Scopa Colletidae

1st m-cu 2nd m-cu 2nd abscissa of Rs 1st r-m 2nd r-m

1st abscissa of R1 2nd abscissa of R1 r and Rs

Rs+M, 2nd and following abscissae of M

It has the drawback that the first is not technically a crossvein, but is thought to be a transverse section of a longitudinal vein, Rs.

Louis (1973) reviewed prior alar terminologies and proposed a new nomenclature for veins, attempting to avoid considerations of homology and phylogeny. He called the submarginal crossveins the first to third RM, or radiomedial veins. I believe that the less technical expression, submarginal crossveins, gives a clearer indication of what these veins are really like and of their relation to the submarginal cells.

Figure 10-8 shows the stigma, marginal cell, and nearby structures, with lines to show how measurements that are used in the keys and by various authors are to be made. The width of the prestigma is measured to the costal margin of the wing, and the expression is thus a misnomer, for it is more than the width of the prestigma

Wing Marginal Cells

Figure 10-8. Forewing of Ceratina rupestris Holmberg, showing how certain measurements are made, as follows: (a) length of stigma; (b) length of costal edge of marginal cell or of margin of cell on the costa or wing margin (not a useful measurement in a wing like this, in which the cell diverges gradually from the wing margin); (c) length of marginal cell beyond stigma; (d) length of marginal cell; (e) length of free part of marginal cell; (f) length of prestigma; (g) width of prestigma (to wing margin); (h) width of stigma. The breaks in the wing veins are the alar fenestrae.

Figure 10-8. Forewing of Ceratina rupestris Holmberg, showing how certain measurements are made, as follows: (a) length of stigma; (b) length of costal edge of marginal cell or of margin of cell on the costa or wing margin (not a useful measurement in a wing like this, in which the cell diverges gradually from the wing margin); (c) length of marginal cell beyond stigma; (d) length of marginal cell; (e) length of free part of marginal cell; (f) length of prestigma; (g) width of prestigma (to wing margin); (h) width of stigma. The breaks in the wing veins are the alar fenestrae.

proper. The length of the costal edge (or margin) of the marginal cell or of the marginal cell on the costa is repeatedly used. This is the measurement from the apex of the stigma distad to the apex of the cell, or, if the cell is truncate, to the point where the cell diverges abruptly from the wing margin. As suggested by Figure 10-8, it is a poor measurement to use for wings in which the marginal cell bends gradually away from the wing margin. A common usage is basal vein for the first abscissa of vein M in the forewing (Fig. 10-7).

The jugal lobe and vannal lobe of the hind wing are both measured from the wing base to the apices of the lobes. Thus, on Figures 10-6 and 10-7 one might say that the jugal lobe is about two-thirds as long as the vannal lobe. The terminology of the veins and cells of the posterior parts of the wings varies; the word "vannal" is sometimes replaced by "anal," and the abbreviation V by A. Thus the vannal lobe is sometimes called the anal lobe.

The alar fenestrae are small, clear areas occupying specific sites in various veins (Fig. 10-8). Flexion lines, often faintly visible in the wing membrane, cross veins at these fenestrae.

The legs. Some authorities advocate a system for identifying parts of legs that assumes that all legs are pulled out laterally at right angles to the long axis of the body. Although I appreciate the logic of that system, I here follow the more traditional system in which the legs are considered to be in their normal positions. Thus, the corbic-ula of corbiculate Apidae is on the outer, not the anterior, surface of the hind tibia, and the two hind tibial spurs are outer and inner, not anterior and posterior. Some additional positional terminology is indicated by numbers in Figure 10-9.

