DIAGNOSIS: The large, backward-tilting mesothoracic paired dorsal tubercles (figs. 6, 8) of the Spinoliella herbsti and S. petuniae separate them immediately from the other taxa in these three genera, as will the more slender but still elongate antennal papilla (fig. 1). Features have not been identified whereby larvae of the two species can be differentiated.

MATERIAL STUDIED: Six predefecating larvae: Chile: Coquimbo Region: Tilama, XI-02-1969 (J.G. Rozen).

REMARKS: Although the body form of the predefecating larva is quite robust, the more slender form of the postdefecating larva of Spinoliella maculata, described below (fig. 8), will likely be found also typical of postdefecating S. herbsti.

The internal head ridges and tentorium on one cleared specimen were considerably modified in that the tentorial bridge was bowed forward in association with the lower end of the postoccipital ridge, and the posterior part of the hypostomal ridge turned mesad and then bent forward. As a consequence, the lower posterior end of the parietal was also bent forward. Since this phenomenon occurred symmetrically on each side of the head capsule, this may be a normal change, perhaps an accommodation allowing the development of the pupal skeleton.

The function of the setae-bearing swellings on the paired dorsal tubercles (figs. 19, 20) is unclear. Initially I thought they might provide traction. Later, after discovering fine setae of similar length without associated swellings on the venter of abdominal segment 10, I considered that they might provide a sensory function. Observations on moving live larvae may provide answers.

BIOLOGICAL NOTES: Spinoliella herbsti, Calliopsis (Liopoeum) trifasciata (Spinola), and Liphanthus (L.) sabulosus Reed nested in a barren area adjacent to an irrigated wheat field at Tilama, Coquimbo Region, Chile. The area, discovered on October 24, 1969, was excavated then and subsequently on November 2, 1969. Bordered by low trees on one side and by low vegetation elsewhere, the nesting area was unshaded during the main part of the day. Surface soil was dry, loose, and without stones but more compact just below the surface and moist from irrigation. Surface slope was about 10%, and pollen plants of none of the species grew adjacent to the site.

NEST STRUCTURE: A single burrow at the edge of a small raised area was identified when a female of Spinoliella herbsti entered it on October 24. Cells were located in the soil beneath in association with a meandering, obliquely descending, open burrow. Cells (table 1) were oriented horizontally and were narrowly pointed (i.e., not evenly rounded as in Calliopsis) at the posterior end. Their smooth surface gave no indication of a special lining. Closures were spirals, concave on the inside, and side tunnels were soil filled. Early provision loads were shaped into uncoated spheres; one incomplete load was 3.0 mm in diameter, while completed spheres were substantially larger (table 1). Homogeneously firm and moist on the inside, they were coated with a thick, shiny, clear, waterproof lining. Completed provisions rested near the center of the cell floor. Eggs were not found, but immobile young larvae, with heads pointed toward cell closures, rested on top of food spheres while eating provisions beneath their heads.

A return to the site on November 2, 1969, yielded further information. Although it was impossible to follow any single nest, they consisted of at least four or five cells with the older cells nearer to the surface of the ground. Main burrows were soil filled most of the way down. When tested, cell walls readily absorb water droplets over the entire surface with the possible exception of the middle surface where food spheres rested. One larva was found lying on its dorsum while eating the remains of the provision on its venter, a clear indication that larvae reorient during the course of feeding.

DIAGNOSIS: The single available larval specimen of Spinoliella maculata was partly destroyed after illustration (figs. 8, 9), so that details of its integument cannot be compared with those of S. herbsti. However, larval body forms of the two (figs. 6–9) were nearly identical except S. maculata was postdefecating and S. herbsti predefecating. Size difference between them was real (note: figs. 6–9 not to same scale), and the seta-bearing swelling on paired dorsal body tubercle were substantially more pronounced on S. maculata (fig. 20). All features in boldface in the description of larval S. herbsti are also characteristic of S. maculata.

MATERIAL STUDIED: One postdefecating larva: Chile: Valparaiso Region: Reñaca, X-22-94 (J.G. Rozen) collected as predefecating larva; preserved XI-9-94 as postdefecating larva.

