No abstract

Plants pollinated by hummingbirds or bats produce dilute nectars even though these animals prefer more concentrated sugar solutions. This mismatch is an unsolved evolutionary paradox. Here we show that lower quality, or more dilute, nectars evolve when the strength of preferring larger quantities or higher qualities of nectar diminishes as magnitudes of the physical stimuli increase. In a virtual evolution experiment conducted in the tropical rainforest, bats visited computer-automated flowers with simulated genomes that evolved relatively dilute nectars. Simulations replicated this evolution only when value functions, which relate the physical stimuli to subjective sensations, were nonlinear. Selection also depended on the supply/demand ratio; bats selected for more dilute nectar when competition for food was higher. We predict such a pattern to generally occur when decision-makers consider multiple value dimensions simultaneously, and increases of psychological value are not fully proportional to increases in physical magnitude.W hen presented with a choice, hummingbirds and nectar-feeding bats prefer concentrated solutions of up to 60% sugar (1-3). Plants that are specialized for vertebrate pollination, however, generally offer relatively dilute nectars of 18 to 23% sugar (4, 5). Nectar value depends on both quality and quantity. Quality is primarily determined by sugar concentration, which is genetically controlled and relatively consistent for individual flowers (6-9). However, multiple foragers normally visit the same flowers, which causes nectar quantity to be highly variable and dependent on the elapsed time since the previous visit. Consumer behavior thus determines availability, introducing a complex dynamic into the decision-making process. To study the factors contributing to the evolution of dilute nectars, we incorporated consumer behavior into a virtual evolution experiment by having free-range bats visit artificial flowers.Field experiments were conducted with freerange adult Glossophaga commissarisi bats that had been outfitted with radio-frequency ID tags in the rainforest at La Selva Biological Station, Costa Rica. We presented bats with a computercontrolled array of 23 artificial flowers (Fig. 1A). Each flower was equipped with an ID sensor and a photogate (a device that recorded the duration of a bat visit) and connected to a nectarpump system capable of delivering nectar of defined sugar concentration and volume (10). The density of the array approximated the distribution of a local bat-pollinated bromeliad (Werauhia gladioliflora) that provides nectar with sugar concentrations between 14 and 16% (11). Visiting bats were rewarded with nectar, and the amount of nectar was determined by the secretion rate and the elapsed time since the previous bat visit.We assumed individual plants in our population to have equal rates of photosynthesis and invest equal amounts of photosynthate, as sugar, into nectar (9). However, the sugar concentration of the nectar was determined by a flower's virtual gen...

No abstract