sign in sign up
<<
Home , Andrenidae, Apoidea, Colletes, Colletes compactus, Colletes inaequalis, Colletes kincaidii

Apidologie 39 (2008) 16–29 Available online at: c INRA/DIB-AGIB/ EDP Sciences, 2008 www.apidologie.org DOI: 10.1051/apido:2007049 Review article

Colletidae nesting biology (: )*

Eduardo A.B. Almeida1,2

1 Comstock Hall, Department of Entomology, Cornell University, Ithaca, NY, 14853, USA 2 Current address: Departamento de Zoologia, Universidade Federal do Paraná, Cx. Postal 19020, CEP: 81531-980, Curitiba, PR Brazil

Received 28 July 2007 – Revised 6 October 2007 – Accepted 10 October 2007

Abstract – are unique among for certain aspects of their nesting biology. In this review, attributes of colletid nesting are evaluated and discussed in light of a novel phylogenetic hypothesis for the . Some predictions made about evolution of certain traits, such as the cocoon-spinning behavior of Diphaglossinae, are confronted with phylogenetic evidence. The cellophane-like cell lining of Colletidae is a synapomorphy of this family, formed by polyester and characterized for being thick and strong, waterproof, and insoluble in different solvents. Historical developments towards the understanding of nature of the cell lining applied by colletids are summarized along with an account of diversity of some aspects of nesting of these bees. bee / nest / Colletidae / / Dufour’s gland

1. OVERVIEW Colletid bees have been long known for the special brood cell lining, which is commonly Nests are essential for the reproductive described as “cellophane-like” (e.g. Benoist, success of bees because they shelter the 1942; Cane, 1983; Michener, 2000). This cell brood, providing essential conditions for de- lining is commonly interpreted as a synapo- velopment of immatures. The majority of morphy of the family (Rozen, 1984; Michener, bee are solitary (Michener, 2000; 2000). Colletidae are also unique among bees Danforth, 2007) and most of them excavate a for their mouthpart morphology, in which the burrow containing brood cells in a substrate, glossa is broad, truncate, bilobed or bifid, and usually soil. Construction of brood cells is par- generally short (Fig. 1). Mouthparts of col- ticularly important and a generalized sequence letid bees have been considered to represent of steps when setting up cell walls comprise the most primitive morphology among bees, (Michener, 2000: 20): (1) lining the surface because the bifid glossa is shared with apoid with a smooth earthen layer, (2) tamping the (Apoidea: , , cell surface smooth with the pygidial plate, Heterogynaidae, and ). These wasps and (3) applying a secreted film of wax-like comprise the most closely related lineages or cellophane-like material to the earthen sur- to bees (Melo, 1999). Mouthpart morphology face. The earthen layer and the secreted lining and brood cell lining are linked to one another are likely synapomorphies of the bee clade, as in Colletidae because the female paints the lin- indicated by their absence in lineages of Hy- ing precursors onto the nest wall using her menoptera closely related to bees (Michener, broadly bifid brush-like glossa (Perkins, 1912; 2000). McGinley, 1980; Michener and Brooks, 1984; Michener, 1992, 2000). Corresponding author: E.A.B. Almeida, [email protected] So much is known about colletid nest ar- * Manuscript editor: Bryan Danforth chitecture that a comprehensive treatment of

Article published by EDP Sciences and available at http://www.apidologie.org or http://dx.doi.org/10.1051/apido:2007049 Colletidae nesting biology 17

