Ecological factors driving pollination success in an orchid that mimics a range of Fabaceae

Abstract

Rewarding plants can enhance the pollination success of co-occurring plants pollinated by food mimicry. However, it is not always possible to readily discern between the effect of model and magnet species. Here, we tested for mimicry of co-occurring Fabaceae by the rewardless Diuris magnifica (Orchidaceae) and whether the number of flowers of Fabaceae, habitat remnant size and frequency of conspecifics, influenced the pollination success of D. magnifica. Trichocolletes bees were the primary pollinators of D. magnifica, on which they displayed similar behaviour as seen when feeding on Fabaceae. Quantification of spectral reflectance suggested that flowers of Bossiaea eriocarpa, Daviesia divaricata and Jacksonia sternbergiana may represent models for D. magnifica, whereas Hardenbergia comptoniana strongly differed in colour. Orchid pollination success was not directly affected by the number of model flowers, but the pollination rate was enhanced by increased numbers of Hardenbergia flowers. Pollination success of the orchid decreased with higher density of conspecifics, but did not exhibit a significant relationship with Trichocolletes occurrence, possibly because of the contribution of sub-optimal pollinator species. Fruit set of the orchid was greater in larger habitat remnants. Overall, pollination success of D. magnifica is affected by ecological factors related to the effectiveness of mimicry, numbers of co-flowering plants and anthropogenic landscape alteration.