Two major challenges for pollination ecologists are: (1) assessing the biodiversity of flower visitors in rainforest canopies, and (2) identifying the factors which organise pollinator community structure. More than 30 methods to survey pollinators are currently available (see Potts et al. 2005) but few are suitable for use in canopies. Some rainforest studies either ignore flower visitors in the canopy due to logistical problems, or have limited access (e.g. rope climbing) or rely extensive equipment (e.g. canopy walkways). These methods can sample relatively few points in the canopy or are extremely expensive. We therefore developed a novel method of canopy access which allows rapid and reliable surveys of pollinator communities to be conducted in a variety of forest systems.Between February 2005 and January 2006 we monitored flower visitors on a monthly basis in a variety of Ghanaian forest systems (undisturbed primary forest, regenerating secondary forest and agricultural matrix) using ‘aerial pantraps’. Water-filled pantraps sprayed with blue, white or yellow UV-bright paint were placed in tree canopies using a dynamic rope system which allowed quick setting and collecting of traps (Figure 1). Pantraps have been shown to be highly effective for collecting a range of pollinating insects (Potts et al. 2005).
We collected more than 6,000 potential pollinators in 18 sites in the three forest systems. Bees, including stingless bees, wasps and beetles were all more abundant in the matrix than the primary and regenerating forest (Figure 2). Bee diversity was also greatest in the matrix than the primary forest which was in turn more diverse than the regenerating forest. All taxa showed distinct colour preferences for the pantraps.
Many potential drivers of flower-visiting community structure have previously been identified, including floral community composition and nest-site availability. Recently, the variety of nectar resources available has been shown to structure bee communities in Mediterranean forests and scrub systems (Potts et al. 2004). In the Ghanaian forest system we sampled nectar volume and concentration in flowering plants in parallel to the pollinator surveys. However, we found no correlation between pollinator diversity and variety of nectar resources; suggesting that either nectar is unimportant in structuring communities in our system, or our design was unable to detect it.
We recommend the use of ‘aerial pantraps’ as a reliable, rapid and cost effective method for sampling flower visitors in tropical ecosystems.
Potts S.G., Kevan P.G. & Boone J.W. (2005) Conservation in Pollination: collecting, surveying and monitoring. Practical Pollination Biology (eds A. Dafni, P. Kevan & C. Husband), pp. 401-434. Enviroquest, Cambridge, Ontario, Canada.
Potts S.G., Vulliamy B., Roberts S., O’Toole C., Dafni A., Ne’eman G. & Willmer P.G. (2004) Nectar resource diversity organises flower-visitor community structure. Entomologia Experimentalis et Applicata 113: 103-107.
Sponsor
British Ecological Society SEPG 2128