The tibial spurs are the movable inferior apical spurs on the tibiae; there is one spur (part of the strigilis) on the front tibia, one on the middle tibia, and in nearly all bees two on the hind tibia. The tibial spines (Fig. 10-9) are immovable, sharp, superior apical projections, usually small in size, often blunt or minute, found in some bees. There are none, one, two, or rarely three spines per tibia; often they are mere angles. The tibial spur of the front leg consists of a main axis or malus, and a thin platelike scraper or velum directed toward the main axis of the leg; the velum usually does not extend to the apex of the malus. In some Apinae a prong or projecting ridge on the anterior side of the malus is termed the anterior velum by Schonitzer (1986) and Schonitzer and Renner (1980). The inner hind tibial spur ofthe hind leg is especially important taxonomically. This spur usually has two toothed margins. It is the inner one that is commonly elaborated in various ways. Following custom, I have described this margin as ciliate if it has slender, almost hairlike projections (usually numerous), although in many cases the appearance is like that ofa fine comb. Because the finely serrate or ciliate (or intermediate) condition is common in Hymenoptera (Gennerich, 1922) and frequently the same as that of the outer hind tibial spur and of the middle tibial spur, such spurs are often described as simple, meaning, I suppose, unmodified. It may be, however, either coarsely serrate or pectinate. Again following custom, I have described a spur as pectinate if its inner margin is produced into several long, coarse, often blunt projections, even though the number of such projections is in some cases reduced to only one or two. The sockets of the hind tibial spurs and their relation to the tibia vary among taxa and were the subject of a study by Cane (1979).

The basitibial plate (Figs. 8-9, 10-9) is on the upper or outer side of the base of the hind tibia of many bees. It is best developed in females and presumably is important for support as bees move up or down their burrows in the soil or tamp the cell surfaces with the pygidial plate. The importance of the latter function is suggested by the observation that most female bees either have both pygidial and basitibial plates or lack both. Commonly, the ba-sitibial plate is surrounded by a carina or a sharp line of some sort and its vestiture (if any) differs from that of adjacent regions, but it may be indicated only by a series of tubercles, as in many Euryglossinae, or even by a single tubercle that indicates its apex; in some cases (as in many Xylocopa) its apex is represented by a structure near the middle of the tibia instead of more or less near the base. In some bees (e.g., certain groups of Centris) the basal or central part of the basitibial plate is sharply elevated above the rest of the plate surface. Such an area is called the secondary basitibial plate.

Most S-T bees possess a pair of brushes or combs, best developed in females, on the middle legs. One is on the underside of the tibia, sometimes also including a brush on the basitarsus. The other is on the basal part of the femur, sometimes extending onto the trochanter. These structures are opposable and are used in cleaning or transferring pollen from the ipsilateral (same side) foreleg (Jander, 1976). They are termed the midtibial brush or comb and the midfemoral brush or comb (see Sec. 13). A comb is a single row of bristles, whereas a brush is less organized.

On the inner surface of the hind tibia of most bees is an area of variable size covered with hairs of uniform length, usually blunt, truncate, or briefly bifid. These hairs, the keirotrichia (Fig. 10-11), appear to serve for cleaning the wings. In some bees they are replaced by longer, more ordinary hairs that may function as part of the scopa in females.

On the hind basitarsus of many female bees is a distal process that extends beyond the base of the second tarsal segment. (For simplicity, I use the expression "tarsal segment" instead of tarsomere.) Sometimes this process bears on its apex a small brush, the penicillus (Fig. 10-9). (This brush bears no relation to the tibial tuft known as the penicillum in the Meliponini.)

Between the tarsal claws there is often a protruding, padlike arolium (Fig. 10-10). Its lower distal surfaces are almost always dark, often black, a fact that helps to distinguish it from associated pale structures. See Michener (1944) and Figure 10-10 for details of the structures between the claws. It is likely that comparative study of these structures would yield new characters of value for bee phylogeny or systematics, although loss of the arolium has occurred repeatedly among bees.

