Bateman–Trivers in the 21st Century: sexual selection in a North American pitviper BRENNA A. LEVINE 1 *, CHARLES F. SMITH 2,3,4 , GORDON W. SCHUETT 3,4,5 , MARLIS R. DOUGLAS 1,6 , MARK A. DAVIS 6 and MICHAEL E. DOUGLAS 1,6 1 Department of Biological Sciences, University of Arkansas, Fayetteville, AR 72701, USA 2 Department of Biology, Wofford College, Spartanburg, SC 29303, USA 3 The Copperhead Institute, PO Box 6755, Spartanburg, SC 29304, USA 4 Chiricahua Desert Museum, PO Box 376, Rodeo, NM 88056, USA 5 Department of Biology and Center for Behavioral Neuroscience, Georgia State University, Atlanta, GA 30303, USA 6 Illinois Natural History Survey, Prairie Research Institute, University of Illinois, Champaign, IL 61820, USA Received 14 July 2014; revised 4 September 2014; accepted for publication 5 September 2014 Assessment of sexual selection in organisms with cryptic life histories is challenging, although accurate parentage assignments using genotypic markers, combined with behavioural observations and a method to account for open population bias, allow for robust estimation of metrics. In the present study, we employed 22 tetranucleotide microsatellite DNA loci to interpret mating and reproductive success in a population of Copperhead (Viperidae, Agkistrodon contortrix) in Connecticut, USA. We sampled DNA from 114 adults (56 males, 58 females) and 137 neonates from known mothers to quantify Bateman gradients (βss), as well as sex-specific opportunities for selection (I) and sexual selection (Is). We also estimated selection on male size [snout-to-vent length (SVL)], a trait important for successful combat and subsequent copulations. Estimates of male I and Is differed significantly from those of females when estimated with four different methods and only males had a significant Bateman gradient. As predicted, male reproductive success was positively correlated with increasing SVL. These results contrast with those derived in another study investigating the same population but based solely on observational data and without correction for open population bias. We thus argue that molecular approaches to quantifying reproductive success and strength of sexual selection provide more accurate results than do behavioural observations alone. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2015, 114, 436–445. ADDITIONAL KEYWORDS: Agkistrodon contortrix – Bateman’s principles – copperhead – male body size – mating systems – reproductive success. INTRODUCTION The characterization of mating systems, quantifica- tion of reproductive success, and identification of phe- notypic characters influenced by sexual selection have a long history of investigation and are clearly impor- tant in both micro- and macro-evolutionary processes (Andersson, 1994; Shuster & Wade, 2003). Bateman’s (1948) principles, derived from laboratory breeding experiments using Drosophila, are particularly useful in understanding mating system evolution by allow- ing sexual selection gradients (i.e. Bateman gradi- ents; Arnold & Duvall, 1994) to be estimated via regression of reproductive success onto mating success, and by facilitating estimation of the oppor- tunity for selection (I; also referred to as opportunity for fecundity selection; Scharf & Meiri, 2013) and the opportunity for sexual selection (I s) in both males and females (for definitions of terminology, see Appendix; Wade, 1979; Jones, 2009; Mobley & Jones, 2012). Bateman (1948) predicted that variances in mating *Corresponding author. E-mail: blevine@email.uark.edu Biological Journal of the Linnean Society, 2015, 114, 436–445. With 2 figures © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2015, 114, 436–445 436