Edinburgh Napier University

  The diet of wild animals has been studied using many different strategies, approaches and methods in recent decades. In this regard, stable isotopes analysis (SIA) is becoming a widespread tool, but no study has yet, to our knowledge, compared diet estimations from SIA with direct observations of the diet of passerine nestlings. Accordingly, our aim was to test the predictive power of SIA for this purpose and identify potential confounding factors such as habitat effects. To do this, we compared isotopic signatures of δ13C and δ15N in the feathers of great tit (Parus major) nestlings, and the corresponding estimates of their diet based on stable isotope mixing models, with prey proportions delivered by their parents obtained through video-recordings. Between-nest differences in isotopic signatures of δ15N were larger than within-nest differences. We found that δ15N signatures of nestling feathers correlated positively with the proportion of spiders and negatively with the proportion of caterpillars in the nestlings’ diet, the most important prey types. On the other hand, between-nest and within-nest differences in δ13C ratios were of similar magnitude and δ13C ratios correlated mainly with the proportion of trees surrounding nest-boxes that were Quercus spp. Estimates of diet composition based on mixing models correlated with the observed nestling diet, yet effect sizes were quite low. Although mixing models are commonly used to ascertain diets, our data show that they can provide valuable information on the relative intake of prey types from different trophic levels; but when complex dietary patterns are recorded (e.g. due to the confounding effects of habitat and/or temporal variation) it can be difficult to draw firm conclusions about diet composition.