BIOLOGICAL NOTES: Spinoliella maculata was discovered nesting in an open sandy area at Reñaca, north of Valparaiso, Chile, on October 22, 1994, where a number of nests were examined. Nests (see table 1 for nest statistics) consisted of a short main tunnel that obliquely entered the nearly horizontal, unshaded ground and then branched into one or more long downcurving laterals each of which ended by abruptly turning and connecting with a single horizontal cell. Laterals were soil filled after cell closure. The entrance hole of the main burrow was crescentic in shape, presumably because of the loose, dry nature of the surface sand, which became moist below 5 cm. When excavating nests, females kicked sand from entrances with their hind legs, the tarsi of which exhibit a row of long setae on each side, thus forming a troughlike structure. At the posterior end of a cell, a 3.5 mm long, narrowly round extension (figs. 34, 35) providing the cell with a narrowly rounded outline similar to that of figure 47 but even more exaggerated. There was no visible special cell lining. All provision masses were spherical and thickly coated with a shiny, transparent, waterproof substance except for a smaller sphere, 2.6 mm in diameter, an indication that preliminary loads of provisions are shaped as well as the final food mass. An elongate, distinctly curved, smooth, whitish egg was attached by both ends to the top of the food sphere in the sagittal plane of the cell. It had a rounded front end and an elongate, pointed posterior end. The pollen source for this species was said to be Cristaria (Malvaceae) although Claude-Joseph (1926) had claimed it to be Medicago sativa L. (Fabaceae) and Leucanthemum (Asteraceae).

DIAGNOSIS: Although mature larvae of Callonychium (C.) flaviventre, C. (C.) petuniae, C. (Paranychium) minutum, and C. (P.) n. sp. agree in many ways with those of Spinoliella and Arhysosage flava, they can immediately be recognized because their very large antennal papillae (figs. 2–4) are longer than the respective labral tubercles on the same specimen. All three genera have large antennal prominences. The more anterior paired abdominal tubercles of larval Callonychium (figs. 21–23) are apically slightly more tapering than those of Spinoliella (figs. 19, 20), which appear blunter owing in part to the prominent seta-bearing swellings.

The differences among the species of known larval Callonychium s.l., are subtle, as follows: Callonychium s.s.: antennal papilla directed nearly straight forward in lateral view (figs. 2, 3); body size larger; and prothoracic paired dorsal tubercles strongly directed backward (figs. 10, 12). Callonychium (C.) petuniae (figs. 12, 13) has more acutely pointed midbody paired dorsal tubercles whereas these tubercles of C. (C.) flaviventre (fig. 10) are more apically rounded. Paranychium (including C. minutum, C. n. sp., and an unknown species from Peru²) with antennal papilla curving upward in lateral view (fig. 4); body size smaller; prothoracic paired dorsal tubercles nearly erect (figs. 14, 16), and mesothoracic paired dorsal tubercles little modified (figs. 14, 16).

MATERIAL STUDIED: TWO postdefecating larvae: Argentina: Tucumán Province: 8 km SW Ticucho, collected as predefecating larvae III-25-1990; preserved IV-12-1990 (J.G. Rozen, A. Roig-Alsina). Five postdefecating larvae: Argentina: Tucumán Province: Amaichá del Valle, III-5, 11-1990 (J.G. Rozen, A. Roig-Alsina).

BIOLOGICAL NOTES: The nesting biology of Callonychium flaviventre was investigated in Tucumán Province, Argentina, first at Amaichá del Valle a number of times between March 5 and 23 and then at 8 km southwest of Ticucho on March 25, all in 1990. At Amaichá del Valle the ground was dry and mostly horizontal on the surface with some large and small surface stones and moderately soft sand with stones below the surface, becoming more compacted below 3 cm. Grindelia (Asteraceae), the larval food source, grew 3–4 m away. However, the unusual nesting biology of the subgenus Callonychium was only fully understood after the nest of C. flaviventre was examined at the Ticucho site. There, because of the even soil texture, the shape of the cell and its entrance were fully explained, as described in the following paragraph. Cells were approximately 8 cm deep at the lower end of main tunnels that first penetrated the surface at a low angle and then descended with little twisting to the cell level.

Burrows at both sites descended in a meandering fashion and were often partly soil filled above but open farther down. All cells were essentially horizontal with an evenly curved posterior end and a front end that was consistently closed with a deeply recessed spiral closure providing a recessed, thimble-shaped space at the front of the cell (figs. 36, 37). It is in this recess (with a maximum diameter of a cell entrance and a lateral depth of 3–5 mm in several cases) where the defecating larva deposits its feces (fig. 38) and not at the cell's posterior end as in all other known calliopsines.

Cell dimensions in table 1 are presented separately for the Amaichá del Valle and Ticucho sites; cell lengths were measured from the posterior end of the cell to its front end but not including the open space created by the deeply concave closure in the entrance tunnel. Dimensions of food spheres for each site are also given; one presumably earlier supply was shaped as an uncoated sphere 2.8 mm in diameter. All completed provisions were approximately spherical and thickly coated with a clear, shiny, waterproof material.

