No abstract

Flower visitation by small mammals is difficult to observe directly on account of their sensitivity to human presence and the nocturnal activity of some species. Their role in pollination has, thus, been inferred mostly from analysis of fur and faecal pollen loads of field-trapped animals. The development of motion-activated remote cameras allows for unprecedented insights into natural behaviours of small mammals on flowers. We used remote cameras to assess the identity of flower visitors, and the frequency and timing of their visitation to flower heads of four poorly studied Protea species (P. cordata, P. decurrens P. scabra and P. subulifolia) that were predicted to be rodent-pollinated on the basis of floral traits. Camera footage revealed that small mammals (typically three to six rodent species per Protea species) were frequent visitors to flower heads over one flowering season, making up 100% of visits to P. cordata and P. subulifolia, 98% of visits to P. decurrens, and 90% of visits to P. scabra. Rodents that are known pollinators of other ground Protea species, such as Acomys subspinosus, Micaelamys namaquensis and Rhabdomys pumulio, were observed to make contact with reproductive parts of the flowers, but further work would be required to demonstrate unequivocally that they are the main pollen vectors of these species. Mus minutoides was recorded for the first time to frequently visit inflorescences of P. subulifolia. Other visitors included genets that visited P. scabra inflorescences to lick nectar off florets, and birds that were relatively infrequent visitors to all species. Together with recent literature, the present study has shown that the utilisation of remote cameras in pollination studies provides an efficient, discreet method for observing vertebrate flower-visiting behaviour.

No abstract

No abstract