considered a candidate for the first lineage in bee diversification (Michener, 2000: 83–87; Danforth et al., 2006b). Resemblance between colletid mouthpart morphology and that of apoid wasps has been taken as evidence that Colletidae are an early diverging branch of bees (the sister to the rest of the bees). As re- lationships between Colletidae and other bee families still is a contentious topic, primi- tiveness of mouthpart morphology remains questionable, as new sources of evidence are added to the debate (Alexander and Michener, 1995; Danforth et al., 2006a, b). Additional plesiomorphic morphological characters rein- force the hypothesis that Colletidae could be the earliest-diverging branch of the bee phy- Figure 1. Bilobed brush-like glossa of logeny, the living sister clade to all other basalis (Smith) (SEM photo, University of Kansas, bees (Michener and Brooks, 1984; Michener, by Robert W. Brooks; courtesy of Charles D. 2000). Michener’s monograph (1944) became Michener). a landmark in the context of contemporary ideas about bee evolution and classification. this topic would not be as satisfactory as the In that article, Michener presented evidence primary literature dealing with it (e.g. Claude- for the recognition of Colletidae, in particu- Joseph, 1926; Malyshev, 1927, 1968; Benoist, lar Paracolletinae, as the group with the most 1942; Janvier, 1933; Rayment, 1935, 1948; plesiomorphies among bees, strongly suggest- Michener and Lange, 1957; Michener, 1960, ing its primitiveness. Furthermore, Colleti- 2002; Eickwort, 1967; Rozen and Michener, dae are primarily distributed in austral con- 1968; Houston, 1969; Sakagami and Zucchi, tinents (, temperate , 1978; Batra, 1980; Roubik and Michener, and South Africa), with the exception of two 1984; Rozen, 1984; Torchio, 1984; Torchio widespread genera: and Hylaeus.This and Burwell, 1987; Torchio et al., 1988; disjunct distribution suggests antiquity, which Maynard and Burwell, 1994; Mader, 1999; could be explained by origin of this family in Michener and Rozen, 1999; Spessa et al., Gondwanaland during the Cretaceous. 2000; Packer, 2004). Malyshev (1935, 1968) The study by Alexander and Michener and Michener (1964, 2000) presented very (1995) cast doubts in regard to the mono- useful partial summaries of nest architecture phyly of Colletidae. However, a wealth of ev- in Colletidae, as well as of other bee families. idences from various sources currently sup- The goal of this paper is to explore some as- ports colletid monophyly (e.g. Michener and pects of the nesting biology of Colletidae. The Brooks, 1984; Michener, 2000: 127; Brady topic to receive the most extensive treatment and Danforth, 2004; Danforth et al., 2006a, b; will be the brood cell lining of colletid nests. Almeida, 2007). Part of this evidence comes The evolution of cocoon-spinning habit, social from the unique cell lining found in nests of behavior, and parasitism are also reviewed. colletid bees, which is discussed in detail in the sections below.

1.1. Bee phylogeny and Colletidae 1.2. Phylogenetic relationships within The history of early diversification of bees Colletidae and the relationships among its major lin- eages have long been filled with uncer- A phylogenetic study of Colletidae, which tainty; virtually every bee family has been included 122 colletid species representing 18 E.A.B. Almeida

Most commonly, Stenotritidae were regarded as either part of or closely related to An- drenidae (Oxaeinae) or Colletidae (particu- larly Diphaglossinae and Paracolletinae). Cur- rently, though, enough evidence invalidates hypotheses in which Stenotritidae are nested within Colletidae, derived from bionomical studies (e.g. Houston and Thorp, 1984), com- parative morphology (McGinley, 1980), and phylogenetic hypotheses based on morphol- ogy (Alexander and Michener, 1995) and on molecular data (Danforth et al., 2006a, b; Almeida, 2007). Most recent studies have Figure 2. Summary of the phylogenetic relation- placed Stenotritidae as sister to Colletidae ships among the main lineages of Colletidae and (e.g. Alexander and Michener, 1995 [implied Stenotritidae. Results based on the combined analy- weights analysis of their “Series II” char- sis of four nuclear protein-coding genes: EF-1alpha, acters]; Danforth et al., 2006a, b; Almeida, LW-rhodopsin, wingless and 28S rRNA (Almeida, 2007). 2007).

