of not being the most accurate approximation to internal body temperature, was the best single predictor to estimate internal body temperature as it was the least variable with the highest correlation coefficient. We speculate this is because the femoral area is closest to the thermocouple location (deep cloaca).
One issue to consider, although we did not experience it, would be an animal that has been in a burrow for a period of time and then climbs out into direct sunlight. In this case it's possible, if not likely, that the external temperature might not reflect the true core temperature. This would be an interesting angle to pursue with either this species or other reptiles. Another factor that should be considered is what is the preferred temperature (T pref ) for the land iguana. Based on the overall health of the animals in this study it may be similar to the T pref of 25-35 C for its mainland relative Iguana iguana, the green iguana (Girling & Raiti, 2004).

CONCLUSIONS
This non-invasive method for measuring body temperature using an infrared temperature gun is useful for approximating body temperature in Galápagos land iguanas. The quantitative assessment provides data that can be most likely be used for monitoring the health status of other similar sized lizard species in captive collections or in the field. Further evaluation is warranted for its applicability to other species of large reptiles.