“…If properly implemented, BSIMMs can produce accurate, probabilistic estimates of animal diets (Moore & Semmens, 2008;Parnell et al, 2013), yet concerns have been raised over misuse and sensitivity to input parameters (Boecklen, Yarnes, Cook, & James, 2011;Derbridge et al, 2015;Franco-Trecu et al, 2013;Martínez Del Rio, Wolf, Carleton, & Gannes, 2009). This has led to attempts to evaluate BSIMMs through experimental and observational studies (Derbridge et al, 2012(Derbridge et al, , 2015Franco-Trecu et al, 2013;Newsome, Collins, & Sharpe, 2015;Resano-Mayor et al, 2014;Weiser & Powell, 2011). Although studies of captive animals in controlled conditions provide a powerful approach to testing mixing model performance (Caut, Angulo, & Courchamp, 2008;Derbridge et al, 2015), they can lack the variation in diet and physiology typical of wild animals (Boecklen et al, 2011).…”