The Origin of Bees from Wasps

It seems reasonably certain that bees arose from forms which, were they alive today, would be considered as Spheciformes—Sphecidae in the broad sense. I know of no synapomorphy of all sphecoid wasps that would justify regarding them as a monophyletic sister group of the bees. The Spheciformes are morphologically diversified to such an extent that various authors have commented on the incongruity of the custom ofdividing the bees into several families while regarding their closest relatives as constituting a single family, the Sphecidae.

Many times in the past (as long ago as Latreille in 1802), authors have regarded some sphecoid groups as families, e.g., Larridae, Crabronidae, Nyssonidae, etc. Lomholdt (1982) addressed the problem in a cladistic study in which he divided the Spheciformes into the Sphecidae s. str. (for Sphecinae and Ampulicinae) and the Larridae, which should be called Crabronidae. More recent studies such as that ofAlexander (1992) showed that the Sphecinae and Ampulicinae do not form a single clade.

The Crabronidae are held together by one strong synapomorphy, the double salivary opening of the larva, which is shared neither with the bees nor with the Sphe-cidae s. str. In the crabronid subfamily Astatinae, the larval salivary opening has been described as single. This appears to have been an observational error; whatever the explanation, the Astatinae do have a double opening like that of other Crabronidae (Evans, 1958). It is thus quite likely that Lomholdt was right in considering this family as monophyletic and the sister group to the bees.

Malyshev (1968) and others have speculated that it is in the Pemphredoninae that one finds the closest sphecid relatives of the bees. This conjecture is based in part on the small, slender body and stem-nesting habits of many of these wasps, superficially suggesting the bee genus Hy-laeus (Colletidae). Malyshev emphasized the idea that the pemphredonines' numerous small homopterous prey, sweet because of their filter chambers and content of hon-eydew, might be a preadaptation to the evolution of dependence upon nectar and pollen by bees. Many other sphecoids provision their cells with only one or a few larger prey items that are not sweet. More important, some pemphredonines, unlike other sphecoid wasps, line their nest cells with a secretion, as do most bees, but the secretion is not of the same origin as that ofbees (McGin-ley, 1980), and the cell linings are therefore not homologous in the two taxa. Adult females of some Pemphre-doninae secrete silk or silklike material from epidermal glands opening on apical terga or sterna (G. Melo, personal communication, 1995) and this material is the source of the cell linings of at least some species. In bees, on the other hand, the cell linings are secreted by Dufour's gland and salivary glands.

To me and evidently also to Radchenko and Pesenko (1994), none of this is convincing evidence of a relationship between Pemphredoninae and bees, but there are morphological indications, probably convergent, of a relationship of the bees to the crabronine branch of the

Crabronidae—the branch that leads to Larrinae, Philan-thinae, and Crabroninae in Bohart and Menke's (1976: 32) dendrogram. For example, in most Hymenoptera, including most sphecoids, the middle coxa is not greatly different in structure from the hind coxa (Fig. 14-1a), there being a short basicoxite separated by a groove from the large disticoxite (Michener, 1981b). In the Crabron-inae (including Oxybelini), Trypoxylonini, and some Larrini (but not in Philanthini) the disticoxite ofthe middle coxa is reduced and the basicoxite enlarged, as in all bees (Fig. 14-1b, c). Moreover, R. McGinley (personal communication, 1981) found similarities in the maxillary (galeal and stipital) structure between bees and certain Philanthini that suggested a possible relationship. Fi-

Figure 14-1. Middle (left) and hind (right) coxae. a, Philanthusgib-bosus (Fabricius) (Crabronidae); b, Tachytes sp. (Crabronidae); c, Anthidium illustre Cresson (Megachilidae). The basicoxite, marked "b," is enlarged at the expense of the disticoxite of the middle coxa of bees and some Crabronidae. From Michener, 1981b.

Figure 14-1. Middle (left) and hind (right) coxae. a, Philanthusgib-bosus (Fabricius) (Crabronidae); b, Tachytes sp. (Crabronidae); c, Anthidium illustre Cresson (Megachilidae). The basicoxite, marked "b," is enlarged at the expense of the disticoxite of the middle coxa of bees and some Crabronidae. From Michener, 1981b.

nally, in Alexander and Michener's (1995) phylogenetic study of S-T bees, the bees arose in all analyses from among the few Crabronidae included in the study, which therefore formed a paraphyletic group. These wasps, however, were included as an outgroup, and the study should not be viewed as informative about the relationships among the wasps. Although none of the adult crabronid synapomorphies is strong, the paired larval salivary openings of Crabronidae are a unique synapo-

morphy, as noted above. The bees have various synapo-morphies, as listed in Section 13. Therefore, it is probable that bees and Crabronidae (Larridae sensu Lomholdt, 1982) are sister groups. Of course the Series Spheci-formes, the group called sphecoid wasps, including the Sphecidae and Ampulicidae as well as the Crabronidae, is indeed paraphyletic, for the Series Apiformes arose from within it.

Was this article helpful?

0 0

Post a comment