The number of cells associated with a nest is uncertain. Only one was clearly found associated with a single burrow, but there is a possibility that a partly open burrow is not a main burrow but a long lateral and that each long lateral is filled after cell closure, and so goes undetected. This hypothesis was suggested by several nests of Spinoliella maculata found later in sandy soil near Valparaiso, Chile. This confusion can be resolved by monitoring a single nest under construction over a number of days.

Two slender, elongate, parallel-sided eggs (see table 1 for statistics) were observed at Amaichá del Valle. Each was attached by its posterior end to the top of the food sphere in the sagittal plane of the cell while the front end, pointing toward the cell front, was well above the food surface. They were curved, whitish, and possessed a clear, shiny chorion. Possibly single attachments of panurgine eggs to provisions are seen with freshly deposited eggs. Later, as the embryo matures, the front end of the egg descends to contact the surface of the provisions, as has been suggested by other panurgines where both fresh and later ovipositions have been detected. This needs confirmation through sequential observations on a single egg during development.

Also at the Amaichá del Valle site several older immatures were encountered including a large larva on its dorsum with its front end close to the cell entrance while it fed on provisions, a fully fed predefecating larva also on it dorsum with its posterior end directed toward the cell entrance, and a pupa again on its dorsum and its metasoma directed toward the feces-filled entrance chamber.³ Finally at the Ticucho site, two postdefecating larvae were found on their dorsa with their front ends directed toward the rear of the cell, their posterior ends pointed to the front, and the two entrance chambers were filled with feces. This series of observations confirmed that this species, like Callonychium petuniae, below, but unlike any other known panurgine, deposits its fecal material at the front of the brood chamber rather than at the rear.

Another remarkable feature of this species was the presence of a clear yellowish to ambercolored liquid detected between the paired prothoracic dorsal tubercles and the mesothoracic tubercles after larvae were finished feeding but not before. The source of this liquid and the role (if any) that it plays in the biology of this bee is unknown, but it has also been detected in the mature larva of Callonychium (Paranychium) minutum from the same locality: see Biology of Callonychium (Paranychium) minutum, below.

This species was first studied at Vila Velha, Brazil, in 1971 and again in 2002. Gabriel A.R. Melo (e-mail: XII-14-12) stated that it is the only known species of Callonychium at this locality.

DIAGNOSIS: See Diagnosis of Callonychium flaviventre and of Spinoliella herbsti.

MATERIAL STUDIED: Ten mature larvae: Brazil: Paraná: Vila Velha, XI-5-71 and XI-21-2002 (J.G. Rozen).

BIOLOGICAL NOTES: The nesting biology of Callonychium petuniae was studied on two widely separated occasions at Vila Velha, Paraná, Brazil., the first on November 5, 1971, and again on November 23, 2002, although most of the following information came from the initial study. At that time the nesting site was along the side of one of the numerous sandy walkways that lead through the state park. The soil was composed of moderately coarse, even-grained sand with no rocks and only a few small roots; dry on the surface, it was moist below 3 cm. The site was exposed to the sun during most of the day, was well elevated, contained a few scattered plants, and had a gently sloping surface. About 20–25 nests occurred within an area of 2 × 1 m. The same site was used for nesting by a species of Lasioglossum (Dialictus) (Halictidae). Neither in 1971 nor in 2002 was the pollen source identified at the site, although the species is known to be oligolectic on Calibrachoa (Solanaceae), a genus closely related to Petunia and perhaps also on Petunia (e-mail: G.A.R. Melo, XII-12-2012).

Nest entrances, irregularly spaced, were marked by conspicuous tumuli, 1.5–2.0 cm in diameter and as much as 1 cm in elevation. Most entrances were obscured by the tumuli but could be detected as eccentric depressions near the edges of tumuli. Pollen-laden females were occasionally seen returning from foraging, briefly landing and then observed “swimming” through tumuli to gain access to nest entrances. Occasionally the burrow entrance was exposed when the tumulus was on the downhill side. Among 10 nests excavated, all main burrows descended usually at about a 45° angle in a meandering fashion. Main tunnels (see table 1 for nest statistics) were filled at least at the surface but may have been only partly clogged with loose soil below. Nests consisted of a number of single cells connected to the main burrow by laterals 1 to 2 cm in length and 2.0 mm in diameter that were soil filled after cell completion. Cells had broadly rounded posterior ends like those of Callonychium flaviventre (fig. 36).

