Influence of functional traits on foraging behaviour and pollination efficiency of wild social and solitary bees visiting coffee (Coffea canephora) flowers in Uganda

An on-farm pollination experiment was conducted during the June–August and November–February blooming seasons of 2007 to 2008, in 30 small-scale coffee fields characterised by different habitat and vegetation types. The study was conducted in order to determine the best pollinator groups for coffee in Uganda and to collect relevant field information and determine the pollination efficiency of different bee species. Results indicate that across blooming seasons, coffee flowers were visited by 24–36 bee species. Hypotrigona gribodoi was the most frequent flower visitor, comprising over 60% of 5941 bee-visits recorded. Foraging rate and pollination speed varied among bee species. Solitary bees foraged on more flowers than social bees, but they spent less time per flower visited. Solitary bees visited more coffee trees and fields, but deposited less pollen, whereas social bees visited less trees and coffee fields in the landscape, but deposited more pollen on flowers. Fruit set was of 87%, 64% and 0.9%, respectively, in hand-cross pollination, open pollination and controlled-pollination treatments. Fruit abortion due to self-pollination was insignificant in this study. There was variability in pollination efficiency of different bee species. Pollination efficiency varied more significantly with sociality than with other bee functional traits and was not significantly influenced by tongue length and bee body size. Single-flower visits by social and solitary bees resulted in 89.7% and 68.14% fruit set, respectively. The most efficient bee species was Meliponula ferruginea (98.3%) followed by Meliponula nebulata (97.1%). Thus, very good pollinator species were wild social bees (mainly stingless bees) as opposed to honeybees and solitary bees that were previously reported to be the best pollinators of coffee in Panama and Indonesia. Morphological and anatomical characteristics of the bee pollen storage features may explain the difference in foraging behaviour activities and in pollination efficiency of social and solitary afrotropical bee species visiting lowland coffee in Uganda. In addition, pollination efficiency was influenced by land-use intensity, field management systems and habitat types found in the immediate surroundings of coffee fields, but not by coffee field size, coffee genotypes and mass blooming wild vegetation. It is recommended to farmers to adopt pollinator-friendly conservation and farming practices such as keeping an uncultivated portion (25%–30%) of their farms as pollinator reservoirs, protecting semi-natural habitats found in the vicinity of coffee fields, as well as promoting high on-farm tree cover to benefit a functionally diverse pollinator community.