2. NESTING BIOLOGY all of its recognized subfamilies and tribes, AND EVOLUTION was conducted using molecular data from 2.1. Nest substrates four nuclear gene loci: elongation factor-1α (F2 copy), wingless, long-wavelength (LW) Colletidae can be divided into two large opsin, and 28S rRNA (Almeida, 2007). Re- groups based on substrate used for nesting. sults of Almeida (2007) are summarized in The first includes soil nesters and encom- Figure 2. Some important results to be high- passes all Diphaglossinae (e.g. Roubik and lighted here were: (1) the removal of the Michener, 1984; Rozen, 1984), Paracolletinae only African endemic , Scrapter, from (e.g. Michener and Lange, 1957; Michener, Colletinae and placement in a new subfam- 1960, 1964; Maynard and Burwell, 1994), and ily, Scrapterinae (as proposed by Melo and Scrapterinae (Rozen and Michener, 1968), and Gonçalves, 2005; see also Ascher and Engel in most Colletinae (e.g. Michener and Lange, Engel, 2005); (2) the recognition of Colletinae 1957; Batra, 1980; Mader, 1999; Michener, s.str. (i.e. Colletes and related South American 2000). Those subfamilies include mostly hairy genera) and Paracolletinae as independent taxa and generally robust bees. All species of (Silveira et al., 2002; Engel, 2005; Melo and Stenotritidae studied so far also nest in the Gonçalves, 2005); (3) Callomelitta,ifkeptas ground (Houston and Thorp, 1984; Houston, part of Paracolletinae, renders the latter para- 1975, 1984, 1987). phyletic. Therefore, Callomelitta is not treated The second group includes species re- here as part of any of the existing subfami- ported to nest inside stems, soft wood, pre- lies. Additional taxonomic adjustments are to existing cavities, as well as in soil. Included be undertaken later. in this group are the species of Callomelitta The present article will also take into ac- (Rayment, 1935), Hylaeinae, Euryglossinae, count available data for Stenotritidae, the clos- and Xeromelissinae. These bees are, on av- est extant relatives of colletid bees (Danforth erage, smaller, and less hairy. The species of et al., 2006a, b; Almeida, 2007). Stenotriti- Euryglossinae and Hylaeinae lack an external dae comprise 21 described species, all dis- -carrying because females trans- tributed in Australia. Throughout history of port pollen internally. bee classification, these enigmatic bees have Most species of Colletinae are ground been grouped with various other bee taxa. nesters, but at least one South American Colletidae nesting biology 19 species nests in pithy stems instead: Colletes from those of Hylaeus. This is quite significant rubicola Benoist (Benoist, 1942). Many but given a well-supported sister-group relation- not all species of Colletes have a quite peculiar ship between Xeromelissinae and Hylaeinae nest architecture compared to other ground- (Fig. 2). Ground-nesting species were more re- nesting bees of their and other families – cently documented for this subfamily: Geodis- the lateral burrows of some Colletes species celis megacephala Michener and Rozen, are subdivided into series of cells (Michener, Chilimelissa australis Toro and Moldenke 2000: 130, and references therein; for an ac- (Michener and Rozen, 1999; Packer, 2004), count of the diversity of nest architectures of and probably also G. longicephala Packer (as seven South American species of Colletes,see suggested by morphological adaptations of Claude-Joseph, 1926: 125–139). In contrast, bees of this species for sand nesting – Packer, lateral burrows of typical soil-nesting bee end 2005). in a single cell (Michener, 1964, 2000). Houston’s (1969) observations of various The linear cell arrangement observed in nests of Euryglossinae species illustrate the di- Colletes nests (Fig. 3a) resembles that of stem- versity of nesting habits for these bees. Some nesting bees (Fig. 3c). Moreover, Colletes groups were found to nest in the ground (Eu- have reduced basitibial and pygidial plates, ryglossasp., Euryglossula chalcosoma (Cock- two generally well-developed morphological erell), and Brachyhesma perlutea (Cockerell)), structures in soil-nesting bees (Batra, 1980: whereas others nest in wood (Euryglossina 525; Radchenko and Pesenko, 1996). These hypochroma Cockerell, Euryglossina pulchra morphological characters plus the nest archi- Exley, Pachyprosopis haematostoma Cock- tecture (in addition to the known case of stem- erell, and Pachyprosopis indicans Cockerell). nesting behavior) suggest the possibility that fasciatella (Cockerell) and Eury- extant species of Colletes descended from a glossa ephippiata Smith also nest in the soil stem-nesting ancestor. Regrettably, nest in- (Rayment, 1935, 1948, respectively). formation is not available for Hemicotelles, Hylaeinae seem to include mostly stem- Mourecotelles,andXanthocotelles,thethree nesting species (Michener, 2000), but some basal-most lineages of Colletinae (Almeida, species use abandoned nests of other , 2007). pre-existing cavities, volcanic rock, and earth Nests of Xeromelissinae are generally banks (Rayment, 1935; Sakagami and Zucchi, built in holes in hollow stems or beetle 1978; Michener, 2000; Daly and Magnacca, burrows in wood or stems, but some re- 2003). Examples of stem-nesting species in- cently described species are ground nesters clude Amphylaeus (Spessa et al., 2000) and (Michener and Rozen, 1999; Packer, 2004). Meroglossa (Michener, 1960). Hylaeus (Neso- All species of Chilicola whose nesting be- prosopis) includes both soil- and stem-nesters havior has been studied were found to nest (Daly and Magnacca, 2003). in woody substrates, i.e. stems, abandoned Michener (1964) compared nesting sub- beetle burrows, etc. (Packer, 2004: 818– strates used by different groups of bees and 819, and references therein). Claude-Joseph concluded that soil nesting is probably a ple- (1926) described nests of Chilicola inermis siomorphic character state for bee nesting. It (Friese) in hollow bamboo stems and of Chili- seems that soil nesting is the ancestral con- cola friesei (Ducke) occupying abandoned dition for the clade formed by Colletidae and stem nests of Manuelia (). Eickwort Stenotritidae, given the soil-nesting behavior (1967) studied nests of Chilicola ashmeadi of Stenotritidae, Diphaglossinae, and Para- (Crawford) and Michener (2002) described colletinae, and the phylogenetic relationships nests of Chilicola espeleticola Michener and within this group (Fig. 2). Chilicola styliventris (Friese). Stem-nests of Callomelitta perpicta Cockerell nests in de- Chilimelissa generally follow the same plan as caying wood, as reported by Rayment (1935: that of Chilicola (Michener, 1995). Michener 97); unfortunately he neither illustrated nor (1995: 33) pointed out that, in general, xe- described in detail the architecture of this romelissine nests do not differ conspicuously bee’s nests. The phylogenetic placement 20 E.A.B. Almeida

