Journal of Herpetology , Vol. 50, No. 4, 627–632, 2016 Copyright 2016 Society for the Study of Amphibians and Reptiles Scanning Snakes to Measure Condition: A Validation of Quantitative Magnetic Resonance JULIA L. RILEY , 1,2 JAMES H. BAXTER-GILBERT , 3,4 CHRISTOPHER G. GUGLIELMO, 5 AND JACQUELINE D. LITZGUS 3 1 Magnetawan First Nation, Britt, Ontario, Canada 3 Department of Biology, Laurentian University, 935 Ramsey Lake Road, Sudbury, Ontario, Canada 5 Department of Biology, University of Western Ontario, 1151 Richmond Street, London, Ontario, Canada ABSTRACT.—Body composition is a measure of an animal’s energetic state that can inform many research fields, yet the analysis traditionally requires individuals to be killed, and chemical analysis is labor intensive. Quantitative magnetic resonance (QMR) measures body composition noninvasively in live and nonanesthetized animals. Our aim was to validate QMR analysis for snakes by comparing it with gravimetric chemical analysis. We collected Northern Watersnakes (Nerodia sipedon sipedon) and Eastern Massasaugas (Sistrurus catenatus catenatus) that were found dead on roads, analyzed their body composition using the QMR scanner, and then by gravimetric chemical analysis. We compared fat mass, wet lean mass, and total water mass between the two methods, and then calculated bias, absolute error (g), and relative error (%) of the QMR analysis. Body composition values from the QMR analyses were highly correlated with the values obtained by gravimetric chemical analysis. Bias and errors were reasonable for wet lean and total water mass values, but the raw QMR data overestimated fat mass. When we calibrated the QMR using the chemical extraction data, it nearly eliminated bias and greatly reduced absolute and relative error. Therefore, following calibration, QMR analysis is an effective method to measure body composition of snakes. QMR very accurately measures wet lean and total water masses and can be used to detect changes in fat mass particularly in longitudinal studies of individuals across seasons. The energetic state of an animal is often quantified and summarized by biologists as their body condition. An animal in good body condition will have greater energy reserves than one in poor body condition (Schulte-Hostedde et al., 2005; Litzgus et al., 2008). The state of an individual’s body condition plays an integral role in its ecology and fitness, and understanding the implications of body condition for an individual is of great interest to ecologists. Body condition is directly related to individual health and can have effects on bodily processes such as immunocompetence (Arsnoe et al., 2011) and senescence (Kirkwood and Austad, 2000). Body condition varies with temporal fluctuations in the environment (Piersma and van Gils, 2011), habitat quality (Oliva-Paterna et al., 2003), maternal effects (Litzgus et al., 2008), animal life-history characteristics (Reed et al., 2008), and genotype (Blanckenhorn and Hosken, 2003). All of the above contribute to individual fitness parameters, and the impact of body condition on fitness parameters has been documented in a variety of taxa including mammals (Schulte-Hostedde et al., 2005), amphibians (Lowe et al., 2006), birds (O’Dwyer et al., 2006), and reptiles (Shine et al., 2001). Generally, individuals with greater energy reserves and better body condition experience higher survival and reproduc- tive success than do individuals with lower body condition (Litzgus et al., 2008). Body condition also is associated with the cost-benefit trade-offs of individual ecological strategies, such as dispersal and mating (Chastel et al., 1995; Cotton et al., 2006). Body condition can fluctuate over various timescales (e.g., seasonal or daily), and as such, ecologists track temporal changes in an animal’s body condition to examine what effects these temporal variations may have on fitness. Overall, body condition is closely tied to an animal’s ecology; hence, a logistically simple, noninvasive, and accurate method of measuring body condition across taxa is critical. The term ‘‘body condition’’ is used in the biological literature to describe various concepts. Most often it is used to refer to an indicator of an individual’s energy reserves. In this paper, we refer to both body condition and ‘‘body composition’’ as measures of an animal’s condition. We use body condition when speaking about estimates of individual energy reserves derived from morphometric measurements (e.g., mass and body length; Labocha et al., 2013) and body composition when referring to estimates of condition based on the relative proportions of protein, fat, water, and mineral components in the body. We use both terms because different fields of biology tend to refer to the energy reserves of an animal differently (e.g., body condition in ecology, and body composition in physiolo- gy). Pros and cons of the various methods of measuring body condition have been debated in the literature (Jakob et al., 1996; Green, 2001; Schulte-Hostedde et al., 2001, 2005), because accurately measuring body condition can be difficult. A commonly used, nondestructive method of estimating body condition is to relate body mass to a linear measure of body size (as reviewed by Brown, 1996). This method of body condition estimation aims to determine the amount of energy reserves an individual has after correcting for structural body size (Schulte- Hostedde et al., 2005). Commonly, body condition indices are generated from the relationship between body mass and a linear measurement of body size using the residuals from this regression (Schulte-Hostedde et al., 2005, but see Garcı ´a- Berthou, 2001, and Green, 2001). Individuals with positive residuals are considered to have better body condition than individuals with negative residuals (Schulte-Hostedde et al., 2001). Yet, confounding factors (e.g., time since last feeding or defecation, gravid females, etc.) decrease the accuracy of this method. Body condition indices derived from morphometric measurements (e.g., mass and body size) may be prone to human error. Also, body condition indices often are inconsis- tently calculated by researchers, which can lead to erroneous 2 Corresponding author. Present address: Department of Biological Sciences, Macquarie University, Sydney, New South Wales, Australia; Email: julia.riley@students.mq.edu.au 4 Present address: Department of Biological Sciences, Macquarie University, Sydney, New South Wales, Australia DOI: 10.1670/15-113