All cells were nearly horizontal and arranged singly. Provisions, coated with a conspicuous, clear, shiny material, were often only roughly spherical, as follows: 2.1 × 2.25 mm; 2.25 × 2.4 mm; 2.25 mm (spherical); and 2.6 mm (spherical), hence contrasting with the uniform spherical shapes of provisions of other calliopsine genera. Cell walls had a special lining somewhat darker than the surrounding sand but also had a coarse surface texture due to the coarseness of the sand. Cells were closed with earthen plugs, which in 1971 could not be observed because all four cell closures found were obscured by feces, but one was a deeply recessed spiral on a single preserved cell collected in 2002 (fig. 39). Early provision loads were shaped as a small, roughly spherical ball that was mealy moist as were the completed provisions on the inside. However, completed provision spheres were coated with a conspicuous, thick, shiny coating.

Two slightly curved eggs (table 1) were transparent white in color and had smooth shiny chorions. One was found attached by only its rear end to the provisions while its front end was elevated and directed toward the cell closure. In another cell a young larva sat on the top of the provisions and ate the provisions immediately below its head. One postdefecating larva rested on its dorsum with its anterior end toward the evenly round posterior end of the cell and with its posterior end directed toward the feces-filled cell closure.

Although observations were brief at Vila Velha, they suggested that females were active only the morning and very early in the afternoon. They were noted flying into the nesting site at 10 A.M., but by 1 P.M. few returning females were observed. The nesting site was not infested with parasitic bees, and furthermore bombyliids and triungulin larvae of meloids were not commonly observed in the area.

DIAGNOSIS: See Diagnosis of Callonychium flaviventre.

MATERIAL STUDIED: Seven postdefecating larvae: Argentina: Tucumán Province: Amaichá del Valle, collected as predefecating larvae III-23-1990; preserved IV-12-90 (J.G. Rozen, A. Roig-Alsina).

BIOLOGICAL NOTES: Observations on nests and nesting of Callonychium (Paranychium) minutum were undertaken at four localities in Argentina. The first two were in Catamarca Province, the first at Copacabana on November 28, 1989, and then at Belén on November 30, 1989. The following year, nests of this species as well as of C. (C.) flaviventre were discovered along the edge of a slightly sloping field at Amaichá del Valle, Tucumán Province, and studied on March 11, 20, and 23, 1990. The final site, 10 km west of Media Agua, San Juan Province, was studied October 30, 1991.

At all sites main tunnels were open, nearly vertical burrows usually with an abundant tumulus (see table 1 for nest statistics). Almost certainly each nest was constructed by a single female and possibly consisted of one cell per nest, although one nest seemed to have had two cells. Cells were oriented horizontally, without a visible special lining. Their shape was noteworthy in that they were elongate with an extended posterior end that, instead of being evenly curved as in most other panurgines, formed an acutely rounded terminus (figs. 40–42), which accommodated the fecal material voided by the mature larva. The inner surface of the cell closure was a rough, deeply concave spiral. Provisions (table 1) were spherical and coated with a clear, shiny surface; early loads were also spherical but substantially smaller and uncoated. Shiny, white, curved eggs were attached by their rear ends to the top surface of the provisions, while the anterior ends were raised above the provisions. A postdefecating larva found resting on its dorsum with head pointed toward cell closure and rear end directed toward the feces at the posterior end of the cell had clear yellowish to ambercolored liquid between the prothoracic and mesothoracic paired dorsal tubercles, as also noticed in the case of Callonychium (C.) flaviventre. The source and nature of this material is unknown. Males and females fly in copulo, often while females carry pollen on their hind legs.

The cleptoparasitic bee Caenoprosopina holmbergi Roig-Alsina (Apidae: Nomadinae: Caenoprosopidini) was collected in association with Callonychium (Paranychium) minutum at all of the sites mentioned above except for Media Agua, and its mature larvae were recovered and described from nests at Amaichä del Valle (Rozen and Roig-Alsina, 1991).

Luisa Ruz (e-mail: 1-07-2013) after examining specimens reported that adults of this bee represent an undescribed species in the subgenus Paranychium close to Callonychium (P.) minutum.

DIAGNOSIS: The mature larva of this species cannot be easily distinguished from that of Callonychium (Paranychium) minutum, although the paired dorsal tubercles (fig. 24) are not quite as attenuated and evenly tapering as those (fig. 23) of C. minutum.

MATERIAL STUDIED: Ten mature larvae: Argentina: Jujuy Province: Tumbaya, XI-18, 19-1989 (J.G. Rozen, A. Roig-Alsina).