Figure 3. Cellophane-like cell lining of Colletidae. (a) Linear array of nest cells of Colletes sp. (Arizona, USA) in a single polyester tubular membrane (picture by James H. Cane). (b) Brood cell of Colletes validus filled with provision and an egg (picture by James H. Cane). (c) Prepupae of (Cockerell) viewed through the nest cell lining (picture by William Nye, USDA-ARS Bee Biology and Systematics Lab; courtesy of James H. Cane). of Callomelitta is not completely under- () (Cane et al., 1983a), and a few stood (Fig. 3), but this genus is part of bees do not line their cells at all (e.g. Hes- clade comprised of Colletinae, Euryglossinae, perapis and Dasypoda: Melittidae) (Rozen, Hylaeinae, Scrapterinae, and Xeromelissinae. 1987; Michener, 2000). The importance of cell It is possible that the common ancestor of this lining certainly has to do with reduction of clade was a stem-nesting bee, and multiple re- water exchange between the brood cells and versals to soil nesting (and to rotten wood) the environment (Michener, 1964, 2000; May, would then have taken place during the diver- 1972; Cane, 1983), and it may also be resistant sification of the group. to fungal hyphae (Albans et al., 1980). Prior to lining the brood cells, some bees apply a mandibular gland secretion that is very effec- 2.2. Brood cell lining tive both for fungistasis as well as for bacte- The application a waterproof lining on the riostasis (Cane et al., 1983b). brood cells wall is a synapomorphy of bees Populations of certain species of Colletes, and is probably associated with the pollen- e.g. C. halophilus Verhoeff, are often flooded feeding habits of bees (Michener, 1964, 2000). while overwintering but the cells are com- The origin of this behavior and the secretion pletely waterproof because of the lining along of the lining by bee’s glands are associated the cell walls and a flap that seals the opening with one another, as no species of apoid wasps (Albans et al., 1980). In some other species of was reported to do it (Hefetz et al., 1979; colletid and stenotritid bees, the main problem Espelie et al., 1992). The glandular lipoidal se- may be desiccation, and a waterproof lining cretion is sometimes replaced by oils collected should work just as well. in flowers, as in the case of species of Cen- Colletidae, in particular Colletes spp., be- tris (Apidae) (Vinson et al., 1997) or Macropis came well known for the apparently unique Colletidae nesting biology 21 lining of their nest cell walls. As early as 1742, Réaumur perceived the uncommon nature of this cell lining and described the multi-layered silky membrane of colletid brood cells (Batra, 1980). Colletid bees (and most other ground- nesting bees) generally apply a waterproof lin- ing to their brood cells but not to the nest Figure 4. Position of the Dufour’s gland and the tunnels. Nonetheless, the hylaeine Meroglossa poison sac in a female Colletes (redrawn from torrida (Smith) can line the entire cavity of Batra, 1980: Fig. 11). a twig with cellophane-like material, and this lining can even extend outside the nest en- trance (Michener, 1960). Hefetz et al. (1979) compared their chem- Cell linings of Colletes are insoluble in ei- ical nature of the contents of the Dufour’s ther aqueous or organic solvents, and are not glands and nest cell lining of three species of degraded by either acidic or basic hydroly- Colletes. The main chemicals found in both sis (Hefetz et al., 1979). Jakobi (1964) tried the gland secretion and the lining were macro- 11 different solvents on nest cell linings of di- cyclic ω-lactones, hydrocarbons, and aldehy- verse bees: Colletes was unique in that lining des. Macrocyclic ω-lactones are the precursors was not soluble in chloroform, but was soluble of lipid polyester of the nest cell lining, the in pyridin (unlike halictids and apids, among latter being composed of ω-hydroxy acid units others). Colletid bees employ their brush-like (Hefetz et al., 1979; Cane, 1981). Although the glossa as the main tool for applying the se- detection of macrocyclic lactones inside the cretion (e.g. Janvier, 1933; Malyshev, 1968; Dufour’s gland of Colletes had already been Batra, 1980). accomplished by Bergström (1974), the com- By the end of the 1970’s, behavioral and parative chemical analysis of nest cell lining morphological evidence suggested that the and gland secretion of the same