Camptopoeum friesei, Mocsary, 1894

Rozen, Jerome G., 2009, Immatures of the Solitary Bee Camptopoeum friesei (Andrenidae: Panurginae: Panurgini) and of its Cleptoparasite Parammobatodes minutus (Apidae: Nomadinae: Ammobatini), American Museum Novitates 3641, pp. 1-14 : 6-12

PUPA OF CAMPTOPOEUM FRIESEI

Figures 16, 17

The following is the first pupal description of any species of Camptopoeum . Terminology and format follow Neff and Rozen (1995) and Rozen and Yanega (1999). Because the more developed pupa was damaged during excavation, the illustration of the anterior end was based on that specimen and the metasoma was drawn from the second specimen. Because the second pupa was considerably smaller than the first, its illustration is slightly more magnified.

DIAGNOSIS: The pupa of this species lacks a tubercle near the base of the outer surface of the hind tibia, whereas the existence of such a tubercle is found in most Andrenidae the pupae of which have been described. The absence of this tubercle also occurs in Melitturga (Yager and Rozen, 1966: figs. 15, 16), Perdita ( Michener and Ordway, 1963; Yager and Rozen, 1966: fig. 18), Neffapis (Rozen and Ruz, 1995) , Oxaea ( Roberts, 1973) , and Protoxaea (Rozen and Rozen, in press). The short but distinct terminal spine (fig. 17) and a clearly identified transverse depression between the mesoscutum and mesoscutellum (fig. 16) found in C. friesei will immediately separate its pupa from that of Oxaea ( Roberts, 1973: figs. 7) and also Protoxaea (Rozen and Rozen, in press), which lack a terminal spine 2 and in which the pupal mesoscutum and mesoscutellum form a single continuous curve in lateral view without a transverse depression separating two curved surfaces. The short and apically rounded terminal spine of C. friesei (fig. 17) easily identifies its pupa compared with that of Perdita with its elongate, apically pointed terminal spine ( Michener and Ordway, 1963; Yager and Rozen, 1966: fig. 18). The lack of mesoscutellar tubercles of C. friesei distinguishes its pupa from that of Melitturga , in which the paired tubercles are pronounced (Yager and Rozen, 1966: figs. 15, 16).

DESCRIPTION: Length 7.9, 8.6 mm; body without setae.

Head: Scape and pedicel unmodified, without tubercles; flagellomeres unmodified. Mandible with large ventral tuberclelike swelling two-thirds distance to apex; malar area without tubercle. Vertex with pair of low, scarcely noticeable, median vertical tubercles mesad of lateral ocelli; median ocellar tubercle conspicuous; lateral vertical tubercles not evident, although area behind upper compound eye faintly varicose; frontal tubercles not evident. Surface of compound eye smooth. Proboscis long (corresponding to that of adult), so that glossa extends to middle of metasoma in lateral view.

Mesosoma: Lateral angles of pronotum faintly varicose; lateral lobes somewhat swol-

2 Roberts (1973: fig. 7) mistook part of the genitalic structure for the terminal spine, which is a projection from the last metasomal tergum. Neither Oxaea nor Protoxaea have a terminal spine (Rozen and Rozen, in press)

len, varicose; mesoscutum without tubercles, but surface somewhat swollen and varicose on each side, so that median line deeply impressed; mesoscutellum without tubercles but produced on each side as low, faintly varicose mound separated from other by median depression; axilla slightly produced; metanotum slightly produced on each side. Tegula without tubercle, surface slightly varicose. Forewing without tubercle. Forecoxa with moderately slender apical tubercle; foretrochanter with ventral, rounded tubercle; forefemur produced at base as rounded projection; foretibia and -tarsus unremarkable. Mid- and hind coxa with apical, sharply pointed tubercle; midtrochanter with low rounded tubercle; hind trochanter apically produced but nontuberculate; mid- and hind femora and tarsi unremarkable.

