Info

glossal lobe

-annulate surface disannulate surface ■

basiglossal sclerite basiglossal sensilla m annulate surface annulate surface

basiglossal sclerite basiglossal sensilla m annulate surface annulate surface disannulate surface seriate line-fr^i /

marginal line-¡3

annulate surface -

seriate hair disannulate surface--^H- '

seriate hairs seriate hair seriate line-fr^i /

disannulate surface--^H- '

seriate hairs

Figure 19-4. Diagrams of colletid and halictid glossae, with structures labeled. a, b, Anterior and posterior surfaces of the glossa of a colletid; c, Cross section of same, the anterior surface above; d, e, Anterior and posterior surfaces of the glossa of a halictid; f, Cross section of same, the anterior surface above. From Michener and Brooks, 1984.

of L-T bees with the labial palpi of S-T bees (Fig. 110-1c, d; Michener and Greenberg, 1980), is a member of the LT bee clade, according to Roig-Alsina and Michener (1993); it probably lost the palpal characteristics of that clade.

Conversely, there are many S-T bees with long mouth-parts. In some groups, such as the Halictinae and Xe-romelissinae, the basal parts of the proboscis, i.e., the cardines, stipites, and prementum, are elongated (Fig. 19-2c), unlike those of L-T bees. In others, like some Pa-nurginae, Nomiinae, Halictinae, and Rophitinae, the glossa is elongate (Fig. 19-5c), and in some ways, e.g., the development of a flabellum in some Panurginae (see Michener and Brooks, 1984), resembles that of L-T bees.

As described above, most S-T bees with long mouth-parts have elongations of major structures (cardines, stip-ites, prementum, glossa). A few S-T bees, however, have unusual elongations of other structures that presumably help in imbibing nectar from deep or protected sources. Examples, all in the Colletidae, are (1) maxillary palpi that fit together to form a tube (a species of Euhesma in the Euryglossinae; Houston, 1983c), (2) labial palpi that form a tube (Niltonia in the Colletinae; Laroca, Michener, and Hofmeister, 1989), and (3) a pencil, formed by enormous galeal setae and the labial palpi, that probably draws nectar by capillary action (Leioproctus subgenus Filiglossa, in the Colletinae; Fig. 19-6).

In spite of the problems indicated in the preceding paragraphs, the division of all bees into two units, the ST and L-T bees, is distinct; it is derived rather than ancestral species or genera in each unit that blur the distinction. Phylogenetic studies show that the L-T bees constitute a holophyletic unit. The S-T bees, however, are paraphyletic; they gave rise to the L-T bees and may have no common synapomorphies; their common characters mentioned above are plesiomorphies as judged by comparisons with sphecoid wasps, as are other characters enumerated in the papers cited, with one possible exception. The galeal comb, widespread among S-T bees, is believed to be not homologous to the galeal comb found in most sphecoid wasps because of its position on the galea (see characters 26 and 27, Alexander and Michener, 1995). It could therefore be a synapomorphy for S-T bees, although lost in some of them. More likely it is a synapo-

Figure 19-5. Proboscides of S-T bees. a, b, Inner surface of maxilla and anterior surface of labium (with enlargements of apical hairs of glossa and paraglossa) of Macropis europaea Warncke; c, d, Labium (with truncate paraglossa at left and labial palpus at right of glossa) and maxilla (base of cardo omitted) of Neffapis longilingua Ruz. Neffapis is a long-tongued S-T bee. From Michener, 1981a, and Rozen and Ruz, 1995.

Figure 19-5. Proboscides of S-T bees. a, b, Inner surface of maxilla and anterior surface of labium (with enlargements of apical hairs of glossa and paraglossa) of Macropis europaea Warncke; c, d, Labium (with truncate paraglossa at left and labial palpus at right of glossa) and maxilla (base of cardo omitted) of Neffapis longilingua Ruz. Neffapis is a long-tongued S-T bee. From Michener, 1981a, and Rozen and Ruz, 1995.

morphy for bees as a whole, lost in some S-T bees and nearly all L-T bees, although indicated even in some L-T bees such as Xeromelecta (Melectini).

