This subfamily is by far the largest and most abundant group ofhairy colletids. It consists of small to rather large, generally hairy andreniform bees, most of them superficially resembling species of the genera Andrena or Halictus. A few small species are only sparsely hairy and are almost hylaeiform, especially in males. The first flagellar segment is shorter than the scape and not recognizably petiolate. The glossa usually has two short lobes (it is thus weakly bilobed, Fig. 19-2b, as in Fig. 37-1) but is sometimes deeply bifid (Fig. 39-12); the preapical fringe (at least in the female) and the glossal lobes (or apical glossal zone bearing the glossal brush) are well developed. The prementum lacks a spiculate depression on the posterior (lower) surface, or has a longitudinal median groove perhaps homologous to the depression; or, in the African genus Scrapter, the spiculate depression is well developed (Fig. 41-3b). The facial fovea is usually broad and ill-defined or absent, but sometimes is sharply defined, and sometimes (as in Callomelitta, some Eulonchopria, and some Scrapter) forms a groove. The episternal groove extends well below the scrobal suture (but forms only a broad, shallow depression in Hesperocolletes). The scopa on the hind leg of the female is large and dense to sparse, forming a corbicula on the underside of the femur; it is also well developed on the tibia and sometimes on the metasomal sterna and on the side of the propodeum. The scopa is almost absent, however, in Leioproctus (Euryglos-sidia) cyanescens (Cockerell), which seems to transport pollen in the crop instead of on the scopa (Houston, 1981b). There are two or three submarginal cells; if two, then the second is two-thirds the length of the first or longer, as if the second submarginal crossvein has been lost. The pygidial and prepygidial fimbriae of the female are usually present, often dense, but they are absent (margins thus similar to those of preceding terga) in Colletes and Mourecotelles.
The larva was characterized by McGinley (1981, 1987).
The distribution of the subfamily is worldwide except for the arctic and antarctic. Bees of this subfamily are uncommon (or in some areas absent) in moist tropics, and probably completely absent in much of tropical Asia and Indonesia. Except for the genus Colletes, the Colletinae are essentially austral, being most abundant and diversified in temperate areas of Australia and South America. One tribe, the Scraptrini, is found only in southern Africa.
Engel (2005) has discussed the features of Scrapter and concluded that they justify recognition of a subfamily, Scraptrinae. Variation within Scrapter and among the remaining genera here placed in the tribe Paracolletini is so great, however, that the only known constant character separating Scrapter from the other genera is the presence of a broad fovea on the posterior surface of the premen-tum (see Fig. 41-3). This character may be a plesiomor-phy rather than a derived feature characterizing Scrapter. For the present I therefore recognize a tribe Scraptrini within the subfamily Colletinae. Other characters that might support recognition of the Scraptrini are those of larvae (McGinley, 1981); some analyses of these characters show Scrapter as the sister group to all other Colleti-nae. A relationship to the Euryglossinae rather than to other Colletinae is suggested not only by the premental fovea but by the nodulose margins of the basitibial plate of some species.
Differences in nesting biology between the Colletini and Paracolletini seem substantial. In the latter group, Lonchopria (Michener and Lange, 1957 and contained references) and Leioproctus (Michener and Lange, 1957; Michener, 1960) make burrows from near the ends of which laterals diverge, each lateral usually ending in a single, more or less horizontal cell, but sometimes, in both genera, there may be two or more cells in series. The cells are similar to those of Halictus or Andrena, homomor-phic, bilaterally symmetrical about a vertical plane (because the lower surface is flatter than the upper), and lined with a thin secreted membrane. The larval food is a firm, subspherical pollen mass, the egg laid on top of it as in Halictus and Andrena. In the Colletini, nests of Colletes are well known; see the account of that genus. The burrow structure may be similar to that described above but more usually ends in burrows subdivided into series of cells, not shaped for particular cells. The cells are therefore heteromorphic. The partitions between cells are of the secreted cell-lining material as in Hylaeinae, not of soil. The provisions are semi-liquid, and the egg is attached by one end to the cellophane-like lining of the cell, above the provisions. In the Scraptrini (Rozen and Michener, 1968) the situation is intermediate, in the sense that the cell is merely the distal part of a lateral burrow, neither flattened on the lower side nor enlarged, but nonetheless probably homomorphic. The provisions fill the distal part of the cell, as they do in Colletes, although they are at first firm, only later, with absorption of water, becoming semi-liquid. The egg, however, is laid on top of the provisions, as in Leioproctus.
The phylogenetic study of families of S-T bees by Alexander and Michener (1995) indicated that the Colletinae are paraphyletic. Because of the very different phy-logenetic hypotheses shown by different analyses for the eight exemplars of Colletinae, no acceptable phylogeny was evident, although Colletes and Mourecotelles consistently emerged as sisters. Scrapter sometimes appeared as the basal branch of the hylaeine clade, instead of as a col-letine, in part because of its possession of a depression or fovea on the prementum. The one Leioproctus and one Lonchopria in the study were widely separated in spite of the existence of an intermediate (not in the analyses) between the two. It is clear that the Colletinae includes diverse elements. A needed step is a phylogenetic study of the forms here placed in Colletinae, using representatives of many more taxa than were used in the Alexander and Michener study.
1. Basitibial and pygidial plates absent; prepygidial and py-gidial fimbriae lacking, in both sexes vestiture ofT5 and T6 thus similar to that of preceding terga (Fig. 40-1); S7 of male with apicolateral lobes greatly enlarged, disc of sternum and apodemes reduced, slender and delicate, the lobes thus constituting the major part of S7 (Fig. 40-
—. Basitibial and pygidial plates present, at least in females (pygidial plate absent in most males; basitibial plate absent in both sexes of a few Australian taxa); prepygidial and pygidial fimbriae of female present; S7 of male with apicolateral lobes of moderate size (Figs. 13-2a; 39-1, -3, -7, -9, -10, etc.) (sometimes greatly reduced or absent), disc of sternum and apodemes thus constituting the major part of S7 2
2(1). Posterior (lower) surface of prementum with a broad longitudinal depression or fovea margined by shiny ridges (Fig. 41-3b) that diverge on basal part of prementum and converge near base of subligular process; galeal comb reduced to three or four small bristles (Fig. 41-2b)
(Africa) Scraptrini (Sec. 41)
—. Posterior (lower) surface of prementum lacking a fovea or with only a narrow medial groove; galeal comb usually well developed Paracolletini (Sec. 39)
Was this article helpful?