Changes in pollinator community of Scadoxus cinnabarinus (Amaryllidaceae) along its elevational range on Mount Cameroon

Abstract

The pollination syndrome hypothesis has been under debate over the past decades. It is criticized as an oversimplification of complex interactions: its validity varies across plant families and depends on spatio-temporal variability of plant–pollinator interactions, yet exact patterns are unclear. We tested the pollination syndrome predictability in a case study of a psychophilous Afrotropical herb species, Scadoxus cinnabarinus, along its complete elevational range (300–1200 m a.s.l.). Through 24-h camera recordings (totalling 500 observed hours), we observed pollinator communities of 21 plant specimens at three elevations on Mount Cameroon, West Africa. Pollination success was quantified through germinated pollen tubes in pistils of selected plants. As predicted by the syndrome, butterflies were the most frequent visitors. Unlike other visiting functional groups, they significantly contributed to the plant’s pollination. We found that the plant’s visitor composition had changed along its elevational range with proportionally more non-pollinating taxa visiting the flowers at both high and low elevations. Additionally, bees at the high elevation probably served as less-effective secondary pollinators, supplementing the lower abundance of butterfly visitors. Even though the psychophilous pollination syndrome correctly predicted the plant’s primary pollinators, its predictability was reduced close to the plant’s elevational limits. These geographic dynamics of specific plants’ pollination systems should be considered in any meta-analyses aiming to test the predictability of pollination syndromes. Especially because the data on pollination systems often originate from a single population.