Chelostoma, Latreille, 1809

Rozen, Jerome G. & Praz, Christophe J., 2016, Mature Larvae and Nesting Biologies of Bees Currently Assigned to the Osmiini (Apoidea: Megachilidae), American Museum Novitates 2016 (3864), pp. 1-48 : 34-38

CHELOSTOMA

Nests of Chelostoma (Gyrodromella) rapunculi (Lepeletier) and C. (Chelostoma) florisomne (Linnaeus) are treated together because of many shared similarities, and both, from The Botanical Gardens of the City of Neuchâtel, were studied together by J.G.R and C.J.P. Nests of these species have been extensively described before ( Westrich, 1989, and references therein); we briefly redescribe them here to precise some aspects related to cocoon spinning, structure, and function. Nests of C. rapunculi were in slender bamboo sections (outside diameter 5–10 mm) from which C.J.P. collected larvae. They contained numerous cells in uninterrupted linear series, end to end (figs. 44, 51). Those of C. florisomne were found in bamboo sections that had slightly larger diameters. Although some cells of this species were in uninterrupted linear series, others were in series in which cells were separated by open spaces somewhat shorter than cell length (fig. 44) (i.e., these are called “intercalary cells” by Krombein, 1967). Cell partitions of both species (figs. 43, 50) were dark soil that appeared to be identical in color to the

12 Reexamination of the cocoon of Haetosmia vechti originally described in Gotlieb et al. (2014) shows that the front end consists only of a single silk layer with fenestrations more or less centrally located where silk touches the rough inner surface of the cell closure (new information), although most of the cocoon’s front end does not contact this surface.

thicker nest closure material but lacked the small, whitish pebbles that the nest maker added to the closure on sealing the completed nest (figs. 41, 48). Almost certainly, cell partitions and nest closure when first constructed are made of mud, perhaps moistened with nectar. The tunnel between the last cell constructed and nest closure (i.e., the “vestibular cell” by Krombein, 1967) was open. Accounts of nests of species of Chelostoma offered by Krombein (1967), Westrich (1989), Müller et al. (1997), and Rozen and Go (2015) provide similar descriptions. Cocoons of C. florisomne (figs. 34–38, 45, 46) and C. rapunculi (figs. 39, 40, 52, 53) are structured and function like the cocoon of Ochreriades fasciatus (Friese) , as described by Rozen et al. (2015). After initiating defecation but while still defecating, the larva deposits a thin nearly clear layer of silk that covers over

FIGURES 54, 55. Heriades truncorum . 54. Nest con- and adheres to the front end of the cell closure taining three cells in trap nest; note orange color of including any fecal pellets that happened to be feces contrasting with black feces of Chelostoma and deposited there (figs. 34, 36). (In the case of partly resin-filled tunnel leading to cells. 55. Close-

up of cell (front end left) showing clear cell partition Heriades carinata, Matthews, 1965 , referred to

of resin (arrow). this layer as the “operimentum”). The inner surface of this layer, though appearing reflective in normal light, is covered with a fine webbing of silk strands (fig. 35). The larva starts spinning the body of the cocoon and attaches the inner layer of the front of the cocoons to this webbing. As a final step in cocoon construction, the larva applies a thin but completely airtight final layer of silk to the inner surface of the entire cocoon except for a cottony mass of silk fibers (figs. 46, 53) centered at the front of the cocoon. This mass, termed the “air-exchange portal” ( Rozen et al., 2015), is thick, but the fibers are not fused and consequently there is an exchange of air between the inside of the cocoon and the space between the two cocoon layers forming the cocoon front (fig. 29). At the periphery of the front end, the inner and outer layers are only loosely attached (figs. 34, 36) with the result that air exchange takes place between the outside of the cocoon and the space between the two front layers of the cocoon. Thus, univoltine larvae are provided with air for their ca. 10 month diapause before emerging the following year, and simultaneously they are protected from parasites and the cell humidity is controlled by the airtight, waterproof cocoon wall. Feces of both species were black or nearly so (figs. 42, 47, 51).

The availability of unmodified surfaces of cell partitions of C. florisomne permitted observations obscured by cocoon silk in other species. Partitions of this species, presumably made of mud, dry quickly, each forming a hard wafer that is slightly concave on the surface directed toward the nest closure and slightly convex on the other side (fig. 47). Both surfaces are rather uneven with the rear surface more or less nodular (fig. 43), so that partition thickness varies from point to point. At minimum, the thinness is a remarkable: 0.3 mm. When J.G.R. visited Neuchâtel in early October 2014, most of the cells of this species were occupied by well-pigmented pupae contrasting with cells of C. rapunculi containing white larvae, suggesting a noteworthy lack of developmental synchronization of the two related species, probably related to the fact that in C. florisomne , the pupa overwinters ( Westrich, 1989), while in C. rapunculi , the larva overwinters. Chelostoma florisomne is active already in April, while C. rapunculi flies late May and June.