Two nests of Hoplitis curvipes are described from Apulia (Italy) and Dagestan (Russia). Both nests consisted of two brood cells placed side by side under a stone. The cells were neither attached to each other nor to the substrate. They were constructed from leaf fragments, which were imbricately arranged, forming a cone-like structure; each leaf fragment consisted of a basal part that was masticated to leaf pulp and an apical part that protruded freely from the cell wall. The cell wall was formed by the fusion of the masticated basal parts of the leaf fragments and thus entirely consisted of leaf pulp. The cell was sealed with a closing plug made of pure leaf pulp; a few leaf fragments were glued to its outer surface. The cocoon consisted of two layers: the outer layer was restricted to the anterior portion of the cell and had several longitudinal air-exchange slits on its lateral surface, while the inner layer had an air-exchange orifice in its most anterior dome-shaped top. Results of measurements of brood cell dimensions and contents are provided. The nesting biology of species of the H. curvipes group is discussed.

Anthophila, Apiformes, bionomics, Caucasus, Gargano, Hoplitis mitis, megachilid bees, Palaearctic region

The genus Hoplitis Klug, 1807 is distributed in the Palaearctic, the Nearctic, and the Afrotropical region; a few species also occur in the Oriental region (Michener 2007). It is the largest genus of the osmiine bees (Hymenoptera, Megachilidae, Osmiini) with 389 species described so far (Müller 2023). The genus is especially diverse in the Palaearctic region, where 14 subgenera and 313 species occur (Praz et al. 2008; Ungricht et al. 2008; Sedivy et al. 2012; Müller 2023). The nesting biology of Hoplitis is extremely diverse and encompasses the whole diversity observed in the osmiine bees (Müller 2023). The 110 species, for which nests have been found so far, build their brood cells in self-excavated burrows in the ground or pithy stems, in various kinds of pre-existing cavities (such as hollow stems, insect burrows in wood or pithy stems, abandoned nests of other bees and wasps, rock and stone crevices, rarely galls or empty snail shells), or construct them freely on the surface of rocks or stones, usually in depressions; a few species are kleptoparasitic. Many species of Hoplitis build brood cells with complete constructed lateral walls. Other species only divide the nest cavity into cells with transverse partitions and seal it with a closing plug at or near its opening. The building material used for nest and cell construction is also very diverse; depending on the subgenus or species, mud and pebbles, leaf pulp, leaf fragments, petals or pith are used alone or in diverse combinations (Michener 2007; Sedivy et al. 2012; Müller 2023).

Alcidamea Cresson, 1864 is one of the largest subgenera of Hoplitis. It occurs in the Palaearctic and the Nearctic region; there are 81 described species, 64 of which occur in the Palaearctic (Michener 2007; Ungricht et al. 2008; Müller 2023). Most representatives of this subgenus nest above ground, mostly in self-excavated burrows in pithy stems or in pre-existing cavities. They usually use plant material for nest and cell construction, particularly leaf pulp, which is sometimes mixed with pith or sand (Müller 2023). One Nearctic species, Hoplitis biscutellae (Cockerell, 1897), is known to collect resin (Rust 1980). The Palaearctic species Hoplitis tuberculata (Nylander, 1848) uses significant amounts of small pebbles and soil particles, which are densely packed between partitions made of leaf pulp (Müller 2015). Bees of the subgenus Alcidamea usually do not construct lateral cell walls except for some species of the Hoplitis fulva group, which make complete cells of plant pulp in either self-excavated or pre-existing cavities in the ground (Marikovskaja 1968; Ivanov and Fateryga 2018).

The most unusual nesting habits are, however, known for Alcidamea species of the Hoplitis curvipes group, which contains five species. Nests of one of them, Hoplitis mitis (Nylander, 1852), have been described so far. This species nests below stones, in rock crevices, in grass tussocks, between dried leaves or in old cells of other bees. The brood cells, which are built singly or in small groups of up to 12, entirely consist of leaf fragments imbricately glued together, forming a cone-like structure. The cell closure is made of leaf pulp, which is occasionally reinforced by sand grains or leaf fragments (Maneval 1925; Koller and Hamann 1950; Bonelli 1967; Müller et al. 1997; Westrich 1989). Trophic relationships are known for four species of the H. curvipes group: H. mitis is oligolectic on Campanulaceae, particularly Campanula L. (Fig. 6G), H. curvipes (Morawitz, 1871) is oligolectic on Allium L. (Amaryllidaceae) (Fig. 1E), H. tricolor (Saunders, 1908) is probably oligolectic on Reseda Tourn. ex L. (Resedaceae), and H. epeoliformis (Ducke, 1899) is polylectic (Müller 2023).

