Test of theory of foraging mode: Goldcrests, Regulus regulus, forage by high‐yield, energy‐expensive hovering flight when food is abundant but use low‐yield, low‐cost methods when food is scarce

1. Here, I describe foraging behavior of goldcrests, Regulus regulus, based on eight years of field observation in a coniferous forest dominated by Norway spruce Picea abies in southwestern Sweden. The aim was to test predictions from theory on the choice of optimal foraging modes in relation to food availability. 2. Mortality from early November to early March amounts to 70–86% among goldcrests in the resident population, suggesting they are food‐limited in winter. Food‐limitation manifests itself as a shortage of time for foraging. It promotes the use of foraging methods that minimize the daily foraging time by maximizing the rate of net energy gain. It increases both individual survival and competitiveness. Elimination of competitors by exploitation occurs when an individual is able to support itself, while food density in the habitat is reduced to levels at which others cannot. 3. Theory shows that when food is abundant, high‐efficiency energy‐expensive search and capture methods give shorter daily foraging times than low‐efficiency low‐cost methods, whereas the latter gives shorter daily foraging times at food shortages (Norberg 2021). Hovering flight is extremely expensive in energy but results in high foraging efficiency. Hover‐foraging should therefore be used when food is abundant. 4. In autumn, there were 85.3 arthropods per kilogram of branch mass, as opposed to 12.9 in spring. The numerical decline of arthropods, their fat metabolism, and size‐biased predation by birds reduced the spring density of food for goldcrests to less than 15.1% of the autumn density. 5. Hover‐foraging occurred 5.29 times per minute in autumn but only 0.23 times per minute in spring, which is 4.4% of the autumn frequency. 6. Foraging conditions are favorable at midsummer because of long days, high temperatures, and an abundance of arthropod prey. Parent birds that were feeding fledglings gathered food at a high rate and hovered 5.42 times per minute. But adults with no young to feed were not compelled to maximize the rate of net energy gain and only hover‐foraged 0.52 times per minute, which is 10% of that of providers. 7. These results are highly consistent from year to year and in qualitative agreement with theory.