High variability within pet foods prevents the identification of native species in pet cats' diets using isotopic evaluation.

Domestic cats preying on wildlife is a frequent conservation concern but typical approaches for assessing impacts rely on owner reports of prey returned home, which can be biased by inaccurate reporting or by cats consuming prey instead of bringing it home. Isotopes offer an alternative way to quantify broad differences in animal diets. By obtaining samples of pet food from cat owners we predicted that we would have high power to identify cats feeding on wild birds or mammals, given that pet food is thought to have higher C isotope values, due to the pervasive use of corn and/or corn by-products as food ingredients, than native prey. We worked with citizen scientists to quantify the isotopes of 202 cat hair samples and 239 pet food samples from the US and UK. We also characterized the isotopes of 11 likely native prey species from the southeastern US and used mixing models to assess the diet of 47 cats from the same region. Variation in C and N isotope values for cat food was very high, even within the same brand/flavor, suggesting that pet food manufacturers use a wide range of ingredients, and that these may change over time. Cat food and cat hair from the UK had lower C values than the US, presumably reflecting differences in the amount of corn used in the food chains of the two countries. This high variation in pet food reduced our ability to classify cats as hunters of native prey, such that only 43% of the animals could be confidently assigned. If feral or free ranging cats were considered, this uncertainty would be even higher as pet food types would be unknown. Our results question the general assumption that anthropogenic foods always have high C isotope values, because of the high variability we documented within one product type (cat food) and between countries (US vs. UK), and emphasize the need to test a variety of standards before making conclusions from isotope ecology studies.