Bees as a Monophyletic Group

The bees have long appeared to constitute a mono-phyletic or holophyletic unit (Michener, 1944; Brothers, 1975; Lomholdt, 1982). This view was strongly supported by the more recent phylogenetic studies of Alexander (1992), Brothers and Carpenter (1993), Alexander and Michener (1995), Brothers (1999), Melo (1999), and others. The following is an annotated list of some syna-pomorphies that demonstrate the monophyly of the bees:

Character a. Some of the hairs of bees are plumose or at least branched (Fig. 13-1). In most Hymenoptera they are simple, although plumose hairs are found in some other groups, e.g., some Mutillidae. Contrary to the usual opinion, I doubt that plumose hairs of bees arose as pollen-collecting and pollen-carrying structures, although of course some bees take advantage of plumosity to enhance these functions. In many bees the scopal hairs are simple, yet nevertheless carry pollen, showing that plumosity is not necessary for a pollen carrier. Moreover, plumose hairs are often found in locations where pollen is never carried, e.g., around the anterior thoracic spiracles and on the male genitalia and hidden sterna. Further, hairs are branched (plumose) in many different ways, some of them not at all suitable for pollen-collecting or pollen-carrying, indicating that the degree or type of plumosity may be under various selective pressures having nothing to do with pollen.

Possibly plumosity first arose as one way (an alternative to a great number of hairs) in which forms in a xeric environment could decrease air flow near the integu-

mental surface, and thus reduce water loss. Simultaneously, since hairs are often pale, plumosity could have been one way to increase the pale coloration often characteristic of insects in xeric environments. Presumably, pale coloration both reflects heat, helping to prevent overheating, and serves for protective coloration on the pale soils and pale vegetation characteristic of deserts.

Character b. With few exceptions, bee larvae eat pollen mixed with nectar or floral oil, or glandular secretions of adults that eat pollen and nectar. The larvae are carnivorous in related Hymenoptera except for Krombeinictus nordeme Leclercq, a Sri Lankan crabronine wasp that feeds pollen to its larvae (Krombein and Norden, 1997a,b). There are three carrion-eating species of TrĂ­gona, as noted in Sections 6 and 120. Adults of other social bees sometimes eat eggs of their competitors, queens eat trophic eggs laid by workers. Larvae of some and adults of other cleptoparasitic bees kill and possibly eat eggs or young larvae of their hosts. Otherwise bees are entirely phytophagous (= plant feeders). Since I do not think that plumose hairs arose for pollen-collecting, I do not think that feeding on pollen is a character that duplicates character a.

Eating of pollen by adults (especially females) is another synapomorphy of bees, unknown in sphecoid wasps, but it is probably a correlate of character b. Adult wasps use prey fluids as a protein source. Bees have no prey and hence eat pollen.

Character c. The hind basitarsus is broader than the subsequent tarsal segments (Fig. 10-9). In other ac-

Figure 13-1. Hairs from tibial scopas (a-g) and sternal scopa (h) of bees of the genus Leio-proctus(Colletinae). a, L. (Perditomorpha) erithrogaster Toro and Rojas; b, L. (Perditomorpha) brunerii (Ashmead); c, L. (Leioproctus) fulvoniger Michener; d, L. (Glos-sopasiphae) plaumanni Michener; e, L. (Protodiscelis) palpalis (Ducke)?; f, L. (Pygopasiphae) wagneri (Vachal); g, L. (Reedapis) bathycyaneus Toro. h, L. (Reedapis) bathycyaneus Toro. From Michener, 1989.

Figure 13-2. Seventh sterna of male bees of the family Colletidae. a, Leioproctus (Perditomorpha) inconspicuus Michener; b, Diphaglossa gayi Spinola. (Dorsal views are at the left.)

From Michener, 1986b, 1989.

Figure 13-2. Seventh sterna of male bees of the family Colletidae. a, Leioproctus (Perditomorpha) inconspicuus Michener; b, Diphaglossa gayi Spinola. (Dorsal views are at the left.)

From Michener, 1986b, 1989.

uleates the hind basitarsus is of about the same width as tarsal segments 2 through 5. The broad hind basitarsus may be related to pollen manipulation and transport. It occurs also, however, in forms that do not collect pollen, such as males and parasitic females, and in forms that do not carry pollen externally on a scopa, such as the Hy-laeinae, although it is not as well developed as in female pollen collectors. In minute Euryglossinae, such as Eu-ryglossula, and in males of some other bees, the hind basitarsus is only about 1.3 times as wide as the second tarsal segment.

Character d. The larval maxilla has one apical papilla, the palpus (Fig. 11-2). In related aculeates there are two papillae, the palpus and the galea. Lack of the galea was therefore considered a larval synapomorphy of bees by Lomholdt (1982). A galea or galea-like projection is found, however, in various bee larvae (some melittids and apines), but this galea is not a fully distinct papilla like the palpus.

Character e. The seventh metasomal tergum (T7) of the female is membranous mid-dorsally. In related ac-uleates this tergum forms a sclerotized arch. In bees the sclerotized parts are reduced to two lateral hemitergites, one on each side, that form part of the sting apparatus (Fig. 10-14).

Character f. A posterior strigil, consisting of hind tibial spurs opposed by a brush of short hairs on a shallow basal concavity of the hind basitarsus, is absent. Such a strigil is present in related aculeates, and its loss must be a synapomorphy for bees (see Lanham, 1960). Lomholdt's (1982) remark that some bees possess the posterior strigil seems to be an error. Wasps use this strigil to clean the hind legs; each leg is pulled through the contralateral strigilis, one after the other. Bees clean hind legs by rubbing the two legs against one another. This behavioral character, although synapomorphic for bees, cannot be considered independent from the strigilar loss.

