DIETARY ANALYSIS FROM FECAL SAMPLES: HOW MANY SCATS ARE ENOUGH?

摘要: Diets of mammals are increasingly being inferred from identification hard parts prey eaten and recovered in fecal remains (scats). Frequencies with which particular species occur among collections scats easily compiled to describe the average diet, can be used compare diets between within geographic regions, across years seasons. Important these analyses is question statistical power. In other words, how many should collected diet species? We addressed this problem by using Monte Carlo simulations frequency occurrence methods analytically determine consequence sample size on dietary analysis scats. considered 2 questions. First, power affected size? Second, what likelihood not identifying a randomly sampled predetermined numbers (n ¼ 10‐200) computer-generated populations containing known frequencies occurrences. also large database field-collected Steller sea lions (Eumetopias jubatus). then standard contingency table tests such as chi-square Fisher’s exact test whether differences our samples were statistically significant. found that minimum 59 necessary identify principal occurring .5% However, 94 required when comparing distinguish moderate effect sizes over time or areas. These findings have significant implications for interpretation published data, well design future scat-based studies pinnipeds species.