bee species, cell-lining secretion was produced by the fe- side-by-side, was only attained later (Hefetz male Dufour’s gland, associated with the sting et al., 1979; Albans et al., 1980). The mem- apparatus (e.g. Batra, 1964, 1966, 1970; Lello, brane made of natural high molecular weight 1971). The Dufour’s gland is a blind sac that polyesters found in Colletes spp. was referred empties its secretions into the base of the sting to as “laminester” (Hefetz et al., 1979). (Fig. 4; e.g. Lello 1971; Batra, 1980; Cane, At a period when important progress was 1981; Duffield et al., 1984). This gland can made on the chemical nature of the Dufour’s be very large in bees that actively secrete gland contents and cell lining of colletid bees, and it may represent up to 10% of the live Batra (1980) conducted a detailed study of the weight of a Colletes bee (Duffield et al., 1984) nesting biology of three species of Colletes. and it may occupy 20–50% of the abdominal She described the process of cell lining appli- cavity of a Colletes (Batra, 1980). Dufour’s cation by Colletes spp. and substantiated the gland functions in Hymenoptera are diverse, evidence of the Dufour’s gland chief role in se- and include defense and alarm pheromones, creting the lining liquid (Batra, 1980). Accord- trail pheromone, host marking and discrimina- ing to her, the process consists of two phases: tion, and sexual attraction (Hefetz et al., 1978). “(1) imbibing the Dufour’s gland secretion Cane (1983: 658) provides a helpful histori- from the partly exserted sting; and (2) regur- cal account of research dealing with Dufour’s gitating the secretion while licking the cell” gland and its secretions. Lello conducted com- (Batra, 1980: 525). The ingestion of the secre- parative studies of the Dufour’s gland in many tion by the female bee happens very rapidly groups of bees (Lello, 1971 [for Colletidae], (0.3–0.5 s) when she “somersaults or reverses other papers by Lello cited in Lello, 1976; her position in the cell” (Batra, 1980: 525). see also Duffield et al., 1984: 403–414 for an Albans et al. (1980) and Torchio et al. (1988) anatomical and functional review of this gland also reported the deposition of a droplet of liq- in bees). uid from the tip of the metasoma. When the 22 E.A.B. Almeida bee applied the secretion to the cell wall, “the discovered the presence of macrocyclic lac- liquid dried and rapidly formed a continuous tones. Later, Espelie et al. (1992) found that membrane”, but it failed to dry and remained the nest cell lining of soluble in organic solvents when experimen- Morawitz consists of a mixture of a lipid tally applied to a glass plate (Albans et al., polyester and silk protein. Chemical analyses 1980). To explain this fact, Albans et al. (1980: of the cell linings of H. leptocephalus indi- 559) hypothesized that “[p]olymerization of cated that these bees use the same Dufour’s the Dufour’s gland secretion may be mediated gland secretion as the other colletids, but it is by an enzyme, which is probably secreted by complemented with silk (Espelie et al., 1992). the thoracic salivary glands”; but the chemical Interestingly, Torchio (1984) did not observe nature of the polymerization has not yet been Hylaeus to curl to acquire Dufour’s secretion completely understood. while in the nest. Batra (1980) and Torchio et al. (1988) give The difference in the relative sizes of the different accounts for the process of deposition Dufour’s gland and the thoracic salivary gland of secretions by the female Colletes. Accord- between Colletes and Hylaeus is noticeable for ing to Batra, salivary and Dufour’s secretions the species studied by Batra (1980: Fig. 11). are mixed in the crop (p. 525). According to The silk filaments produced by Hylaeus are Torchio et al. (1988: 608), these two secretions secreted by the salivary gland (also known as are separately applied onto the cell wall and the labial gland – e.g. Duffield et al., 1984) the polymerization occurs after the bee coats and its hypertrophy, as compared to that in the nest surface with the Dufour’s gland secre- Colletes, is compensated for by a not so-well tion on the top of a previously applied layer developed Dufour’s gland. Silk production is of salivary secretion. In Torchio et al.’s words: observed in a diverse array of Hymenoptera “[t]he (presumable) salivary liquid is applied (e.g. members of