The scopae. Female bees have scopae for holding and transporting pollen; males do not. Exceptions are the Hy-laeinae and Euryglossinae in the Colletidae, parasitic and robber bees in various families, and queens of highly eu-

basitibial ^

Figure 10-9. Hind leg of a female bee, hairs omitted except those that form the penicillus. The numeral 1 indicates the posterior or upper margin of the tibia; 2, the outer surface; and 3, the distal or apical margin. Modified from Michener, McGinley, and Danforth, 1994.

social bees (Meliponinae and Apinae), all of which lack scopae. In the keys and descriptive comments I often refer to the scopa without reminding the user that scopae are found only on females. The scopa consists of pollen-carrying hairs. These are not usually the hairs and brushes with which pollen is removed from flowers, but are the brushes on which pollen is carried back to the nest. Some pollen may be carried on various parts of the body, but scopae occur principally on the hind legs (Figs. 6-4, 8-5) or on the metasomal sterna (Fig. 8-7). In most bees the scopal hairs are on the hind legs, but in nonparasitic Megachilidae they are on the metasomal sterna; in some colletids and halictids they are on both the underside of the metasoma and on the hind legs. If fringes of scopal hairs surround a space in which pollen is carried, they are said to form a corbicula. The best-known corbicula, on the outer side of the hind tibia of the corbiculate Apidae (Euglossini, Bombini, Meliponini, Apini), consists not only of the fringes of hairs but also of the enclosed concave or flat surface (Fig. 10-11). Other corbiculae are on the undersides of the hind femora ofAndreninae, Halic-tidae, Colletidae, and others, and on the sides of the propodeum of many species of Andrena, some halictids, and some colletids.

The metasoma. For simplicity and to save space, as noted in Section 9, metasomal terga and sterna are referred to as T1, T2, etc., and S1, S2, etc., T1 and S1 constituting the basal segment of the metasoma (Fig. 10-12). Each metasomal tergum or sternum (except for the ante-riormost and the reduced apical ones) consists of a plate commonly marked by some transverse lines, as follows. First, across the anterior margin, always completely hidden in the intact metasoma, is the antecostal suture. The equivalent internal ridge is the site of attachment of longitudinal intersegmental muscles. The very narrow rims of the tergum and sternum anterior to the antecostal suture are the acrotergite and acrosternite, plus the apode-mal margin that expands to form apodemes laterally. Second, nearer to the middle of each plate is another transverse line, the gradulus. Typically, the surface basal

^distilarsus

^distilarsus

Figure 10-10. Dorsal views of apices of tarsi. a, Andrena mimetica Cockerell; b, Anthophora edwardsii Cresson; c, Xylocopa orpifex Smith; d, Anthidium atripes Cresson. Arolia are well developed in a and b, absent or greatly reduced in c and d. The median, arolium-like structure in d is membranous, not obvious in dry specimens, and Anthidium is considered to lack arolia. From Michener, 1944.

to the gradulus, i.e., the pregradular area or disc, is slightly elevated compared to that behind it, rendering the gradulus a minute step, as shown in Figure 10-12b. The ends of the tergal graduli, unless bent strongly to the rear, are usually near the spiracles. If bent strongly to the rear, the resultant longitudinal lines are called lateral parts or lateral arms of the graduli or, if carinate, lateral gradular carinae, sometimes elevated to form lateral gradular lamellae or lateral gradular spines. The graduli—except for their lateral arms, when present— are often concealed by the preceding terga or sterna on the intact metasoma but, especially on T2 and S2, are sometimes exposed or can be exposed easily by slightly extending the metasoma artificially. Third, near the posterior margin of each tergum and sternum is usually another transverse line, the premarginal line, separating the marginal zone (posterior marginal area of Michener, 1944; apical depression of Timberlake, 1980b) from the rest of the sclerite (Fig. 10-12). This zone is often depressed but in other cases differs only in sculpturing from the area basal to it; sometimes the marginal zone is not differentiated at all. The region between the gradulus and the premarginal line can be called the disc when a name is needed for it. Sometimes, e.g., on T2 of Exomalopsis, the premarginal line is arched far forward, so that the marginal zone is broad and the disc reduced to a transverse zone. The dorsolateral parts of the tergal discs (between the graduli and the premarginal lines), especially on T2 to T4, are often somewhat elevated, convex, and frequently shiny. These dorsolateral convexities frequently accentuate the premarginal lines, which limit the convexities posteriorly. The narrow posterior margin of the marginal zone is often recognizably different from the rest of the zone in sculpture or is elevated; this is termed simply the margin. Packer (2004a) provided an account of sternal variation among bees.