Metasoma: Terga I–VI each with transverse row of very small, often sharply pointed tubercles; those of tergum I few in number; those of all terga most pronounced sublaterally, tending to be small or absent medially. Sterna without tubercles. Terminal spine apically rounded, perhaps slightly varicose apically, short, poorly differentiated from tergum VIII, which is well developed.

MATERIAL EXAMINED: 1 male pupa, AUSTRIA: Burgenland: Illmitz, Reserve Seewinkel-Neusiedlersee , VII-18-2007 (J.S. Ascher); 1 male pupa, same except: (M.G. Rightmyer) .

REMARKS: Unfortunately the female pupa has not been collected, so that it is unknown if the leg tubercles might be different to accommodate adult setal development.

OVARIAN STATISTICS AND MATURE OOCYTE OF PARAMMOBATODES MINUTUS

Figures 18–21

that of the mature oocytes) by the distance between the outer rims of the tegulae of the female from which the egg was taken. One can then compare the resulting figures (indices) of different sized adults to determine if their eggs are larger, smaller, or of the same size irrespective of body size. Iwata and Sakagami (1966: table 2) proposed a classification of the indices consisting of five size categories. Indices of both P. rozeni and P. minutus fall well within dwarf, the smallest category. This is not surprising since eggs of cleptoparasitic bees tend to be among the smallest of all groups of bees (Rozen, 2003). Furthermore, eggs of all other Ammobatini (representing five additional genera) that have

TABLE 1 Number of Ovarioles and Number and Size of Mature Oocytes of Two Species of Parammobatodes Data for P. rozeni from Rozen and Özbek (2003) and Rozen (2003). Numbers in first three columns are average values of the two specimens of P. minutus been examined are either dwarf or small (ibid.).

The values of the other statistics in table 1 are consistent with those of other Nomadinae (Rozen, 2003) in that the ovariole number (number of ovarioles per ovary) is elevated compared with the plesiomorphic number (4:4) for Apidae including other cleptoparasitic lineages. Also, the total number of mature oocytes is high compared with solitary bees, but this is usually the case with cleptoparasitic bees irrespective of their families (Rozen, 2003). The adaptive significance of these phenomena has been discussed by Alexander and Rozen (1987) and Rozen (2003).

DIAGNOSIS: As with other Ammobatini ( Rozen and Özbek, 2003) the mature oocyte of this species has an operculum the surface of which slants obliquely from the long axis of the oocyte. The oocyte is slightly smaller that that of P. rozeni , but otherwise they are nearly identical, although the more regular patterning of the operculum of Parammobatodes minutus (fig. 20) seems to contrast with the less uniform patterning of that of P. rozeni ( Rozen and Özbek, 2003: fig. 38). Both species are unusual in that the operculum is strongly tinted; that of P. rozeni ebony colored while that of P minutus is more or less amber but always paler than ebony. The range of color depth of the operculum is likely due to the age of the oocyte after it reached maturity: the older the mature oocyte, the darker its color.

Because mature oocytes have opercula that are tinted amber, they are easy to identify although very earlier mature oocytes may be colorless because of the gradations in ages of their maturity. Oocytes, counted through the follicular tissue, take on the shape of the mature oocytes before becoming mature. Hence, a few mature oocytes may have been mistaken for a premature oocyte because they had not yet acquired any color. In addition to those with tinted opercula, two tinted opercula unattached to oocytes were found in separate ovarioles next to the calyx of one ovary of one specimen. With Parammobatodes rozeni , such opercula without attached oocytes were considered reabsorbed ( Rozen and Özbek, 2003), as seems to be the case here.