Chalcidoidea, , sparingly but during long periods of time, Apoidea: Melo, 1997 and references therein). whereas the (presumable) Dufour’s gland liq- Hylaeus species generally live inside stems uid is deposited in larger quantities but in- and that the production of Dufour’s gland se- frequently. . . the primary purpose of somer- cretion seems to be more closely associated saulting usually followed extended periods of with earthen nests. The shift to stem nesting salivary deposition when those liquids began may be related to silk secretion (Cane, 1983). solidifying directly on the mouthparts before In Stenotritidae, cells are coated with a thin, they were deposited on the nest wall. . . sali- waterproof membrane much like a plastic film vary material deposited by Colletes carries en- that is not readily separable from the earthen zymes that open and then cross-esterfy the wall (Houston, 1984; Houston and Thorp, lactones produced in the Dufour’s gland into 1984: 164–165); the membrane was insoluble the polyester layers found in Colletes cell lin- in organic solvents did not appear to be fibrous ings. These enzymes apparently remain active nor silky, and did not melt when heated. This well after salivary liquids harden and, as a lining was hypothesized by Houston (1984) consequence, Dufour’s gland liquids solidify to be made of polyester, as in Colletidae. only after contact with the salivary coating” No chemical analyses of either the stenotri- (Torchio et al., 1988: 622). tid Dufour’s gland secretions nor their cell lin- In addition to the investigation of nest- ing have been undertaken. So far, Colletidae ing behavior of Colletes spp., Batra (1980) are the only bees known to have representative also compared X-ray and infrared character- taxa whose brood cell walls are coated with istics of nest cell linings of Colletes and Hy- polyester derived from macrocyclic lactones. laeus. She concluded that they were only su- Rozen (1984) noticed that there is a phys- perficially similar, because nest membranes of ical difference between the lining of Dipha- Hylaeus appeared to be constituted by silk glossinae and of the remaining groups of Col- (Batra, 1980: 521; see also Torchio, 1984). letidae. In the Diphaglossinae (as well as in Duffield et al. (1980) studied the Dufour’s Stenotritidae), it is thinner, more fragile, and gland secretion of Say and closely connected to the substrate; in all other Colletidae nesting biology 23 colletid bees it is loosely attached to the sub- morphology for the former, indicating that nest strate (Rozen and Favreau, 1968; Roubik and construction imposes selective pressures on Michener, 1984; Rozen, 1984; Michener and glossal morphology (Michener and Brooks, Rozen, 1999). As Rozen (1984: 26) portrayed 1984). The role of colletid glossal morphology it: “[t]his lining seems to be homologous with for nest cell construction has long been appre- that of the cells of other colletid subfami- ciated (e.g. Kirby, 1802). lies, and probably is applied with the spe- Colletid bees fold their glossa longitudi- cialized glossae characteristic of all colletid nally in repose and, depending on the shape bees. However, there are no air spaces between of the glossal lobes, the glossa may resemble the lining and the cell wall nor are its lay- a pointed glossa, which may even have a ers separated by air spaces and fibrous strands, value for the bees. Claude-Joseph (1926: as seems to be typical of Colletes, Scrapter, Fig. 25, p. 146) illustrated a folded glossa Hylaeinae, and Xeromelissinae [multi-layered of Cadeguala occidentalis (Haliday). If there lining was first described for Colletes com- are morphological constraints to maintain the pactus Cresson by Rozen and Favreau, 1968]. broad colletid glossa to function for the appli- The lining of diphaglossine cells therefore cation of the nest lining, they may prevent the lacks the glistening, reflective appearance of lobes from being abnormally long. these other colletids and much more closely re- Batra (1972) studied halictids and apids sembles the “varnished” cell surface of other when lining their cells and called attention to families with a conspicuous, nonwaxlike lin- the use those bees make of their pygidial plate ing”. and penicillus. Colletes lack a penicillius and, instead, use their brush-like glossa to apply the cell lining (Batra, 1980). 2.3. Mouthparts and nesting