Figure 1. 

Bionomics of Hoplitis curvipes A habitat in Dagestan B medium-sized stone, under which a female bee entered (arrow) C position of the nest under the stone after its removal D extracted nest consisting of two brood cells E male sleeping in an inflorescence of Allium rotundum s. l. F nest from Apulia after removal of covering stone consisting of two brood cells with the female entering a cell. Scale bar: 1 cm (D).

Hoplitis curvipes is known from Spain, France, Italy (including Sicily), Greece, Bulgaria, Russia (Dagestan), Azerbaijan, Turkey, and Syria (Müller 2023). In spite of its relatively large distribution area, nests of H. curvipes have not been described so far. The purpose of the present contribution is to report the nesting biology of this rare species based on two nests found in Italy and Russia.

Field observations were carried out in Apulia (Italy) in the vicinity of San Giovanni Rotondo (Monte Gargano, Province of Foggia, 41°42'44"N, 15°44'11"E, ca. 600 m a.s.l.) on 5 July 1994 and in Dagestan (Russia) in the vicinity of Talgi (foothills of the Greater Caucasus, Makhachkala urban okrug, 42°52'36"N, 47°26'42"E, ca. 270 m a.s.l., Fig. 1A) on 13 and 27 June 2021.

The nest from Dagestan was first recorded on 13 June when it was provisioned by the female bee. During the second visit on 27 June, the nest was completed and consisted of two brood cells, which were transported to the laboratory, where they were kept in outdoor conditions in the shade. In January 2022, the two cells were separated from each other, softened in a humid environment, and dismantled. Leaf fragments were detached from the cell walls, pressed between sheets of paper, and dried. They were measured with an ocular micrometer scale of an MBS-9 stereomicroscope and weighed with a precise torsion balance. The cells with the outer coverage of leaf fragments removed were subjected to longitudinal dissection. The cocoons with fecal pellets were removed from the cells. The thickness and dimensions of cell walls and cocoons were measured. The cell walls, the cocoons with fecal pellets, and the prepupae found inside the cocoons were weighed. The prepupae were placed into glass tubes sealed with cotton plugs and kept under outdoor conditions. An emerged bee specimen was deposited in the collection of the V.I. Vernadsky Crimean Federal University.

To ascertain how much of the initial leaf fragments was masticated to leaf pulp, the following calculation was performed. We supposed that the initial average length of the leaf fragments of a brood cell consisted of the sum of the average length of the basal parts of the fragments, which had been masticated to leaf pulp (l₁), and the average length of the intact ends of the fragments (l₂). The total mass of the building material consisted of the mass of the cell wall consisting of leaf pulp (m₁) and the total mass of the removed intact ends of the leaf fragments (m₂). If all leaf fragments would be parallel-sided, l₁ could be calculated as m₁ multiplied by l₂ and divided by m₂. However, the leaf fragments were approximately 1.5 times broader at their base than at their apex due to apical narrowing. Therefore, the average length of the masticated basal part of a leaf fragment was calculated according to the following formula:

$$l_1=\frac{m_{1}x{l}_2}{1.5m_2}$$.

Photographs of the nest were taken with a Canon EOS RP and a Canon EOS Rebel T2i digital camera, a Sigma AF 105 mm f/2.8 and a Tamron SP AF 90 mm f/2.8 Di macro lens, and a Yongnuo YN-14EX macro flash.

Nest architecture and brood cell structure of Hoplitis curvipes were compared with those of H. mitis based on literature data (Maneval 1925; Koller and Hamann 1950; Bonelli 1967; Westrich 1989) and four nests discovered in the Swiss and Italian Alps (Zeneggen/Valais, 15 July 1990; Val Piora/Ticino, 30 July 1991; St. Pierre/Aosta Valley, 8 July 1996; Surses/Grisons, 10 July 2021).