BIOLOGICAL NOTES: This bee was discovered in copulo on blossoms of Prosopis (Fabaceae) and Heliotropium (Boraginaceae) and on the ground at Tumbaya, Jujuy Province, on November 18, 1989, and studied then and the following day. Nests were broadly scattered and those excavated on the first day were unusually deep and in a dry substrate (in table 1, top row refers to the dry substrate and bottom row, to moist substrate). In contrast those studied the next day 100 m away were in a very moist substrate and occurred at a depth of 4 cm, making nests of the species at this site the shallowest in this study. The dry area was removed from a source of water and the soil was sandy. The moist area was at the bottom of an artificial channel next to a river. In both situations, each nest was constructed by a single female, and the main tunnels were open and of about the same diameter. Cells were arranged singly at the end of short laterals (ca. 1–3 cm long) that were soil filled after cell closure. Closures were deeply concave spirals on the inner surface. Burrows descended at about a 45° angle in the dry substrate, but the rate of descent seemed more variable under the moist condition. In both situations, cell orientation was the same (approximately horizontal); cell shape and size did not vary. Cells were narrowed at the posterior end (fig. 43), into which larval feces were deposited rather than on the posterior ceiling of the cell. In lateral outline cells seemed about as pointed at the front end as at the back end. Because copulating individuals were observed on both Prosopis and Heliotropium, possibly both plant species may be the larval food source. No cleptoparasites were associated with either of the nesting sites.

The most noteworthy discovery with this species was the substantial difference in cell depth that appears to be determined by the ground moisture level. Does this suggest cell depth is an adjustment for internal cell humidity, evidence that excavating moist substrate is more physically demanding, or…?

DIAGNOSIS: The antennal papilla of the larva of Arhysosage flava is distinctly shorter than that of any other known larva in the Spinoliella clade and is slightly shorter than its own basal diameter. Furthermore, instead of having a smooth ventral apical edge to its mandible like others in the clade, its ventral apical mandibular edge is finely serrate. The paired dorsal mesothoracic tubercles are slender with the long axis of each directed more laterad (fig. 18) while its extreme apex curves posteriorly (fig. 17). The paired dorsal metathoracic and larger (more anterior) paired dorsal abdominal tubercles taper apically and have scattered short setae that arise from very small, inconspicuous swellings.

MATERIAL STUDIED: TWO predefecating larvae: Argentina: Salta Province: Cruz Quemada, 40 km S. GRL. Güemes, XI-10-1989 (J.G. Rozen).

BIOLOGICAL NOTES: The nesting site of Arhysosage flava and another, darker species, probably A. ochracea (Friese), was discovered at Cruz Quemada, 40 km south of General Güemes, Salta Province, Argentina, on November 10, 1989. The site of approximately 10 nests scattered over 2 m² of horizontal ground surface was first identified when an adult Caenoprosopis cleptoparasite was noticed sitting on a tumulus and matings of A. flava were observed in and around yellow flowers of the pollen plant, a low-growing cactus similar to or actually Opuntia (Cactaceae). Males flew rapidly from one cactus flower to another, alighted, and seemed to wait for females there. While in copulo, a male was seen to hold a female with his mandibles clasping her waist while she continued gathering pollen, a behavior also reported by Schlindwein and Wittmann (1995) for A. cactorum Moure. After being observed for a minute, they flew off in copulo, not unlike the mating behavior of females of Calliopsis (Nomadopsis) (Rozen 1958) except in the latter spiculate volsellae, not mandibles, seem involved with attachment. Nests appeared in early stages of construction; none yet had cells.

When the site was again visited on November 20, 1989, four nests of Arhysosage flava (see table 1 for statistics) were excavated. All were moderately shallow with cells occurring singly (i.e., not in linear series) short distances away (ca. 1–2 mm) from meandering, open main burrows. Cells, horizontal to sloping backward 10°, had rather evenly curved, perhaps only slightly narrowed posterior ends (fig. 44), and cell walls were fine grained and nonwaterproof when tested. However, importantly, the cell floor possessed a water-retardant area 5 mm long and 4 mm wide on its lower surface; see Discussion. Provisions were formed into a sphere, on top of which eggs were placed and young larvae ate; predefecating larvae were found resting on their dorsa on the cell floor with their heads pointed toward cell closures. Provisions were thickly coated with a clear, waterproof material.

The cleptoparasitic bee Caenoprosopis crabronina Holmberg was found attacking nests of Arhysosage flava at this site (Rozen and Roig-Alsina, 1991). Figure 44 shows not only the shape of the rear of the cell, but also the egg insertion hole of the cleptoparasite, which embedded her egg in the cell wall, so that only the anterior end is seen (fig. 45) along with marks created by the host female where she excised a cleptoparasite egg (fig. 46) presumably with her mandibles.