Fig 1.
(A). Side view of a flower, showing the four parameters in the equation. All parameters (except for c, which is dimensionless) are in millimeters. (B). 3D rendering of some representative flowers with variation in nectary diameter and corolla curvature. The corolla tube length L is fixed at 20 mm and the overall diameter 2(R + r0) is fixed at 55 mm.
Fig 2.
Second stage experimental setup.
(A). Hawkmoth flight chamber. Only the center flower is instrumented and supplied with nectar. The two flowers on the sides are distractors. (B). Instrumented flower model, which is an enlarged view of the center flower in (A).
Fig 3.
Hawkmoth foraging success rate and flower visitation frequency in the first stage experiment.
(A). Hawkmoth foraging success rate across two dimensions of flower morphospace. Panels B and C show some flowers that capture the range of variation along the two morphological axes from the full dataset (S1 Table). (B). Hawkmoth foraging success rate. (C) Flower visitation frequency. Error bars represent ± 1 SEM. The sample size of each flower morphology is in S1 Table.
Fig 4.
Fitness of pollinator and plant measured by visit quality.
(A). Hawkmoth pollinator’s fitness; (B). Plant’s fitness. Error bars are ±1 SEM. Sample size (number of visits) of each flower morph is: c = -∞, N = 198; c = -3, N = 428; c = -1, N = 384; c = 1, N = 806.