Apidologie 40 (2009) 347–354 Available online at: c INRA/DIB-AGIB/EDP Sciences, 2009 www.apidologie.org DOI: 10.1051/apido/2009018 Review article Diversity and conservation status of native Australian bees* Michael Batley 1,KatjaHogendoorn2 1 Australian Museum, 6 College St, Sydney, N.S.W., 2010, Australia and Department of Biological Sciences, Division of Environmental and Life Sciences, Macquarie University, North Ryde, N.S.W. 2109, Australia 2 School of Agriculture, Food and Wine, University of Adelaide, SA 5005, Australia Received 8 August 2008 – Revised 14 October 2008 – Accepted 26 October 2008 Abstract – Australia’s idiosyncratic bee fauna is characterised by a combination of numerous endemic taxa and by the complete absence of some families. Many species, and in particular several oligolectic species, remain undescribed and more than half the named taxa are in need of revision. The main threats to the native bee fauna include removal of nesting and foraging opportunities through land clearing and agriculture, the spread of exotic plant species and the consequences of climate change. Early steps to conserve the native bee fauna include commercial applications, the raising of public awareness and preservation of natural habitat. However, these actions are severely hampered by a lack of both identification keys and taxonomic expertise. Considerable investment in taxonomic research is needed to improve this state of affairs. native bees / Australia / conservation / taxonomy / Apoidea 1. DIVERSITY OF BEES 100 bee species are either known or suspected IN AUSTRALIA to be oligolectic (AFD, 2006). Michener (1965) noted a clear difference Australia has a distinctive bee fauna char- between the “typically Australian element” acterized by an unusually high proportion found throughout most of the continent and of species in the family Colletidae (Tab. I). the depauperate bee fauna of the northern rain- Several groups with widespread distributions forests that shares features with that of New throughout the rest of the World, notably An- Guinea and the Indo-Malaysian region. At drenidae and Mellitidae, are not found in Aus- the species level, there is a less distinct divi- tralia while the family Stenotritidae, and the sion between the fauna of the open eucalypt subfamily Euryglossinae are endemic to the forests and woodland of eastern and south- region (Michener, 1965, 2000). More than eastern coastal regions and that of the arid three-quarters of the known species belong zone which covers most of the remaining area. to the three most highly derived bee fami- There are no sharp lines of demarcation and lies, Stenotritidae, Colletidae and Halictidae species belonging to groups with significant (Danforth et al., 2006) and a small number of diversity in New Guinea have been found in colonisation events have given rise to lineages, rainforest remnants as far south as Sydney. like the genus Exoneura (Schwarz et al., 2006) or parts of the genus Lasioglossum (Danforth While the taxonomy of Australian bees and Ji, 2001), that are endemic to Australia has received some attention in the last 20 and adjacent areas. Despite the incomplete na- years, there is still a large number of unde- ture of flower-visiting records, approximately scribed species. The Australian Faunal Direc- tory (AFD, 2006) lists 1647 names, of which Corresponding author: K. Hogendoorn, about 45% come from revisions published [email protected] after 1965, 18% have been treated in doc- * Manuscript editor: Mark Brown toral theses that remain unpublished, and the Article published by EDP Sciences 348 M. Batley, K. Hogendoorn Table I. The number of Australian genera and named species (AFD, 2006) covered by keys. References to the keys are given in Michener (2000). family # genera # genera with keys # named # species covered % of species covered to at least part species by keys by keys Apidae 13 4 197 49 25% Colletidae 31 20 879 434 49% Halictidae 11 3 385 188 49% Megachilidae 5 1 168 25 15% Stenotritidae 2 1 21 10 48% remaining 37% are greatly in need of revi- affect bee populations, with demonstrable eco- sion. Genera such as Exoneura, Lipotriches, nomic consequences (Kremen et al., 2004). Stenotritus, Paracolletes, Trichocolletes and Similar studies have shown that the presence much of Euhesma have not been studied in re- of native Australian rainforest is important cent times. for the unmanaged pollination of tropical fruit Extrapolation of the results of four recent (Blanche et al., 2006). investigations, including unpublished theses, Conservation of native vegetation in Aus- covering almost 30% of the fauna suggests tralia has received considerable public and that between 300 and 400 species remain un- regulatory attention over the past 20 years, pri- described, but