DESCRIPTION OF MATURE OOCYTE (figs. 18– 21): Length of longest oocyte (measured from front edge of operculum to posterior tip) 0.68 mm, maximum width behind operculum 0.19 mm; average egg index 0.37 (N 5 2) (dwarf). Shape (fig. 18) nearly straight, with anterior thick, flat operculum slanting posteriorly on dorsal surface; operculum oval with distinct flange that drapes over and covers sides of oocyte; opercular length (exclusive of draped flange) (0.21 mm) greater than width; opercular surface somewhat uneven as seen through stereomicroscope; as seen with SEM, opercular surface with large circular area pebbled in low relief, becoming smooth toward periphery (fig. 20). Micropyle (fig. 21) a somewhat irregular and nonelevated hole near anterior edge of operculum (for further description, see Remarks below). Operculum amber, transparent; rest of chorion clear, seemingly colorless, shiny, and without surface sculpturing when examined with SEM; dorsal surface of oocyte with series of transverse folds (as in Rozen and Özbek, 2003: figs. 37, 44).

MATERIAL EXAMINED: Oocytes from specimens collected: AUSTRIA: Burgenland: Illmitz, Reserve Seewinkel-Neusiedlersee , VII- 22-2007 (G. Hölzler) .

REMARKS: Almost certainly the egg of this species when found in the host cell will have its posterior end bent under (or over) the anterior part of the egg as has been reported for Pasites maculatus Jurine ( Rozen, 1986: figs. 6, 9) and Oreopasites ( Bohart, 1970) . Evidence of this is the dorsal transverse folding of the chorion at midsection that allows the stretching of the chorion necessary to accommodate the Ushaped bending of the egg. This phenomenon is unique to the Ammobatini genera, the eggs or oocytes of which have been studied. However, it is not found in Sphecodopsis (Pseudodichroa) , 3 perhaps because of the cell lining of the host, Scrapter (Colletidae) ( Rozen and Michener, 1968). The lining is cellophanelike, similar to that of Colletes . If it is separated from the cell wall as reported by Rozen and Favreau (1968) for Colletes , then perhaps the egg does not need to be bent Ushaped since it would lie in the empty space between lining and cell wall.

The micropylar duct of Parammobatodes minutus appears to penetrate the thick chorion of the operculum diagonally, and the parallel ridges at its entrance on the surface (similar to those of Pasites maculatus [ Rozen and Özbek, 2003: fig. 47]) suggests that duct may lead to a multipored, true micropyle below or it may be a remnant of it. This suggestion is not unreasonable since by far most bee taxa have multipored micropyles even including some Ammobatini such as Ammobates carinatus Morawitz and a number of species of Oreopasites (Rozen, 2003: figs. 26, 31, 36). A recessed micropyle of this sort, thus, might have come about through selection pressure for the bee egg to have a thick exposed chorion as protection against the mandibles of the host female. The ontogenetic changes in the oocyte of Leiopodus abnormis Jörgensen (Rozen, 2003: figs. 16–19) (a presumably unrelated cleptoparasitic bee in the Protepeolini) during late oogenesis give credence to this idea.

POSTDEFECATING LARVA OF PARAMMOBATODES MINUTUS

Figures 22–25

DIAGNOSIS: The mature larva of Parammobatodes minutus strongly resembles those of other Ammobatini in having a body form

3 Perhaps non-coincidentally, species of this subgenus have an egg the chorion of which is uniformly covered with a network of fine interconnecting ridges ( Rozen and Michener, 1968: figs.9, 10; Rozen and Özbek, 2003: figs. 48, 49). Thus, it is probably stiffened and presumably not sufficiently elastic to bend.

that tapers both anteriorly and posteriorly and that often has pronounced body segmentation ( McGinley, 1981; Rozen, 1954, 1966, 2008; Rozen and McGinley, 1974). Internal head ridges tend to be reduced or even lacking, especially the postoccipital ridge, and the labiomaxillary region is greatly fused and recessed, with the labial palpi scarcely evident. On all, the hypopharynx is nonspiculate, and mandibles tend to be short and robust basally. The following key will separate the larvae of taxa that have been described to date (references to their descriptions cited parenthetically after their name).

