A major unresolved question for omnivorous carnivores, like most species of bears, is to what degree are populations influenced by bottom–up (food supply) or top–down (human-caused mortality) processes. Most previous work on bear populations has focused on factors that limit survival (top–down) assuming little effect of food resource supply. When food resources are considered, most often they consider only the availability/supply of a single resource, particularly marine-subsidized or terrestrial sources of protein (carnivory) or alternately hard or soft mast (frugivory). Little has been done to compare the importance of each of these factors for omnivorous bears or test whether complementary resources better explain individual animal and population measures such as density, vital rates, and body size. We compared landscape patterns of digestible energy (kcal) for buffaloberry (a key source of carbohydrate) and ungulate matter (a key source of protein and lipid) to local measures in grizzly bear Ursus arctos abundance at DNA hair snag sites in west-central Alberta, Canada. We tested support for bottom–up hypotheses in either single (carnivory [meat] versus frugivory [fruit]) or complementary (additive or multiplicative) food resources, while accounting for a well-known top–down limiting factor affecting bear survival (road density). We found support for both top–down and bottom–up factors with complementary resources (co-limitation) supported over single resource supplies of either meat or fruit. Our study suggests that the availability of food resources that provide complementary nutrients is more important in predicting local bear abundance than single foods or nutrients (e.g. protein) or simply energy per se. This suggests a nutritionally multidimensional bottom–up limitation for a low density interior population of grizzly bears.