Many animal ecology studies rely on methods to determine the body condition of different individuals in a population (Stevenson & Woods 2006). Both destructive (body composition) and non-destructive (body mass and linear measures of body size) are used to estimate or determine the condition index (CI) of an individual. Body condition is assumed to influence an animal's health and fitness and may affect many aspects in an organism's life such as social status (dark-bellied brent geese Branta bernicla bernicla, Poisbleau, 2006), reproductive success (Crimson Finch Neochmia phaeton, Milenkaya et al. 2015), foraging strategy (White-tailed Deer Odocoileus virginianus, Taillon, & Côté, 2007; Meerkat Suricata suricatta, Thornton, 2008), survival through stressed periods (Golden-mantled Ground Squirrels Callospermophilus lateralis, Wells et al. 2019), disease status (green sea turtle Chelonia mydas, Rossi et al. 2019) and dispersal (Viviparous Lizard Zootoca vivipara, Meylan et al. 2002). It is highly desirable to understand body condition, both temporally and ontogenetically in order to provide supporting evidence and mechanistic linkages for population studies (Stevenson & Woods 2006). In population studies of species of conservation importance, new and improved non-destructive methods for body condition indices are increasingly important (Speakman 2001, Peig & Green 2009).
Body fat, due to its high energy content, is a good estimator of body condition of an animal (Hadley, 1985). Body fat reserves directly influence fitness and are highly dependent on season, reproductive status and periods of fasting (e.g., Atkinson & Ramsay 1995, Stephenson et al. 2002). The most accurate measure of body fat is the direct approach wherein several individual animals are euthanized and have the fat extracted chemically from the carcasses (e.g., Xuefeng & Yilian 2003, Prestrud & Pond 2003). This destructive approach is, however, complicated, time consuming and does not allow for comparisons of body composition within and between seasons on the same individuals. Non-destructive techniques include various body condition indices (ratio of body mass to a linear dimension of body size, or the residuals of the regression between body mass and size; Green 2001; Secor & Nagy, 2003), isotope dilution (Servello, et al. 2005), bioelectrical impedance analysis (BIA, Hwang et al. 2005), total body electrical conductivity (TOBEC, Angilletta 1999, Scott et al. 2001), lipid-soluble gas absorption (Henen, 2001), quantitative magnetic resonance (QMR) (Riley et al. 2016), and dual-energy X-ray absorptiometry (DXA) (Nagy 2001).
Bioelectrical Impedance Analysis (BIA) appears to be a better predictor of energy stores than body condition estimates calculated from mass and SVL (Wirsing et al. 2002). However, the repeatability and accuracy are not sufficient to monitor small changes in Lean Body Mass (LBM) and lipid stores (Secor and Nagy, 2003). Among the alternative techniques, DXA, holds the most promise as an easy and accurate measure, especially for smaller animals. DXA scans the body with two X-ray beams of different energy levels and uses the attenuation of the energy of those two X-ray beams to determine the tissue signature and to quantify total body mass, lean mass and fat mass of the organism (Korine et al. 2004).
DXA has been used to assess nutritional status in captive rhesus monkeys Macaca mulatta (compared with stable isotope dilution, no validation, Blanc et al. 2005); determine bone density distribution patterns in museum skulls of delphinids and beaked whales (Cozzi et al. 2010); measure fat mass in small migratory birds (validated with freeze-dried carcasses, Korine et al. 2004), identify metabolic bone disease and bone mineral density in captive green iguanas Iguana iguana (no validation, Zotti et al. 2004), determine the body composition of diamondback water snakes Nerodia rhombifer (validated with euthanized individuals, Secor & Nagy, 2003) and channel catfish Ictalurus punctatus (validated with euthanized individuals, Johnson et al. 2017). In reptiles, body condition is generally determined by morphometric measurements (Hayes & Shonkwiler 2001, Stevenson & Woods 2006) which are rarely tested against other CIs or validated (but see Secor & Nagy, 2003, Falk et al. 2017). The use of DXA has not been tested for its applicability for studies of small lizards, thus our primary objective is to assess the use of DXA as a practical method for determining body fat composition, and therefore use as a CI, in a small lizard. Our specific goals are (1) to evaluate the accuracy of DXA in predicting body fat composition of a small lizard by comparing DXA to chemically extracted fat, (2) to compare the accuracy of this method for males, females and juveniles, and (3) to quantify the relationship between five common estimators of body condition (based on mass and/or length relationships) with body fat determined from DXA.