T1 differs from other terga because its base is constricted for the narrow connection with the thorax. Its dorsal, horizontal surface is nonetheless similar to that of

Figure 10-10. Dorsal views of apices of tarsi. a, Andrena mimetica Cockerell; b, Anthophora edwardsii Cresson; c, Xylocopa orpifex Smith; d, Anthidium atripes Cresson. Arolia are well developed in a and b, absent or greatly reduced in c and d. The median, arolium-like structure in d is membranous, not obvious in dry specimens, and Anthidium is considered to lack arolia. From Michener, 1944.

succeeding terga, often having a marginal zone, premar-ginal line, and disc, each easily recognizable. Toward the base, T1 is strongly declivous. Often there is a transverse line or carina near the summit of the declivous surface and more or less separating it from the horizontal surface. This line or carina may be the gradulus of T1, although I do not use that term for it. It is the line or carina delimiting the anterior surface or anterior concavity of T1, and it is well developed in Heriades and certain other small

Figure 10-11. Hind tibia and basitarsus of a worker of Plebeia frontalis (Friese). a, Outer surfaces, showing the scopa reduced to fringes around the smooth and largely hairless tibial corbicula; b, Inner surfaces. From Michener, McGinley, and Danforth, 1994.

premarginal premarginal

Osmiini. In addition, the lateral line of T1 (Fig. 10-12a) is usually present.

In various colletids and andrenids, a lateral fovea of T2 (Fig. 10-12a) is usually developed if the facial foveae are developed. The lateral fovea ofT2, like the facial fovea, is lined with apparently secretory cells (Ramos et al., 2004). It is likely that whatever the function of facial foveae, that of the lateral T2 foveae is the same. Both foveae are little evident in males, but often distinct in females.

The terga are often provided with transverse bands of pale hair. These are often nonhomologous, being on different parts of the terga. The terms metasomal bands and fasciae are applied indiscriminately, apical bands if on or overlapping the marginal zones, basal bands if on the discs.

The pygidial plate (Fig. 10-13) is a usually flat plate, commonly surrounded laterally and posteriorly by a carina or a line and in some cases produced as an apical projection, on T6 of females and T7 of males. The line demarcating the pygidial plate may be a median elaboration of the gradulus of T6 (females) or T7 (males), as is well shown, for example, in female Eucerini. In many bees the gradulus is absent laterally, so that only the part demarcating the plate is present. In other cases, the gradulus is entirely transverse, i.e., it extends across the tergum basal to the pygidial plate, which nonetheless is margined. (This observation makes one wonder if pygidial plates could have two, nonhomologous origins.) Sometimes the pygidial plate is reduced to a flat-topped spine or is completely absent. It is more often absent or rudimentary in males than in females. The prepygidial fimbria is a band of dense hairs across the apex of T5 of females. It is conspicuously different from, usually denser than, the apical hair bands or fasciae that may be present on preceding terga (Fig. 10-13). The prepygidial fimbria is considered

Figure 10-12. Metasomal structures of a male bee. a, Diagrammatic lateral view of the metasoma; b, Lateral view of T3; c, Ventral view of S3; d, Dorsal view of T3. a, b, and c, modified from Michener, 1944.

absent if the hair band on T5 is like that on T4 and more anterior terga. In females of Nomadinae the prepygidial fimbria is modified, consisting of uniform short hairs on the often sloping surface of the apical median part ofT5, forming an area suggestive of a pygidial plate, sometimes distinctly outlined by a ridge or change of surface slope and texture. This area is called the pseudopygidial area. Dense hairs on T6 of females, on each side of the pygidial plate, constitute the pygidial fimbria (Fig. 10-13), which is divided into two parts by the plate.

