Ursids evolved early and continuously to be low-protein macronutrient omnivores

Ursids evolved dietary diversity without major alterations in metabolic rates

The diets of the eight species of ursids range from carnivory (e.g., polar bears, Ursus maritimus) to insectivory (e.g., sloth bears, Melursus ursinus), omnivory (e.g., brown bears, U. arctos), and herbivory (e.g., giant pandas, Ailuropoda melanoleuca). Dietary energy availability ranges from the high-fat, highly digestible, calorically dense diet of polar bears (~ 6.4 kcal digestible energy/g fresh weight) to the high-fiber, poorly digestible, calorically restricted diet (~ 0.7) of giant pandas. Thus, ursids provide the opportunity to examine the extent to which dietary energy drives evolution of energy metabolism in a closely related group of animals. We measured the daily energy expenditure (DEE) of captive brown bears in a relatively large, zoo-type enclosure and compared those values to previously published results on captive brown bears, captive and free-ranging polar bears, and captive and free-ranging giant pandas. We found that all three species have similar mass-specific DEE when travel distances and energy intake are normalized even though their diets differ dramatically and phylogenetic lineages are separated by millions of years. For giant pandas, the ability to engage in low-cost stationary foraging relative to more wide-ranging bears likely provided the necessary energy savings to become bamboo specialists without greatly altering their metabolic rate.

Testing foraging optimization models in brown bears: Time for a paradigm shift in nutritional ecology?

How organisms obtain energy to survive and reproduce is fundamental to ecology, yet researchers use theoretical concepts represented by simplified models to estimate diet and predict community interactions. Such simplistic models can sometimes limit our understanding of ecological principles. We used a polyphagous species with a wide distribution, the brown bear (Ursus arctos), to illustrate how disparate theoretical frameworks in ecology can affect conclusions regarding ecological communities. We used stable isotope measurements (δ13 C, δ15 N) from hairs of individually monitored bears in Sweden and Bayesian mixing models to estimate dietary proportions of ants, moose, and three berry species to compare with other brown bear populations. We also developed three hypotheses based on predominant foraging literature, and then compared predicted diets to field estimates. Our three models assumed (1) bears forage to optimize caloric efficiency (optimum foraging model), predicting bears predominately eat berries (~70% of diet) and opportunistically feed on moose (Alces alces) and ants (Formica spp. and Camponotus spp; ~15% each); (2) bears maximize meat intake (maximizing fitness model), predicting a diet of 35%-50% moose, followed by ants (~30%), and berries (~15%); (3) bears forage to optimize macronutrient balance (macronutrient model), predicting a diet of ~22% (dry weight) or 17% metabolizable energy from proteins, with the rest made up of carbohydrates and lipids (~49% and 29% dry matter or 53% and 30% metabolizable energy, respectively). Bears primarily consumed bilberries (Vaccinium myrtillus; 50%-55%), followed by lingonberries (V. vitis-idaea; 22%-30%), crowberries (Empetrum nigrum; 8%-15%), ants (5%-8%), and moose (3%-4%). Dry matter dietary protein was lower than predicted by the maximizing fitness model and the macronutrient balancing model, but protein made up a larger proportion of the metabolizable energy than predicted. While diets most closely resembled predictions from optimal foraging theory, none of the foraging hypotheses fully described the relationship between foraging and ecological niches in brown bears. Acknowledging and broadening models based on foraging theories is more likely to foster novel discoveries and insights into the role of polyphagous species in ecosystems and we encourage this approach.