REVISION OF THE CARPENTER-BEES (XYLOCOPA LATREILLE) OF THE MOLUCCAN ISLANDS, WITH NOTES ON OTHER INDO -AUSTRALIAN SPECIES BY M. A. LIEFTINCK Rijksmuseum van Natuurlijke Historie, Leiden The present paper is a continuation of my recent survey of the Carpenter-bees of the Lesser Sunda Islands (1955), in which, however, no species were discussed that occur also in the islands of the Moluccas, between Celebes and New Guinea. The Moluccas, as here understood, comprise the Sula Islands (Taliabu, Mangole and Sanana), the northern Moluccas (Morotai, the Halmahera group and Obi), and the southern Moluccas (Bum, Ambon with adjacent small islands, and Ceram). The Banda, Kei and western Papuan islands are left out of consideration here as the species occurring there differ markedly from those of the Moluccas proper. There is a very noticeable decrease in the number of species of Xylocopa east-wards of the Sunda Islands and Celebes. For instance, according to van der Vecht's recent synopsis (1953), the island Celebes still has one representative each of the subgenera Biluna and Zonohirsuta, and four of Maiella^), not count-ing a number of subspecies of some of the last species-group. Remarkably enough, these are absent from the Moluccan islands, with the exception of one, viz. X. (Maiella) nobtlis tricolor Ritsema, the only carpenter-bee known from the Sula Islands. This, also, is the only connecting-link between the faunae of Celebes and the Moluccas, differing scarcely from nobtlis F. Smith, which itself occurs on Celebes in a number of races. The remaining four (of which possibly only three deserve full specific rank) are likewise members of the subgenus Maiella and form a compact group of rather uniformly coloured species which appear much more closely related inter se than with either the Malaysian group of confusa J. Perez, or the Papuan cluster of provida F. Smith. Although perhaps one of them may prove to have penetrated eastwards some way into the western part of New Guinea, all species or subspecies are peculiar to one island, or group of islands, and there is no definite proof of two forms occurring together on one of them. Like most other members of Maiella. 1 ) The validity of the subgeneric name Maiella Michener 1942 (= Orbitella Maa 1938, olim: nom. praeocc.) has recently been called in question by Maa (Vidensk. Medd. Dansk naturh. Foren., 1954, vol. 116, p. 192 footnote), who revalidated the name Koptortosoma Gribodo 1894 (typ. subgen. gabonica Gribodo), or its emendation Coptortbosoma J. Perez, as a probable substitute for Maiella. However, since gabonica Grib. does not appear to have been re-characterized in recent times, I prefer for the present to retain the name Maiella to denote a large group of Old World species of Xylocopa. 55
BioStor