The examination of four nests of Hoplitis mitis revealed close similarities, but also some differences compared to H. curvipes. The structure of the cells, which were 10–12 mm long and 8–9 mm wide, proved to be largely identical as revealed by the following characteristics: i) the cells of H. mitis were constructed from imbricately arranged leaf fragments, which formed a cone-like structure (Fig. 6A, B, D, E, F); ii) the leaf fragments consisted of a basal part that was chewed to leaf pulp and an apical part that protruded from the cell; the cell wall was formed by the fusion of the masticated basal parts of the leaf fragments and thus had a smooth inner surface (Fig. 6C); iii) the cells were sealed with a 1 mm thick plug made of leaf pulp, into which a few pieces of leaf fragments were incorporated by their chewed base; these fragments were directed in longitudinal direction to the cell and slightly curved inwards, so that the nest plug was no longer visible (Fig. 6B, D).

Figure 6. 

Bionomics of Hoplitis mitis A nest in dorsal view consisting of three brood cells, which were hidden among dead blades in a grass tussock (Surses, Grisons, Switzerland) B same nest as above in frontal view with one brood cell still being provisioned C inner surface of brood cell D nest consisting of three brood cells, which were hidden under a removed stone (Zeneggen, Valais, Switzerland) E single brood cell, which was hidden in dense plant litter (Val Piora, Ticino, Switzerland) F single brood cell attached to a stem (St. Pierre, Aosta Valley, Italy) G female collecting pollen on Campanula rotundifolia (Bräntschu, Valais, Switzerland).

In contrast, some nests of H. mitis differed from those of H. curvipes in the following characteristics: iv) the brood cells of H. mitis may adhere to each other (Fig. 6A, B, D) by smoothed-out leaf pulp from chewed leaf margins, whereas the brood cells in both discovered nests of H. curvipes lay freely side by side; v) while in most nests of H. mitis the brood cells did not adhere to the substrate as in H. curvipes, one brood cell of H. mitis was found attached to a plant stem (Fig. 6F); interestingly, in this cell the leaf fragments were more irregularly arranged and not all aligned longitudinally to the cell axis as is the case for the free-lying cells (Fig. 6A, B, D, E); this more irregular arrangement of the leaf fragments was probably due to the necessity to build part of the cell from the side rather than evenly from the posterior to the anterior end.

Our examination of the four nests of H. mitis confirmed the findings of other authors (Westrich 1989 and references therein) except for characteristic ii), which is described here for the first time.

In the present study, nest architecture and brood cell structure of Hoplitis curvipes are described for the first time. The nesting biology of H. curvipes is very similar to that of H. mitis, which also belongs to the H. curvipes species group of the subgenus Alcidamea and whose nesting biology has been described by several authors (Maneval 1925; Koller and Hamann 1950; Bonelli 1967; Müller et al. 1997; Westrich 1989). Both species construct the brood cells from leaf fragments, which are imbricately glued together resulting in a cone-like structure of the cell.

Both nests of H. curvipes discovered at two localities 2500 km apart were largely identical: they consisted of two brood cells, which lay freely side by side under a stone and adhered neither to each other nor to the surrounding substrate. Whether these characteristics are universal for H. curvipes across the species’ entire distribution range is unclear due to the small number of nests discovered so far. It may be possible that the species is more flexible with respect to nesting site or nest architecture as is the case for H. mitis, which shows an amazing variability in its nesting behaviour (Maneval 1925; Koller and Hamann 1950; Bonelli 1967; Müller et al. 1997; Westrich 1989; present study): i) H. mitis constructs its brood cells either in pre-existing cavities (below stones, rock crevices, abandoned cells of other bees), in vegetation (grass tussocks, plant litter) or attached to plant stems or in angles of pine crotches; ii) the nests contain a varying number of brood cells ranging from 1 to 12, which lie freely side by side or are (partly) attached to each other; iii) the leaf fragments for cell construction are collected on various plant taxa; and iv) the cell closure made from leaf pulp is occasionally reinforced by sand grains.

