The canopies of tropical rain forests support highly diverse, yet poorly known, animal and plant communities. It is vital that researchers who invest the time needed to gain access to the high canopy are able efficiently to survey the animals and plants that they find there. Here, we develop diversity assessment protocols for one of the most ecologically important canopy animal groups, the ants, in lowland dipterocarp rain forest in Sabah, Malaysia. We design and test a novel trap (the purse-string trap) that can be remotely collected, thus avoiding disturbance to ants. We compare this modified trap with two other methods for surveying canopy ants: precision insecticide fogging and baited pitfall trapping. In total, we collected 39,351 ants belonging to 173 species in 38 genera. Fogging collected the most individuals and species, followed by purse-string trapping with baited pitfall trapping catching the fewest. Fogging also resulted in samples with a different species composition to purse-string trapping and baited pitfall trapping, which were not different from one another. Using a ‘greedy algorithm’, which guides the selection of inventory methods in order to maximize new species discovered per researcher-hour, we show that projects allocating fewer than 132 researcher-hours to canopy ant collection and identification should sample exclusively using fogging. Those with more time should use a combination of methods. This prioritization technique could be used to accelerate species discovery in future rapid biodiversity assessments.
Abstract in Malay is available in the online version of this article.