Eremapis, PARVULA OGLOBLIN

Rozen, Jerome G., 2011, Immatures of Exomalopsine Bees with Notes on Nesting Biology and a Tribal Key to Mature Larvae of Noncorbiculate, Nonparasitic Apinae (Hymenoptera: Apidae), American Museum Novitates 2011 (3726), pp. 1-52 : 30-32

EREMAPIS PARVULA OGLOBLIN

Table 2

Neff (1984) provided extensive information on nesting habits of this species. The following observations supplement his original account.

I investigated two nesting sites of Eremapis parvula . One at Pampa Vieja near San José de Jáchal, San Juan Province, Argentina, I discovered on November 2, 1991, and took brief notes on one active nest and vacated cocoons. I visited it again on November 28, 1993, when I carried out a more thorough observation. Although females were carrying pollen loads into nests, no new cells were encountered, but numerous cells whether vacated or with mature larvae gave insight into cocoon construction and phenology. The other site, at Guandacol, 42 km southwest of Unión, La Rioja Province, Argentina, I discovered and examined on November 29, 1993, and the following day. Almost no vacated cells were encountered, suggesting that the site was newly established, but numerous cells containing feeding larvae were recovered, giving insight into provisioning and larval development. Both sites were densely populated .

The Pampa Vieja site was restricted to a nearly barren horizontal area 2–3 m wide and 8–10 m long, on the east side of a row of trees and bushes that bordered an irrigation ditch. Fully exposed to the sun during the morning, at midday the site was partly shaded by the trees, some of which were Prosopis , the known pollen source of Eremapis parvula . Although at first they were presumed to be the sole source, a low-growing species of Prosopis also occurred some distance away. The fact that males were collected there suggests that females may have foraged from that species as well.

Soil at both sites was fine grained with few inclusions. Analysis of soil from Pampa Vieja indicated a composition of 50% sand, 42% silt, 8% clay, 0.5% coarse fragments, with a texture of sandy loam/loam. There the soil was dry even at the cell level; moisture at Guandacol resulting from recent irrigation was evident.

NEST ARCHITECTURE: Nests at both sites were shallow, with cells occurring 5–10 cm below the surface. 6 At both, burrow entrances were open as were main tunnels, but laterals leading to cells were soil filled. Small concentric tumuli were also evident at Guandacol and at Pampa Vieja in 1991 but not in 1993, perhaps because there were fewer active nests. I suspected in 1991 that nests seemed to contain single females, as had already been reported by Neff (1984).

Main tunnels tended to descend vertically but then turned and meandered at the cell level, where many branched and perhaps anastomozed. Burrow walls were unlined and obviously (though not tested) water absorbent. Many cells were arranged singly but some were in linear series of 2. All were horizontal or tilted as much as 30°; diameter of entrance was 1.5 mm (N = 2). In lateral outline, cell shape was an elongate oval, somewhat more elongate than the one diagrammed by Neff (1986: figs. 2, 3), and cell closures were deeply concave spirals of about 3 coils to the radius, deeper than depicted by Neff (ibid.). Cell walls appeared unlined (unlike those of any other exomalopsine taxon described in this paper), as was also reported by Neff. They were tested with a water droplet, which was absorbed immediately. 7 See table 2 for other nest statistics.

PROVISIONING AND DEVELOPMENT: So far as known, the pollen sources of Eremapis parvula are several species of Prosopis (Fabaceae) . The reader is referred to the paper by Neff (1984), who offers considerable information concerning floral preferences and discusses the bands of conspicuous specialized setae on adult female metasomal sterna 2–5, which he refers to collectively as a “metasomal scopa.” These setae are uniformly arranged and directed, and each is apically curved and points posteriad. One wonders whether they serve as a comb to manipulate dry pollen found in the curved, long and long-branch, scopal setae of the hind basitarsus and tibia.

Provisions are shaped in the form of a loaf (figs. 49, 50) with a curved top on which the female deposits an egg in the sagittal plane of the cell. The loaf is positioned on the floor (i.e., lower cell wall) of the cell, but whether it touches the floor with only its front and back edges (as depicted by Neff (ibid.: fig. 2) or rests its full length on the floor (as indicated by notes taken in 1993) needs to be confirmed. Although a distinct “foot” on the lower front edge of the provisions was not evident, on some preserved masses that edge was faintly produced. Young larvae crawl over the food mass while they eat making visible grooves on the surface of the provisions.

6 A nesting aggregation of Eremapis parvula subsequently encountered but not studied at Basilio Nievas , San Juan Province, on November 10, 1998, contained cells with larvae and old cocoons at a depth of ca. 8 cm .

7 This test was performed on a sample that had been stored in the museum for 17 years. Although one might question whether the waterproof substance would not deteriorate over such a long period, other cells in this study, stored equally as long, had retained their hydrophobic nature when tested.

Intermediate-stage larvae have dorsal paired tubercles on most body segments, but these tubercles all but disappear later. Older larvae were encountered encircling the provisions so that the provisions no longer touched any part of the cell wall. These larvae had positioned themselves approximately in the equatorial plane of the cell (i.e., at right angle to the cell’s long axis) so that they touched the cell wall with most of their dorsal surface while they circled the provisions. At this stage a larva starts defecating while still feeding, excreting greenish-brown elongate fecal pellets. Because the larva crawls (presumably with remnants of its dorsal tubercles) in the equatorial plane of the cell, almost all of the feces are applied as a broad band circling the cell wall. This band when complete forms a dark grayish-green belt up to 2 mm wide composed of more or less parallel fecal pellets that circles the cell. Larvae subsequently applies a pale, thin cocoon (fig. 63) the like of which, because of fecal placement, has not been reported for any other known bee, as pointed out by Neff (1984). The cocoon fabric adheres tightly both to the inner surface of the fecal band and to the rear and front of the cell and inner surface of the closure. The soil is glued to and hides the silk from the outside, but the inner surface of the cocoon glistens, and with SEM examination is seen to be composed of one or more thin sheets of material with embedded strands of silk, as in Anthophorula completa .

Future studies of this bee should consider the possibility that its larval behavior functions to remove the stored provision as soon as possible from contact with the cell wall that has no waterproof lining. See commentary on this matter in Discussion of Exomalopsine Nesting Biology, below.

ADULT ACTIVITY: Although males of this species were encountered at both nesting sites, matings were not seen to take place there, an observation paralleling Neff ’s (1984). Males and females were observed at Pampa Vieja during the heat of the day. Late in the day at the Guandacol site, numerous males and females were seen over a section of the nesting site, which was judged to be about a meter square. The following morning a Larrea bush densely shaded the same area, and little or no activity was observed to take place in the shade, although elsewhere in sunny portions males and females were observed.

PARASITISM: Cleptoparasitic bees were not found at nesting sites.