ABSTRACT: Apennine brown bears (Ursus arctos marsicanus) survive in an isolated and critically endangered population, and their food habits have been studied using traditional scat analysis. To complement current dietary knowledge, we applied Stable Isotope Analysis (SIA) to non-invasively collected bear hairs that had been individually recognized through multilocus genotyping. We analysed carbon (δ13C) and nitrogen (δ15N) stable isotopes of hair sections and bear key foods in a Bayesian mixing models framework to reconstruct the assimilated diet on a seasonal basis and to assess gender and management status effects. In total, we analysed 34 different seasonal bear key foods and 35 hair samples belonging to 27 different bears (16 females and 11 males) collected during a population survey in 2014. Most bears showed wide δ15N and δ13C ranges and individual differences in seasonal isotopic patterns. Vegetable matter (herbs, fleshy fruits and hard mast) represented the major component of the assimilated diet across the dietary seasons, whereas vegetable crops were rarely and C4 plants (i.e., corn) never consumed. We confirmed an overall low consumption of large mammals by Apennine bears consistently between sexes, with highest values in spring followed by early summer but null in the other seasons. We also confirmed that consumption of fleshy fruits peaked in late summer, when wild predominated over cultivated fleshy fruits, even though the latter tended to be consumed in higher proportion in autumn. Male bears had higher δ 15N values than females in spring and autumn. Our findings also hint at additional differences in the assimilated diet between sexes, with females likely consuming more herbs during spring, ants during early summer, and hard mast during fall compared to males. In addition, although effect sizes were small and credibility intervals overlapped considerably, management bears on average were 0.9‰ lower in δ 13C and 2.9‰ higher in δ 15N compared to non-management bears, with differences in isotopic values between the two bear categories peaking in autumn. While non-management bears consumed more herbs, wild fleshy fruits, and hard mast, management bears tended to consume higher proportions of cultivated fruits, ants, and large mammals, possibly including livestock. Although multi-year sampling and larger sample sizes are needed to support our findings, our application confirms that SIA can effectively integrate previous knowledge and be efficiently conducted using samples non-invasively collected during population surveys.