T7 of the female, always completely hidden, consists of two weakly sclerotized plates called T7 hemitergites; these are part of the sting apparatus. Each contains a spiracle, which thus readily identifies the T7 hemitergites (Fig. 10-14). The T8 hemitergites are similar-sized plates lacking spiracles. The second valvifers or female gono-coxites which are also similar-sized plates that lack spiracles, give rise basally to the rami of the second valvulae and apically to the third valvulae or female gonostyli, also called sting sheaths. Other terminology includes stylet for the fused second valvulae, lancets for the first valvu-lae, and valve for the dorsal flap near the base of each

Figure 10-13. Diagram of apex of metasoma of a female bee, such as a eucerine. The tergal numbers are indicated on the righthand side.

Arolia Bee

lancet. The small first valvifer is called the triangular plate in literature on honey bees.

There is now good evidence that the sting apparatus varies among taxa and should be routinely extracted and studied in comparative examinations of genus-level relationships among bees. The principal paper on this topic is by Packer (2003). Rightmyer (2004) introduced to melittology a more recent terminology based on ho-mologies recognized by Scudder (1971), which do not agree with the homologies accepted by Michener (1944, 2000). The principal changes in terminology are shown in Table 10-2. With the currently increased use of characters in the female terminalia for systematic studies (Packer, 2003, 2004a; see also Fig. 27-1), the terminology becomes important. As with wing venation, the non-commital terminology will not change with differing opinions about homologies and may sometimes be more useful than the morphologically correct terminology.

The genitalia and S7 and S8 of male bees exhibit many interesting characters and may be dissected out for study, although they are almost always retracted in killed specimens. In some groups, e.g., Heriades in the Megachilinae, more anterior sterna such as S6 and S5, and rarely even S4 and S3, are also hidden and modified, but the expression "hidden sterna" is commonly used for S7 and S8. On a fresh or freshly relaxed specimen it is usually possible to reach between the apical exposed tergum and sternum

Table 10-2. Terminology of Sting and Associated Parts.

Noncommittal Terms accepted terms (e.g., by Michener

Snodgrass, 1956) (1944, 2000)

Terms used by Scudder (1971) and Rightmyer (2004)

Triangular plate 1s valvifer Oblong plate 2nd valvifer Sting sheath 3rd valvula or gonostylus Lancet 1st valvula

Stylet gonangulum 2nd gonocoxa gonoplac

1st gonapophysis fused 2nd valvulae fused 2nd gonapophyses and, with a hooked needle, pull out the genitalia and hidden sterna. In most cases, such dissection is not too difficult, but in the Megachilini the numerous hidden sterna are firmly connected to one another and laterally to the terga; moreover, they are often delicate and easily torn apart medially, so that successful dissection may be difficult. Beginners should start with other groups. Packer (2004a) has treated the variation in metasomal sterna of females for a broad selection of the Apoidea, thus providing a basis for recognition of previously little used or unrecognized characters.

S8 of males usually has a median basal point or angle for muscle attachment that is absent on other sterna. It is called the spiculum (Fig. 10-15b). Sterna and terga of both sexes, except for T1 and S1, have a basolateral projection or apodeme on each side (Fig. 10-12).