The nests of H. curvipes and H. mitis are not only unique among species of the subgenus Alcidamea, but also among all other osmiine bees. As reviewed in the Introduction, most other Alcidamea species nest in self-excavated or pre-existing cavities and use leaf pulp as building material. Representatives of other taxa of osmiine bees build their nests also from materials other than leaf pulp, such as mud and pebbles or petals. Whole petals or large petal pieces are used by Hoplitis species of the subgenus Anthocopa Lepeletier & Serville, 1825 and Osmia subgenus Tergosmia Warncke, 1988 (Rozen et al. 2010; Müller 2020). Although petals are quite similar to leaf fragments, they are applied in a different way for cell construction by Anthocopa and Tergosmia species than the leaf fragments in H. curvipes and H. mitis. Species of Anthocopa use the petals to merely line the walls of brood cell cavities and nest burrows, while species of Tergosmia construct three-layered cells consisting of two layers of petals that sandwich a thin middle layer of mud (Ivanov and Filatov 2008; Rozen et al. 2010; Müller 2020). The unusual structure of the brood cells of species of the H. curvipes group, which do not require the support of the surrounding substrate, enables the construction of nests in cavities of various sizes. One nest of H. mitis consisting of six loosely attached cells was even found in the angle of the lowest crotch of a 0.8 m tall pine tree (Koller and Hamann 1950).

Most probably, the ancestors of the H. curvipes species group constructed their brood cells from leaf pulp alone, but then evolved to leave the apical parts of the collected leaves unmasticated. This is consistent with the reconstructed phylogeny of the genus Hoplitis, which placed H. curvipes and H. mitis as members of the same clade amidst and not basal to species of Alcidamea, which use leaf pulp as nest building material (Sedivy et al. 2012). Within the family Megachilidae, the use of leaf fragments for brood cell construction is most typical for numerous species of the genus Megachile Latreille, 1802 (Megachilini), which, however, differ from H. curvipes and H. mitis by their habit to usually cut regularly circular to elliptical leaf pieces (Michener 2007; Ivanov and Zhidkov 2010; Gonzalez et al. 2019). Some Megachile species, however, masticate the margins of the cut leaf pieces in order to glue them together (Ivanov and Filatov 2008). As in the H. curvipes species group, the leaf-cutting behaviour of Megachile probably evolved from ancestors, which used leaf pulp as the main nest building material, but did not completely masticate the cut leaves (Michener 2007; Gonzalez et al. 2019).

Interestingly, the brood cells of species of the H. curvipes group are most similar to those of communal wasps of the Indo-Malayan genus Calligaster de Saussure, 1852 (Hymenoptera, Vespidae, Zethinae). These wasps construct aerial nests with several brood cells attached together side by side; the cells are cone-shaped and made from leaf fragments, which are imbricately arranged (Nugroho et al. 2016). The difference of these nests from those of the H. curvipes species group is that the free apical ends of the leaf fragments are always directed to the posterior end of the cell rather than to the anterior one. Thus, the nests of these wasps look like “inverted” cones, which is an interesting example of convergent evolution of a similar cell structure in solitary bees and vespid wasps.

The cocoon of H. curvipes has a structure which corresponds to the generalized scheme of the osmiine bee cocoon with an outer and an inner layer (Rozen and Praz 2016). The outer layer is restricted to the anterior portion of the cell and separated from the inner layer by a narrow space. The inner layer is mostly airtight except for an air-exchange portal at the anterior top. In H. curvipes, additional air-exchange slits occur at the lateral sides of the outer layer. In combination, these structures provide humidity control and air exchange and simultaneously serve as a barrier against parasites and predators (Rozen and Praz 2016). A similar structure of the cocoon (“cocoon with an arch”) is found in some chrysidid wasps of the genera Omalus Panzer, 1801, Pseudomalus Aschmead, 1902, and Chrysellampus Semenov, 1932 (Hymenoptera, Chrysididae) (Martynova 2020).

Comments by Laurence Packer on the first version of this paper led to improvements. The research was carried out within the state assignment of the Ministry of Science and Higher Education of the Russian Federation, No. 121032300023-7 (for A.F.).