Abstract: Bees are key pollinators in almost all terrestrial ecosystems and can have major roles in agricultural production. Records of bees in the Southwest Pacific indicate a very low diversity, with the Fijian bee fauna one of the least diverse, despite an otherwise rich biota. Megachilid bees represent a large proportion of the bee fauna for almost all island groups in the Southwest Pacific and, because they are wood- and stem-nesting, their wide distribution is likely to have been influenced by rafting and anthropogenic maritime trade. Our study is the first to apply molecular techniques to the study of megachilid bees in this region and indicates between four and five recent introductions to Fiji, likely from Southeast Asia. The study also provides the first record of Heriades (Michenerella) in the Southwest Pacific and the first record of the subgenus Megachile (Callomegachile) in Fiji. These results indicate that a large proportion of the Fijian bee fauna is likely to have been introduced only very recently and, therefore, has had only a very recent role in Fijian ecosystems, despite their current abundance. This has very wide implications for understanding Fijian plant-pollinator relationships. We argue that there is a strong need to understand ancient plant-pollinator relationships that may have evolved in Fiji before the mid-late Pleistocene and Holocene and whether these could be disrupted by recent bee introductions.
BEES COMPOSE one of the most important groups of pollinators (Michener 2007) and have coevolved with angiosperms since the mid-Cretaceous (Crepet and Nixon 1998, Engel 2000). They play a critical role in most terrestrial ecosystems, yet recent studies suggest a variety of threats to both their abundance and diversity (Memmott et al. 2007, Kaiser-Bunbury et al. 2010, Potts et al. 2010). Understanding these threats is important in maintaining pollination networks that have evolved over long periods of evolutionary time. However, it is also possible that some terrestrial ecosystems have evolved in the absence of bees, or where bee abundances and diversities were very different before human activities. Identifying these circumstances is also important for conservation and for understanding the genesis of plant-pollinator ecosystems.
The ecosystems of continental landmasses, such as those associated with tectonic plates originating from early Gondwanan and Laurasian supercontinents, have seen bees and angiosperms share very long evolutionary histories, and we might therefore expect a range of both broad and narrow pollinator suites for different plant groups (Danforth et al. 2006, Thien et al. 2009). However, many island ecosystems are characterized by recent and relatively complex geological histories, where plant-pollinator relationships are likely to have been heavily influenced by patterns of colonization from older continental regions.
Previous studies have suggested that the bee fauna in the Southwest Pacific is depauperate (Perkins and Cheesman 1928, Michener 1965). This is unexpected, considering the otherwise complex geographical and biotic history of the region (Groom and Schwarz 2011). For Fiji, a total of only 14 non-Apis bee species have been listed (Evenhuis 2007) and summarized by Groom and Schwarz (2011), but this figure is more likely...