| Summary: | Generalist populations often harbor individual dietary specialists. Whether using a narrower set of resources than the population (= specialization) affects specialists' nutrient intake remains underexplored. We evaluated variation in ingredient and nutrient specialization in a European brown bear Ursus arctos population via the Proportional Similarity Index (PS i , from 0 = highly specialized to 1 = not specialized) and assessed associations of specialization with year, season and reproductive class. Different methodologies concerning the organization of raw data for PS i calculations were evaluated (i.e. the resolution of diet compositional data (feces vs the average of all feces per individual) and temporal restrictions for the population (year‐round vs within‐season). Overall, a tendency for ingredient specialization (PS i 0.37 ± 0.14) and absence of nutrient specialization (PS i 0.79 ± 0.10) was observed. Ingredient specialization was mainly influenced by the proportion of berries, graminoids, oats and moose in the diet. Annual, seasonal and reproductive class effects were moderate and did not strongly affect PS i for both ingredients and nutrients. Organizing diet compositional data from a ‘feces resolution' to ‘individual resolution' decreased specialization. Changing the comparative population in PS i calculations from ‘all‐year‐round' to ‘within year and season' also resulted in less pronounced specialization. The degree of specialization was not caused by individuals exhibiting consistent ingredient preferences over the years (low repeatability of PS i ) except in spring. Our results suggested absence of nutrient specialization and mild ingredient specialization, which appeared to be an outcome of the ecological circumstances rather than specific individual traits. Additionally, we demonstrated that the methods applied can have substantial influence for the calculation of specialization indices.
|
|---|