KEY TO KNOWN MATURE LARVAE OF AMMOBATINI

1. Body with distinct paired dorsolateral tubercles (fig. 22; Rozen, 2008: fig. 12)......... 2

– Body without distinct paired body tubercles ( Rozen, 1954: fig. 1, 1966: fig. 59; Rozen and McGinley, 1974: figs. 1, 10, 18), although in Oreopasites dorsolateral area on many terga slightly produced forming dorsolateral angle (or ‘‘hip’’) on each side.............. 3

2(1). Each parietal swollen, directed forward as seen in lateral view (Rozen, 2008: fig. 15); mature larva extremely small, less than 4 mm long. Chiasmognathus pashupati Engel (Rozen, 2008)

– Each parietal normally sloping posteriorly as seen in lateral view (fig. 25); mature larva larger, longer than 6.5 mm long......... Parammobatodes minutus (Mocsáry) (present study)

3(1). Labral tubercles far apart, distance between them about four-times basal diameter ( Rozen and McGinley, 1974: fig. 15); upper frons with median swelling (ibid.: figs. 15, 16)............................ Ammobates carinatus Morawitz ( Rozen and McGinley, 1974)

– Labral tubercles more closely set, distance between them at most twice, to less than, basal diameter ( Rozen, 1954: fig. 5; Rozen and McGinley, 1974: figs. 16, 23 4); upper frons without median swelling ( Rozen, 1954: fig. 6; Rozen and McGinley, 1974: figs. 8, 24).. 4

4 Although the illustration ( Rozen and McGinley, 1974: figs. 23) of the labral tubercles of Pasites histrio (Gerstaecker) (5 Morgania histrio transvaalensis Bischoff ) seems to indicate that they are separated by a distance of more the twice their basal diameter, re-examination of the specimens confirms their separation by no more than two diameters.

4(3). Labral tubercles separated by distance equal to, to twice, their basal diameter ( Rozen and McGinley, 1974: fig. 23).... Pasites histrio (Gerstaecker) (5 Morgania histrio transvaalensis Bischoff [ Rozen and McGinley, 1974])

– Labral tubercles separated by distance clearly less than their basal diameter ( Rozen, 1954: fig. 5; Rozen and McGinley, 1974: figs. 7)....... Oreopasites and Sphecodopsis (Pseudodichroa) fumipennis (Bischoff) (as Pseudodichroa fumipennis Bischoff ) 5 ( Rozen and McGinley, 1974)

5 Rozen and McGinley (1974) regarded larval Sphecodopsis (Pseudodichroa) fumipennis (Bischoff) (as Pseudodichroa fumipennis Bischoff ) as quite distinct from other ammobatines. Because their limited material consisted of only predefecating specimens, which have now deteriorated to some degree, we do not attempt here to separate this species from Oreopasites in the key. However, whereas all known predefecating larvae of the North American genus Oreopasites are normally slender,they can immediately be distinguished from physogastric predefecating larvae of the southern African S. fumipennis , which when fully fed has a maximum body width 4–5 times greater than its maximum head width.

DESCRIPTION: Length approximately 6.5 mm. Head: Unpigmented except for mandibular apices, mandibular internal ridges, and hypostomal ridge, which are weakly pigment- ed; sensilla nonsetiform; spiculation, including that of hypopharynx, absent.

Head size small compared with body; head capsule wider than long in frontal view (length measured from vertex to level of posterior mandibular articulation); as seen in lateral view (fig. 25), parietals at top of head sloping backward, not projecting forward as in Chiasmognathus pashupati Engel (Rozen, 2008: fig. 12). Tentorium absent except at pits (possibly because specimen was about to molt), existing tentorial arms short, fragile; anterior tentorial pit very small, close to anterior mandibular articulation; posterior tentorial pit small, below and posterior to curve of hypostomal ridge as ridge bends upward; postoccipital ridge not evident so that hypostomal ridge fading completely at its posterior end and posterior margin of head capsule a faint line extending dorsolaterally to cervical fold; integument in front of posterior margin nonspiculate, integument behind margin becoming conspicuously spiculate. Median longitudinal thickening of head capsule absent; hypostomal ridge evident, becoming weaker posteriorly; pleurostomal ridge weak, not clearly defined; epistomal ridge mesad of anterior tentorial pits absent, weak laterad of pits. Parietal band not evident. Antennal prominence low; antenna positioned as in figs. 24, 25; antennal disc and papilla not clearly differentiated; papilla bearing two sensilla. Labroclypeal region strongly projecting beyond frontal area with labrum large, without sclerite, with two sharply defined, apically acute, small, sensilla-bearing tubercles that are separated by distance slightly less than their basal diameter; lower margin of labrum extending downward to meet hypopharynx where they join pharynx; lower part of labrum (epipharyngeal surface) overridden laterally by mandibular apices when mandibles are closed.