The male genitalia (Fig. 10-15a) have on each side, distal to the gonobase, a gonocoxite, to the distal end of which is usually attached the gonostylus. Although the gonocoxite may have some hairs, the gonostylus is frequently quite hairy and thus easily recognized. In most cases the gonocoxite and gonostylus are partly fused, often showing their articulation only on one side ofthe area of union. Frequently the fusion is complete or the gono-stylus is lost, so that instead of a two-segmented appendage, one finds an unsegmented appendage; in this case one may not know whether the gonostylus is absent or fully fused to the gonocoxite. The structure in this case is called the gonoforceps. Commonly the distal part is hairy, in which case that part probably represents the gonostylus. In various groups one can find related forms of both sorts, some with distinct gonostyli and gonocox-ites, others with the fusion or the loss of gonostyli complete.

In many groups of bees the male gonostylus is divided to its base, so that there appear to be two gonostyli on each side, arising from adjacent parts of the gonocoxite (Fig. 10-15c). These are called the upper or dorsal gonostylus and the lower or ventral gonostylus. The latter is often absent, in which case the upper gonostylus is simply called the gonostylus. The upper gonostylus is the only

Figure 10-15. Terminal structures of male bees. a, Diagram of ventral view of genitalia; b, Diagram of S8; c, Lateral view of male genitalia of Coeliox-oides exulans (Holmberg), showing upper and lower gonostyli. a, b, modified from Michener, McGinley, and Dan-forth, 1994; c, from Roig-Alsina, 1990.

Bee Genitalia

Figure 10-15. Terminal structures of male bees. a, Diagram of ventral view of genitalia; b, Diagram of S8; c, Lateral view of male genitalia of Coeliox-oides exulans (Holmberg), showing upper and lower gonostyli. a, b, modified from Michener, McGinley, and Dan-forth, 1994; c, from Roig-Alsina, 1990.

one called a gonostylus in most literature. The lower gonostylus takes on various shapes and aspects in different bees and may not be homologous in all cases. It looks quite like a gonostylus, distally directed and hairy, in such bees as Epicharis and Eufriesea, and in Coelioxoid.es (Fig. 10-15c) (Apidae). Commonly, the base is in contact with or even in common with the base of the upper gonosty-lus. In some halictines the lower gonostylus is similar to the upper, with minute hairs, but in many other halictines the lower gonostylus is directed basad and in such cases has usually been called the retrorse lobe. In allodapine bees the ventroapical plate, which bears peglike setae, may be the lower gonostylus.

The volsella (Fig. 10-15a) is often most easily identified by the heavily sclerotized dark teeth on the opposable surfaces of the digitus and cuspis. In other Hymenoptera these parts clearly function as pincers, but in bees the volsellae are reduced (often wanting) and the digitus becomes fused to the body of the volsella and thus immovable. Nonetheless, Snodgrass (1941) showed muscles that move the digitus in Andrena and Macropis. In Figure 10-15a the distal mesal volsellar structure is the digitus; lateral to it is the cuspis.

The penis valves (Fig. 10-15a) are connected on the dorsal surface, near their bases, by a bridge. In many Ap-inae this bridge is expanded posteriorly to form a dorsal plate called the spatha. Some bees have an often large and complex, largely membranous endophallus that is usually invaginated within what is here called the penis (see Roig-Alsina, 1993). The genus Apis is unique among all Hymenoptera in its enormous endophallus and the reduction of all other external parts of the male genitalia (Fig. 121-3).

Some special terms used only for the male genitalia of Meliponini, and defined in Section 120, are amphigonal, rectigonal, and schizogonal. Special terminologies used in the past for genitalic structures of Bombini are explained in Section 119, especially Table 119-1.

Genitalia and hidden sterna ofmany bees have been illustrated by Snodgrass (1941), and in the multitude of taxonomic works cited in the accounts of bee taxa in subsequent sections of this book. Particularly large collections of illustrations of these features in diverse taxa are to be found in Saunders (1882, 1884), Strohl (1908), Mitchell (1960, 1962), and Michener (1954b, 1965b). The various works of Radoszkowski are also rich in genitalic illustrations.

+1 0

Post a comment