that the total number of valid marily with the objectives of controlling sec- names will remain virtually unchanged when ondary dryland salinity, maintaining biodiver- synonyms are removed. However, not only sity and reducing greenhouse gas emissions. are there large uncertainties in such extrapo- A recent review of progress towards national lations, but there is probably a bias towards objectives for biodiversity conservation (Grif- polylectic species. Oligolectic species are of- fin NRM Pty Ltd., 2004) concluded that, de- ten overlooked until the nature of their special- spite patchy success in many areas, clearing of isation is discovered (Houston, 1989, 1992, native vegetation will be effectively controlled 1993;Exley1998, 2004; reviewed in Murray in most jurisdictions by 2010, provided that et al., 2009). currently proposed and enacted regulations are A measure of the extent of missing informa- properly implemented. Australia is still some tion is given by the number of subgenera for distance from achieving a no net loss target which there are no published keys to species for native vegetation, but the recent policy and (Tab. I). Overall, 57% of all named species are regulatory interventions should ensure that all not covered by keys. future clearing proposals are subject to appro- Knowledge of the taxonomy of Australian priate levels of assessment relative to the con- bees is much more extensive than information servation status of the vegetation community. about their biology. Many species are known from very few specimens and distributions and Some areas have been less affected by clear- flowering-visiting records for regions outside ing than others so that only 13% of the to- Western Australia are limited (Houston, 2000: tal land area has been cleared. The largest AFD, 2006). proportion of Australia, the rangelands of the arid zone, remains relatively free of broad- scale clearing. The eastern and south-eastern 2. THREATS TO BEES coastal areas were predominantly eucalyptus IN AUSTRALIA forests and woodlands. Approximately 80% of the pre-European extent of eucalypt forests 2.1. Deforestation and agriculture and woodlands remain, despite high rates of clearing in Queensland and New South Wales The loss of floral resources and nesting sites that were a cause for concern. Between 1981 associated with deforestation will obviously and 1999 the average rate of clearing was Bees in Australia 349 38000 ha/year in NSW and 343 000 ha/year in nectar-producing flowers, making pesticides Queensland, though some of this was removal of secondary importance. of regrowth. The most severely affected vege- tation groups are the low closed forests, closed shrublands and the heaths, of which only 50% 2.2. Exotic species now remains, but these covered only 1% of the continental area. Australia is the only country yet to be in- However, vegetation removal is not the only vaded by Varroa destructor and has a large form of ecological disturbance and an area population of feral Apis mellifera. The num- may retain the same vegetation type and yet bers of feral honeybees are such that they may have a different species composition. For ex- be affecting the populations of birds, mammals ample, it has been estimated that of the two and other insects by virtue of the sheer amount million ha of lowland grassland in south- of floral resources that they consume (Paton, eastern Australia at the time of European set- 1996; Stout and Morales, 2009). Paini (2004) tlement, the remaining natural remnants cover has reviewed the 10 studies that addressed var- 10000 ha. (McDougall and Kirkpatrick, 1994). ious aspects of the impact of European honey- Furthermore, in the arid centre of Australia, bees on native bees in Australia and concluded substantial changes in plant composition have that many of the studies suffered from low been attributed to the impacts of water-centred replication or confounding factors. It is likely grazing. The effect of such changes on bee that the most important impact of honeybees populations could be considerable. For exam- on native bees is indirect, through their role in ple in the southern arid regions of Australia, the propagation of serious exotic weeds such the lack of regeneration of Western Myall as Scotch Broom (Cytisus scoparius, Simpson due to rabbit, goat and sheep grazing (Lange et al., 2005), and lantana (Lantana camara, and Purdie, 1976), may seriously affect the Goulson and Derwent, 2004) at the cost of survival of the endemic allodapine bee Ex- native vegetation (Stout and Morales, 2009). oneurella tridentata, because this species de- The imminent arrival of the Varroa mite in pends largely on the dead branches of this Australia, which is expected to largely wipe tree for nesting substrate (Hurst,