Mandible (too small to be dissected from head but interpreted through cleared head capsule) moderately long for an ammobatine, with base moderately developed (but not globular); apex strongly curved downward and sharply pointed; cusp and apical concavity not differentiated, without denticles; outer surface with scattered denticles; dorsal and ventral apical edges with conspicuous, sharply pointed, apically directed teeth. Maxilla strongly fused to labial region, its apex (except for palpus) small, surpassed by hypopharynx; galea and maxillary sclerites absent; palpus large at base, tapering rapidly to pointed apex, so that palpal length clearly shorter than basal diameter. Labium not divided into pre- and postmentum; premental sclerite absent; labial palpus presumably represented only by a sensillum somewhat posterior to salivary opening. Salivary opening simple, circular, without lips. Hypopharynx defined by mouth above, by maxillae laterally, but without hypopharyngeal groove below.

Body (figs. 22, 23): Without setae, with fine spicules on prothorax and on venters of other thoracic segments and anterior abdominal segments; integument without spines or scler- otized tubercles. Body form slender, elongate; intersegmental lines weakly incised between most body segments; dorsolateral intrasegmental lines not evident; thoracic and abdominal segments 1–8 with paired, low, conical (i.e., not transverse) dorsolateral tubercles, which do not project dorsally as far as body midline; paired tubercles absent on abdominal segments 9, 10; venter of abdominal segment 9 not produced; segment 10 longer than basal diameter in lateral view, not bulging ventrally, attached centrally to 9; anus a transverse slit, apical on 10; perianal area with dorsal and ventral lips (fig. 23). Spiracles faintly pigment- ed, subequal, moderately small; spiracles projecting beyond body wall, with rims; peritreme present, moderately wide; atrium globular; atrial wall without denticles or rings; primary spiracular opening with collar; subatrium (fig. 15) moderately long, with about 14 chambers, all of approximately equal diameter; thus, subatrium not tapering from outer most to inner most chamber as in Camptopoeum friesei (fig. 14). Sex characters unknown.

MATERIAL EXAMINED: 1 postdefecating larva, AUSTRIA: Burgenland: Illmitz, Reserve Seewinkel-Neusiedlersee , VII-18-2007 (J.S. Ascher) .

Figures

Figs. 16, 17. Diagrams of pupa of Camptopoeum friesei, lateral views, drawn from different specimens to different scales. 16. Anterior part of body. 17. Posterior part of metasoma, with short terminal spine. Scales (5 1.0 mm) refer to closest figures.

Figs. 18–21. Oocytes of Parammobatodes minutus. 18. Entire oocyte, anterior end toward right, lateral view, with chorion too badly damaged to demonstrate transverse dorsal folds. 19. Close-up of operculum, showing flange draped over and clinging to rest of oocytes; flange almost certainly outstretched and continuous with surface of operculum after deposition in host cell, as in Pasites maculatus (see Rozen, 1986: figs. 5, 6), presumably an adaptation to obscure seam in cell wall so that returning host female will not detect parasite egg. 20. Close-up of another oocytes, end view, showing texture of surface and micropyle on left. 21. Close-up of micropyle.

Figs. 22–25. Diagrams of postdefecating larva of Parammobatodes minutus. 22. Entire larva, lateral view. 23. Close-up of terminal abdominal segments, lateral view. 24, 25. Head, frontal and lateral views, respectively. PTP 5 posterior tentorial pit. Scale (5 1.0 mm) refers to fig. 22.