Species status and new distribution records for Lithurgus huberi Ducke (Hymenoptera, Megachilidae, Lithurginae)

Notes are provided on the morphology of males and females of the enigmatic Lithurgus huberi Ducke (Lithurginae: Lithurgini), a species historically believed to have been introduced into South America from Asia and to be a possible synonym of the more widespread L. atratus Smith. Distinctive differences are documented between L. huberi and L. atratus, perhaps indicative of separate species. In addition, we provide new records of L. huberi in Argentina and Paraguay.


Introduction
Megachilid bees of the subfamily Lithurginae Newman are commonly known as wood-boring bees because they usually excavate burrows in dead, dry, often decayed wood (Michener 2007).The subfamily is unquestionable monophyletic and the sis-ter group of Megachilinae (Roig-Alsina and Michener 1993, Engel 2001, Michener 2007, Gonzalez et al. 2012).Lithurginae consists of two tribes, Protolithurgini Engel, an extinct lineage preserved in mid-Eocene Baltic amber, and Lithurgini Newman, an extant taxon containing about 60 species in three genera: Lithurgus Berthold (currently with two subgenera), Microthurge Michener, and Trichothurgus Moure.Although the subfamily is relatively small in number of species, it is found on all continents except Antarctica.Only species of Lithurgus s.str.are found in the Eastern Hemisphere except for Lithurgus huberi Ducke, which occurs in Brazil andArgentina (Roig-Alsina 2006, Moure andMelo 2007).Snelling (1983) suggested that L. huberi was adventive to South America and probably a synonym of the Indo-Australian species L. atratus Smith given the morphological similarity between the two.Such a hypothesis is also supported by the wood-nesting habits of Lithurgus, which facilitates dispersion across great distances.
We confirmed the close morphological similarity between L. atratus and L. huberi, as noted by Snelling (1983).However, we also noted some morphological features in both sexes of L. huberi that seem to be consistent and that may prove to be useful in species recognition.Accordingly, the purpose of this note is to document and illustrate those characters, as well as record L. huberi for the first time for Paraguay and supplementing this with new distribution records for Argentina (vide infra).

Materials and methods
We examined (V.H.G. & M.S.E) the type series of L. atratus as well as of L. dentipes Smith, a species that has been considered a synonym of L. atratus, deposited in The Natural History Museum, London (NHML).Additionally, we examined 42 specimens of L. atratus from India, Celebes, Borneo, Solomon Islands, and Australia deposited in the Snow Entomological Collection (SEMC), Division of Entomology, University of Kansas Natural History Museum, Lawrence, Kansas, USA, and the U.S. National Pollinating Insects Collection, Bee Biology and Systematics Laboratory (BBSL), Utah State University, Logan, Utah.
Photomicrographs were prepared using a Canon 7D digital camera attached to an Infinity K-2 long-distance microscope lens, and were assembled with the Com-bineZM TM software package.Specimens are deposited in SEMC and the entomological collection of the Museo de La Plata, Argentina (MLP).Occurrence data for L. huberi were plotted using CorelDRAW ® X5 and were taken from the literature (i.e., Ducke 1907, 1908, 1910, Camillo et al. 1983, 1994, Wittmann and Hoffmann 1990, Roig-Alsina 2006, Pick and Schlindwein 2011) as well as from specimen data retrieved from the Inter-American Biodiversity Information Network (IABIN) provided by the USDA, Agricultural Research Service Pollinating Insect Research Unit, Logan, USA, and the Coleção de Entomologia do Laboratório de Biologia Vegetal, Universidade Federal de Pernambuco, Recife, Brazil.

Results
Females of L. huberi primarily differ from those of L. atratus by the facial prominence.In L. huberi it is more strongly punctate and more depressed along the epistomal sulcus than in L. atratus (cf.Figs 1, 3 vs. 2, 4).Likewise, males of L. huberi differ from those of L. atratus in the inner glabrous surface of the metabasitarsus, which is broader and more projected (cf.Figs 5 vs. 6).Although subtle, such morphological differences appear to be consistent across the specimens examined.However, we have had access to a limited number of individuals from a restricted set of localities and cannot rule out that such differences represent mere geographic variations.
Despite the small number of species in Lithurgus s.str., the status of most of them remains questionable, particularly those from Southeast Asia.For example, at least eight 'species' that are closely related to L. atratus have been suggested to represent a single taxonomic unit (Michener 1965), although they could be a complex of rather cryptic species.Some of these are practically indistinguishable from L. atratus, differing mostly in body size (e.g., L. atratiformis Cockerell).Undoubtedly, a revision of the group that includes a great number of specimens from multiple locations (thereby necessitating extensive fieldwork as existing collections have significant gaps for Lithurgus across its distribution) is needed before any taxonomic action can be taken with certainty.Such a work would ideally be accompanied by molecular data to further test putative species boundaries as well as to elucidate possible routes and times of dispersion.
In conclusion, we support the view that L. huberi is closely related to the Indo-Australian L. atratus species or species group and, if dispersed by human activity, it might have reached South America at least 100 years ago when it was described (Ducke 1907).However, we show here consistent, albeit somewhat subtle, morphological features to distinguish L. huberi from L. atratus.For the time being we recommend that they should be treated as separate species.

Figure 7 .
Figure 7. Distribution records of Lithurgus huberi Ducke, including the new records discussed herein.