A few other character states of adults exhibit a similar distribution and, although variable, appear to be generally derived among S-T bees, while ancestral in L-T bees. Examples are the elevated basal area of the labrum; the short basistipital process; the large, diverging seriate hairs of the glossa (Fig. 19-4e), possibly lost in female and most male colletids; and the midtibial and midfemoral brushes or combs. These are all found in S-T bees and are seemingly apomorphic compared to sphecoid wasps and to the L-T families of bees, which lack these features.

A viewpoint that I have abandoned but which nonetheless requires discussion is that L-T bees were derived from the Panurginae or their antecedents. This possibility is suggested by certain panurgine characters similar to those

Figure 19-6. Proboscis of Leioproctus (Filiglossa) filamentosus (Rayment). a, Labium, with mentum, lorum, and most of one palpus omitted; enlargement of apex of palpus below. b, Outer view of maxilla, with cardo omitted. (Labial palpal segments are numbered; an arrow indicates the unusually large lacinia.) Presumably, the pencil of filaments consisting of palpi and galeal setae draws nectar by capillarity.

of L-T bees. As indicated above, in certain panurgines the first segment of the labial palpus is long and (rarely) the first two segments are elongate, suggestive of L-T bees. Moreover, in those Panurginae with an elongate glossa the disannulate surface is usually somewhat invaginated (strongly so in Melitturga; see Michener and Brooks, 1984) and the apex has a well-formed flabellum.

This flabellum, especially in some species of Perdita (Fig. 28-2h, i) is incredibly similar to that of many L-T bees, the resemblance including not only its shape but its anterior curvature, the row of setae across its anterior surface, the hairs elsewhere on it, etc. (Michener and Brooks, 1984). It seems most unlikely that such an elaborate structure could have evolved independently in these pa-nurgines and in L-T bees. In many doubtful cases elsewhere it seems hopeless to suppose that we will ever de cide which similar characters of two organisms are homologous and which result from convergence. In this case, however, the answer is clear. Among S-T bees only certain panurgines have such characters, which are clearly apomorphic, suggesting L-T bees. If L-T bees arose from panurgines, then the L-T bees are a sister group to the rather long-tongued panurgines— Calliopsis, Melitturga, Perdita, etc. Yet panurgines have striking synapomor-phies not shared by L-T bees. For example, they lack or nearly lack the male gonobase and they have two suban-tennal sutures under each antenna. To regain the gonobase after losing it seems most improbable. Furthermore, the relationship of L-T bees to the short-tongued Melittidae seems well established. The common characters of panurgines and L-T bees, then, appear to be a remarkable case of convergence (see Secs. 27, 28).

Michener's (1944: 231) reasons for deriving the melit-tid/L-T bees from the Panurginae or their immediate ancestors are also not tenable. The tapering mentum characteristic of L-T bees and panurgine bees, on which I based this conclusion, is also found in some Colletinae and even in some Andreninae (Michener, 1985a) and therefore does not point to a panurgine origin.

In a study of bee larvae, Michener (1953a) observed that many characters seemed to be derived in the S-T bees and ancestral (like those ofmore ancestral Hymenoptera) in the L-T bees. This is the reverse of the relationship that would be anticipated if derived larval characters were associated with the long tongue and other derived adult characters. For example, a rather broad mandibular apex, often with two large teeth; the presence of body setae and hair-like spicules; and well-developed, more or less cylindrical antennae are plesiomorphic for larval bees as judged by their presence in other Hymenoptera. These are widespread features in L-T bees. The corresponding apomorphies, slender mandibular apices usually lacking large teeth, the near absence of body setae, and the weak antennal papillae are features of S-T bees (and certain derived L-T bees like Apis). If the phylogeny with Colleti-dae near the root (Fig. 20-1a, b) is correct, then the L-T bee seemingly ancestral larval attributes must have reappeared in the melitted-L-T bee clade or they were retained from ancestral aculeates but independently lost in S-T bees. However, as indicated in subsequent sections, it seems likely that the Colletidae is not near the root of the bee phylogeny. In this case the larval features listed above for S-T bees may have arisen early in the evolution of an S-T clade (see Fig. 20-1d), wheras the melittid-L-T clade retained ancestral features. It must be remembered, however, that within both L-T and S-T bees, the larval characters mentioned above vary; the conclusions suggested are therefore not very strongly supported by larval characters.

Was this article helpful?

0 0

Post a comment