Failure to adapt to a changing nutritional environment comes at a cost, as evidenced by the modern human obesity crisis. Consumption of energy-rich diets can lead to obesity and is associated with deleterious consequences not only in humans but also in many other animals, including insects. The question thus arises whether animals restricted over multiple generations to high-energy diets can evolve mechanisms to limit the deposition of adverse levels of body fat. We show that Plutella xylostella caterpillars reared for multiple generations on carbohydrate-rich foods (either a chemically defined artificial diet or a high-starch Arabidopsis mutant) progressively developed the ability to eat excess carbohydrate without laying it down as fat, providing strong evidence that excess fat storage has a fitness cost. In contrast, caterpillars reared in carbohydrate-scarce environments (a chemically defined artificial diet or a low-starch Arabidopsis mutant) had a greater propensity to store ingested carbohydrate as fat. Additionally, insects reared on the low-starch Arabidopsis mutant evolved a preference for laying their eggs on this plant, whereas those selected on the high-starch Arabidopsis mutant showed no preference. Our results provide an experimental example of metabolic adaptation in the face of changes in the nutritional environment and suggest that changes in plant macronutrient profiles may promote hostassociated population divergence.geometric framework ͉ nutrition ͉ Plutella

We used a strain of diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae), that had been reared for approximately 350 generations in a precisely characterised environment to test hypotheses regarding the influence of nutritional heterogeneity on the evolution of nutrient regulatory responses. Caterpillars were maintained with ad libitum access to a diet that emulated that of an extreme nutritional specialist, comprising a homogeneous food of fixed nutrient composition. We measured performance (survival, development rate, and pupal mass), as well as the protein and carbohydrate intake of individual caterpillars confined to one of a range of single foods differing in their protein, carbohydrate, and water content. In a separate experiment, we measured the amount and balance of protein and carbohydrate self‐selected by caterpillars presented with nutritionally complementary foods. Results showed a close fit with three of four predictions about the nutritional responses of ‘nutrient specialist’ feeders: (1) survival, development rate, and pupal mass were highest for animals given diets with the protein:carbohydrate composition of the ancestral culture diet, and dropped off sharply with higher and lower protein:carbohydrate balance, (2) caterpillars coped poorly with dietary dilution by water, irrespective of the macronutrient balance, and (3) the self‐selected intake point corresponded with the macronutrient balance that gave peak performance (i.e., that of the ancestral culture diet). The fourth prediction, that caterpillars would be disinclined to over‐ingest nutrients on imbalanced diets, was at best weakly met. We hypothesise that the evolution and maintenance of the specialist strategy might, paradoxically, require some degree of environmental heterogeneity.