Character g. Only one sperm cell develops from each spermatocyte. Wasps produce four sperm cells from each spermatocyte. This character, cited by Lomholdt (1982), has been examined in so few taxa that no certainty exists as to its distribution.

Character h. The foreleg is cleaned (or pollen is transferred from it to the middle leg) by drawing it through the flexed (at femorotibial joint) middle leg (Jander, 1976). This movement is not seen in other Hymenoptera. It is associated in females and some males with the mid-femoral brush, a brush on the lower side of the middle fe mur-trochanter, and the midtibial brush, a brush of often ordinary hairs on the lower side of the middle tibia and sometimes basitarsus. When the leg is flexed, these brushes are opposable and both surfaces of the foreleg are cleaned at one stroke (see Secs. 6 [pollen], 10). In the Hy-laeinae and Euryglossinae these brushes are weak and limited to the distal part of the tibia and the base of the femur (and sometimes the trochanter). These bees carry pollen to the nest in the crop, thus eating it rather than transferring it to a scopa. The midleg apparatus therefore probably serves only for grooming. The same is true of many male bees. In bees the ancestral cleaning movements of the foreleg by the mouthparts serve for transfer of pollen to the mouthparts for eating.

Character i. S7 and S8 of the male are modified and concealed by S6, or only the apical process of S8 (and sometimes that of S7) is exposed. In bees, S7 is commonly greatly elaborated, with lobes and hairy surfaces that may have tactile or evaporative functions (Fig. 13-2). These are common features in the Colletidae, Andrenidae, Rophitinae, Fideliini, etc. Conversely, S7 may be reduced to little more than a transverse ribbon, as in the Xylo-copini. In sphecoids S7 and S8 are relatively unmodified, S7 being usually exposed.

Character j. The bristles on the outer surfaces of the tibiae are usually absent or weak, although strong, presumably secondarily, in some parasitic forms like the No-madinae. Such bristles are very common in sphecoid wasps.

Character k. The basitibial plate (Fig. 10-9) is present, especially in females. It is well developed in almost all bees that excavate nests and shape cells in the soil, and is no doubt an ancestral bee feature, although it has been lost in many bees that do not make their own cells in the soil. It is absent in sphecoid wasps or it is present as a convergent development in a ground-nesting species of Pem-phredoninae, a subfamily whose members commonly nest in holes in wood or stems (McCorquodale and Naumann, 1988).

Character l. The mandible of the larva is simple to minutely denticulate or ends in two teeth. In related ac-uleates the mandible ends in three or more teeth. Such reduction in bee larvae is not surprising, considering that they mostly feed on pollen and nectar and do not eat arthropod prey.

Character m. G. Melo has pointed out to me that the cleft claws (Fig. 10-10) found in most bees may be a synapomorphy for bees, although they revert to simple claws in some bees, principally females. Claws are not cleft in sphecoid wasps.

Character n. A character used by Gauld and Bolton (1988) to distinguish bees from sphecoid wasps is the course of flexion lines in the forewing. In bees a line cuts across the first recurrent vein and no line cuts across vein M, i.e., the posterior margin of the submarginal cells. Presumably this pattern is synapomorphic for bees. In sphe-coid wasps the pattern is reversed. Unfortunately, these lines and the fenestrae where they cross veins are often difficult or impossible to see, and the universality of this character is not verified.

Lomholdt (1982) used the presence of a single spur on the middle tibia as a synapomorphy of bees separating them from the Larridae, properly called Crabronidae (Menke, 1993). In Lomholdt's sense, the Larridae consist of the Sphecidae s. l. minus Sphecinae and Ampulicinae. Many sphecoids, however, including those in the Larr-inae, Philanthinae, and Pemphredoninae, also have only one middle tibial spur (Bohart and Menke, 1976). Perhaps Lomholdt derived his statement on middle tibial spurs from Brothers (1975); it has to be remembered that Brothers' lists of characters for each taxon contain ancestral states, not states that arose within the taxon, so that when he says sphecids have two middle tibial spurs, he is saying that the presence of two spurs is an ancestral character for sphecids. He is not excluding the fact that many of them, like all bees, have lost one of the spurs.

The above list is substantial enough to assure mono-phyly for bees. Several of the characters, however, may not be independent. The broad hind basitarsus (character c) may have evolved as an enhancement of pollen-manipu lation or pollen-carrying capacity. In pollen-collecting females the breadth and flatness of the basitarsus are striking; the characteristic may have been carried over in a less noticeable condition to males, parasitic females, and the Hylaeinae and Euryglossinae (which carry pollen in the crop rather than externally). The broadening and this novel function of the hind basitarsus may have led to the loss of the posterior strigil (character f). If character c has to do with pollen, then obviously it is also related to b, the larval food, which of course is related to l. Character h is also associated with the use of pollen, and hence related to character c. Finally, if (as I doubt) the plumose hairs (character a) arose as pollen holders, and thereafter spread to males and were retained in the evolution of parasitic bees, then character a also is related to c.

Nonetheless, the monophyly of the bees seems clear. There is no group that could have evolved from the bees, and they therefore cannot be paraphyletic. Robertson (1904) suggested that the bees may be diphyletic, that is, that those with pygidial plates and those without could have arisen from sphecoid wasps with and without such plates. He even gave names (Pygidialia and Apygidialia) to the two groups of bees. This idea is understandable for the limited fauna known to him (eastern North America) but is not tenable when the world fauna is examined, as stated by Michener (1944). For example, colletids in eastern North America are apygidialate, but numerous related colletids, mostly in the Southern Hemisphere, have pygidial plates. As will be shown later, pygidial plates are plesiomorphic for the Apoidea. They have been lost in diverse groups of bees and sphecoid wasps.

Was this article helpful?

0 0

Post a comment