Appendix A. Methodology of flower-visiting insect observations, pollination experiments, and functional group classification.
The coffee flower-visiting bees were observed and number of species and individuals were counted on three different days for 25 minute periods in 15 (C. canephora) and 24 (C. arabica) coffee agroforests in Central Sulawesi, Indonesia (for more details see Klein et al. 2003a,b). As we do not have replicated observation data on the spatio-temporal pollinator behaviour in coffee, we used a priori classification with two morphological traits, (i) body size and (ii) tongue lengths, and one behavioural trait (sociality).We measured the tongue lengths and the intertegular span of the pronotum of a female and a male individual of each species to define body size classes. We divided species in three tongue length classes: short tongue (<3 mm), medium tongue (3–8 mm), and long tongue (>8 mm), and four body size classes: very small (>2 mm), small (2–3 mm), medium sized (3–5 mm), and large (>5 mm). For sociality, we used two classes: social (colony building) and solitary bees (one to few females per nest). Flower-visiting nitiduliid beetles in atemoya were separated in two body size classes (body length), small beetles (<2.1 mm), large beetles (>2.1 mm) (see also Blanche and Cunningham 2005).
We define pollination service as the service provided by bees, calculated as the percentage initial fruit set resulted from open pollination with insects had access to flowers minus bagged pollination with only passive self-pollination lead to fruit development of the same plant individual. Therefore, we excluded the effects of passive self-pollination. Pollination experiments were conducted using four replicated coffee plants in each of the 15 and 24 agroforests for C. canephora and C. arabica, respectively (methods described in Klein et al. 2003a,b,c).
LITERATURE CITED
Blanche K. R., and S. A. Cunningham. 2005. Rainforest provides pollinating beetles for atemoya crops. Journal of Economic Entomology 98:1193–1201.
Klein, A. M., I. Steffan-Dewenter, and T. Tscharntke. 2003a. Fruit set of highland coffee increases with the diversity of pollinating bees. Proceedings of the Royal Society of London, B. 270:955–961.
Klein, A. M., I. Steffan-Dewenter, and T. Tscharntke. 2003b. Pollination of Coffea canephora in relation to local and regional agroforestry management. Journal of Applied Ecology 40:837–845.
Klein, A. M., I. Steffan-Dewenter, and T. Tscharntke. 2003c. Bee pollination and fruit set of Coffea arabica and C. canephora (Rubiaceae). American Journal of Botany 90:153–157.