1 | INTRODUCTION
Some terms in ecology hold specific connotations, that when used, can limit our perceptions about ecological processes, bias our results, and hinder our understanding of the natural world. For instance, the term “carnivore” can be confusing in ecology, either referring specifically to members of the mammalian order Carnivora, or more generally referring to animals (typically mammals) that kill and consume other vertebrates (i.e., meat-eaters; Gittleman, 2013; Yoshimura et al., 2021). However, outside a few families of obligate carnivores (felids and hyenids; Clauss et al., 2010; Yoshimura et al., 2021), members of Carnivora eat a wide variety of foods, including invertebrates, plants and fungi (Edwards et al., 2011). Additionally, many species in groups outside Carnivora, such as primates, rodents, raptors, reptiles, amphibians, fish, invertebrates, and plants (Román-Palacios et al., 2019), kill and consume vertebrates. Thus, classic foraging categories of carnivore, omnivore, and herbivore may be overly simplistic (Harrer & Levi, 2018; Leigh et al., 2018; Vazquez et al., 2023) and lead to a narrow focus on an organism’s diet. Using a more precise term, such as polyphagous (Loxdale & Harvey, 2023), may combat this issue.
Food webs have traditionally modelled polyphagous species as static consumers exerting constant pressure on multiple resources, but diet specificity changes with environmental conditions and is essential to ecosystem stability (Gutgesell et al., 2022). Primary productivity varies annually related to climate and insect or pathogen abundance (Bjerke et al., 2014) and organisms can respond to this variation by changing their diet. Thus, polyphagous species can rapidly respond to changes in resource availability (Deacy et al., 2018), relieving pressure from an exhausted resource while increasing pressure on another, more abundant resource, even if the second resource is a less-preferred food (Zhang et al., 2021). The historical focus on predators through simple interactions between large, charismatic, obligate carnivores and their prey over-simplifies many polyphagous species’ roles within ecological communities (Miller et al., 2001). This likely overlooks trophic interactions which may be important to community stability (Kratina et al., 2012) and essential to predict community shifts in a changing climate (Gutgesell et al., 2022).
Despite the need for a deeper understanding of polyphagous species beyond their roles as predators, measuring diet in wild species is difficult (Davis & Pineda‐Munoz, 2016) and modeling environmental features associated with diet changes is challenging due to the complexity of ecological communities. One way to estimate diet in wildlife is through stable isotope analysis (Tieszen & Boutton, 1989). This method is based on the principal that the ratio of naturally occurring stable isotopes varies across the earth and among different food types, such as C-4 and C-3 plants or marine and terrestrial animals, and all organisms must build their tissues from molecules they consume (Tieszen & Boutton, 1989). Thus, the stable isotope value of an organisms’ tissue will most closely resemble that of its dominant foods (Semmens et al., 2009).
In addition to diet estimation, accurate measures of food availability are difficult to obtain, especially over time periods long enough to detect change (Davis & Pineda‐Munoz, 2016). Even within the same population of a single species, there will be differences in diet among individuals (Edwards et al., 2011), demographic classes (Beck et al., 2007), across space (Stern et al., 2024), and time (Davis & Pineda-Munoz, 2016), which can obscure general patterns. For example, within a species different populations may have different responses to variation in primary productivity, such as mast crops (Hertel et al., 2019; Schwartz et al., 2010). Thus, determining the effect of variation in primary productivity on diet of polyphagous species is challenging.
We used a dataset on a polyphagous mammal, the brown bear (Ursus arctos ), to estimate annual diet proportions of common foods as well as identify diet patterns and drivers over 25 years. We used carbon (δ 13C) and nitrogen (δ 15N) stable isotopes measured in hair of known individual bears in south-central Sweden to estimate annual diet among different demographic classes. We focused on the five primary diet components of bears in this system: ants (Formnica spp. and Camponotus spp. ), bilberry (Vaccinium myrtillus ), crowberry (Empetrum nigrum), lingonberry(Vaccinium vitus-vitae), and moose (Alces alces ; Stenset et al., 2016). Berry production in Scandinavia is variable both temporally and spatially (Hertel et al., 2016), while moose populations showed less variation (Jensen et al., 2020). Based on previous stable isotope analyses (Mikkelsen et al., 2023), we expected bilberry to make up the greatest proportion of the brown bear diet, however meat is considered higher quality food than berries (Pritchard & Robbins, 1990). Thus, while it accounts for a smaller diet proportion, we expected brown bear diet to respond to moose availability (Hypothesis 1); years with greater numbers of moose will be associated with greater proportions of moose in the diet. Moose populations have been increasing across Scandinavia (Jensen et al., 2020), and proportions of moose in the diet should increase through time as bears respond to increasing moose availability (Hypothesis 2). Annual variation in food productivity will also result in annual variation in stable isotope values. Specifically, we expected δ 15N values to be related to indices of moose population size andδ 13C values to be related to annual bilberry production (Hypothesis 3).