Ambient temperature does not affect fuelling rate in absence of digestive constraints in long-distance migrant shorebird fuelling up in captivity

Pre-flight fuelling rates in free-living red knots Calidris canutus, a specialized long-distance migrating shorebird species, are positively correlated with latitude and negatively with temperature. The single published hypothesis to explain these relationships is the heat load hypothesis that states that in warm climates red knots may overheat during fuelling. To limit endogenous heat production (measurable as basal metabolic rate BMR), birds would minimize the growth of digestive organs at a time they need. This hypothesis makes the implicit assumption that BMR is mainly driven by digestive organ size variation during pre-flight fuelling. To test the validity of this assumption, we fed captive knots with trout pellet food, a diet previously shown to quickly lead to atrophied digestive organs, during a fuelling episode. Birds were exposed to two thermal treatments (6 and 24°C) previously shown to generate different fuelling rates in knots. We made two predictions. First, easily digested trout pellet food rather than hard-shelled prey removes the heat contribution of the gut and would therefore eliminate an ambient temperature effect on fuelling rate. Second, if digestive organs were the main contributors to variations in BMR but did not change in size during fuelling, we would expect no or little change in BMR in birds fed ad libitum with trout pellets. We show that cold-acclimated birds maintained higher body mass and food intake (8 and 51%) than warm-acclimated birds. Air temperature had no effect on fuelling rate, timing of fuelling, timing of peak body mass or BMR. During fuelling, average body mass increased by 32% while average BMR increased by 15% at peak of mass and 26% by the end of the experiment. Our results show that the small digestive organs characteristic of a trout pellet diet did not prevent BMR from increasing during premigratory fuelling. Our results are not consistent with the heat load hypothesis as currently formulated.