Apoid wasps and bees that possess a short 2.4. Provisions and truncate glossa face limitations in the ex- tent to which they can reach flower bases Most bees provision their brood cells with a when collecting (Krenn et al., 2005). mixture of pollen and nectar. The pollen con- Limitations can be overcome in a number tent of the food mass, relative to the amount of colletid bees by modification of different of nectar, determines the consistency of the mouthpart structures. There are no cases, how- mixture. Michener (1964) suggested that semi- ever, of a greatly elongate glossa in Colleti- solid provisions reduce the contact between dae (e.g. Michener, 2000; Krenn et al., 2005). the food ball and the cell wall, thus reduc- Diphaglossinae and some taxa of Paracol- ing points of contact with the nest wall and letinae (Glossopasiphae, Lonchopria [Porter- risk of fungal colonization. Firm provision apis], and Tetraglossula) are peculiar for their has been considered a plesiomorphy for bees elongate lateral glossal expansions (Michener, (Radchenko and Pesenko, 1996), and most 2000), and some species may have a basally groups of bees (with the exception of Col- longer glossa (e.g. capitus species letidae) predominantly have firm food masses. group – Houston, 1990), but it is never as ex- Stenotritids, for example, make a solid smooth treme as in the case of Apidae and Megachili- and moist uncoated pollen loaf of characteris- dae. Colletid mouthpart evolution apparently tic shape (Houston, 1984; Houston and Thorp, occurred under a constraint for a short and 1984). blunt glossa (Fig. 1), which is largely as- Colletidae are known for producing more sociated with painting the secreted linings liquid provisions than most other bees. A rea- onto the inner walls of the cells (McGinley, sonable assumption to make is that fluid pro- 1980; Michener, 1992, 2000; Krenn et al., visions can only be found in nests of which 2005). Indeed, comparisons of the glossa of brood cells are lined with a waterproof mem- cleptoparasitic bees (which never build nests) brane (like those of colletids), because liquid and pollen-collecting bees showed a simpler food masses would not be well preserved in 24 E.A.B. Almeida a chamber whose surface were permeable. As in extant groups of bees as the result of mul- Albans et al. (1980: 562) put it: “female col- tiple independent losses the cocoon-spinning letids are freed from the labor of shaping their behavior during bee evolution. A parsimo- pollen stores”. nious interpretation for the absence of cocoon Most colletid subfamilies have provisions spinning in all colletids except Diphaglossinae that vary from liquid to semi-liquid: Colletinae would be its single loss in a group formed by (e.g. Claude-Joseph, 1926; Malyshev, 1927; all subfamilies except Diphaglossinae (Rozen, Michener and Lange, 1957; Batra, 1980); 1984), which makes sense in light of the phy- Diphaglossinae (Janvier, 1933; Roberts, 1971; logenetic hypothesis shown in Figure 2. Roubik and Michener, 1984; Rozen, 1984; Brood cells have their opening sealed by an Torchio and Burwell, 1987), Euryglossi- operculum, which might serve for the same nae (Rayment, 1935: 27 – described provi- purpose as the leathery cocoon wall, i.e. pro- sions of Euhesma fasciatella (Cockerell) as tection against predators and nest parasites “pudding”; Michener, 1960; Houston, 1969), (Rozen, 1984). In Diphaglossinae, cells have Hylaeinae (Michener, 1960; Sakagami and a closure made of soil. The operculum is the Zucchi, 1978; Torchio, 1984), Xeromelissi- top of the cocoon (made of silk) that presum- nae (Claude-Joseph, 1926; Eickwort, 1967; ably permits exchange of gases and at the same Michener and Rozen, 1999; Packer, 2004). time might serve to exclude parasites (Rozen, Paracolletinae is the only subfamily for which 1984). semi-solid provisions are reported (Michener Michener and Lange (1957) reported hav- and Lange, 1957; Michener, 1960). How- ing found a cocoon-spinning species of Col- ever, Janvier (1933: Fig. 47, p. 328) described letes, but this must have been an erroneous ob- nests of Perditomorpha tristis (Spinola) and servation because Michener (2000: 128) stated based on his account of the provisions and that Diphaglossinae are the only colletids that the illustration of the nest, provisions of this retained this behavior. species appear to be semi-liquid and viscous. In Scrapter, the consistency of the provisions 2.6. Sociality was found to be intermediate between a semi- solid pollen ball (typical of Paracolletinae), No species of Colletidae are known to be and a semi-liquid mass as in most other col- eusocial (sensu Crespi and Yanega, 1995). letids (Rozen and Michener, 1968). The great majority of the species are soli- Roberts (1971) wondered how provisions of tary, as are most of the short-tongued bees. the diphaglossine guinnae Roberts All the , Melittidae, Stenotritidae, could have small amount of pollen and still and Rophitinae () are solitary. Eu- contain a high-enough protein content to be sociality, among short-tongued bees, evolved nutritious to the immatures. According to three independent times in Halictidae, a phe- Roberts (1971: 287), it is possible that yeast nomenon restricted to the subfamily Halictinae found fermenting the provisions might, in a (Brady et al., 2006). situation like that, substitute the ordinary pro- For Stenotritidae and Colletidae, it is com- tein source of bees (pollen). mon to observe nest aggregations, but com- munal nests and primitive levels of social- 2.5. Cocoon spinning ity are remarkably uncommon. Sakagami and Cocoon spinning is considered to be Zucchi (1978) first reported on a possibly so- a plesiomorphic characteristic among bees cial species of Colletidae. Hylaeus (Hylaeop- (Rozen, 1984; Radchenko and Pesenko, 1996; sis) tricolor (Schrottky) was found reusing Michener, 2000). Stenotritid larvae do not nests of apoid and vespid wasps in Brazil spin a cocoon (Houston, 1984; Houston (Sakagami and Zucchi, 1978). There were and Thorp, 1984) and, among Colletidae, more females in the nests studied than cells Diphaglossinae are the only known examples being provisioned, which along with other bio- of cocoon-spinning bees. Rozen (1984) inter- nomical characteristics allowed Sakagami and preted the presence/absence of this behavior Zucchi to propose that Hylaeus tricolor is Colletidae nesting biology 25 facultatively quasisocial (following the clas- eration the phylogenetic relationships within sification by Michener, 1974). Females of Colletidae, as illustrated in Figure 5. In light Meroglossa spp. (Hylaeinae) may co-exist in a of our knowledge, the polyester cell lining nest for sometime, but Michener (1960) found was present in the common ancestor of col- no strong evidence for cooperation among letid bees, a subsequent modification having them. happened in the clade formed by all colletids The second possibly parasocial colletid except Diphaglossinae. Cocoon-spinning was species was reported by Spessa et al. (2000). lost once, and a single transition from semi- The authors found sub-tropical populations of liquid to firm provisions happened in the evo- Amphylaeus morosus (Smith) in Australia to lution of Colletidae (Fig. 5). Hylaeinae are have multiple females sharing a nest. How- known to add silk to their cellophane-like cell ever it was not clear whether division of la- lining (Fig. 5), but future studies may indicate bor takes place. Relatedness between females that this trait is more widely distributed if it sharing a nest was too low to suggest kin se- is found to occur in other colletid lineages. lection, and ovary differentiation was not no- Further research is needed to understand the ticeable (Spessa et al., 2000). K. Hogendoorn evolution of nesting substrate preference in (unpublished, in litt.) studied two-female nests the clade comprising Callomelitta, Colletinae, and observed labor division in which one fe- Euryglossinae, Hylaeinae, Scrapterinae, and male assumed reproductive position. Egg lay- Xeromelissinae. It is possible that the shift ing would then be suppressed on the second fe- from soil-nesting to wood-nesting (especially male unless the first did not return to the nest. in stem) happened once (Fig. 5), followed by Parasitism and nest reutilization may consti- multiple reversals to soil nesting, but this is tute the main selective pressures for nest shar- not yet clear. Uncertainty about the evolu- ing (Spessa et al., 2000). tion of nest substrate preference will hopefully be dissipated as nesting habits are described 2.7. Cleptoparasitism for more taxa and phylogenetic relationships within colletid subfamilies are studied. About 20% of all bee species are clep- Detailed accounts were made for vari- toparasites (cuckoo bees): instead of building ous aspects of nesting behavior Colletes, Hy- their own nests, these bees lay eggs in brood laeus, and Diphaglossinae (e.g. Batra, 1980; cells built by other bees (Michener, 2000; Rozen, 1984; Torchio, 1984; Torchio et al., Danforth, 2007). Cleptoparasitism is estimated 1988). However, Xeromelissinae, Euryglossi- to have arisen over 25 independent times in nae, and Paracolletinae remain largely unex- bees and much of the diversity of cleptopar- plored, and no information exists about nests asites is concentrated in Apidae, Megachili- of Dissoglottini (Diphaglossinae). Very im- dae, and Halictidae (Danforth, 2007). As for portant chemical studies were conducted with Colletidae, only five species of Hylaeus (Ne- various lineages of Colletidae, most promi- soprosopis) from Hawaii are presumably clep- nently of Colletes and Hylaeus.TheAus- toparasites, based on circumstantial evidence tralian groups of Colletidae (Euryglossinae, (Daly and Magnacca, 2003), whose hosts are Australian Paracolletinae; various endemic other species of Hawaiian H. (Nesoprosopis). lineages of Hylaeinae), Xeromelissinae, and All other Colletidae and all species of Stenotri- Stenotritidae remain completely unexplored tidae are free-living, but many serve as hosts of chemically. Additionally, research of the bees of various tribes of Nomadinae and Isepe- chemical nature of the cell lining was con- olini (e.g. Rozen and Michener, 1968; Rozen, ducted for a small proportion of taxa for which 1984; Michener, 2000), as well as other Hy- the Dufour’s gland secretion was studied. menoptera (e.g. , ). As indicated in Cane (1983), the production of macrocyclic lactones to line nesting cells is 2.8. Concluding remarks not restricted to the Colletidae. Many of the Various topics discussed in this review ap- same molecules are possessed by representa- pear to make sense when taking into consid- tive species of the Halictidae (Halictinae and 26 E.A.B. Almeida

Figure 5. Phylogenetic relationships within Colletidae as a framework to understand the evolution of nest- ing behavior (see text for details). The transition from soil to wood nesting, if properly represented in this figure, was followed by multiple reversals to soil nesting in Colletinae, Euryglossinae, Hylaeinae, Scrapteri- nae, and Xeromelissinae (see Sects. 2.1 and 2.7 for further discussion).

Nomiinae) and Andrenidae (Oxaeinae), al- throughout the development of this research. I though they are uncommon molecules among am also thankful to Rick Harrison, Jim Liebherr, eukaryotes. Cell linings of these groups very Laurence Packer, Jerome Rozen, and two anony- much resemble those of the Diphaglossinae. mous referees for comments on this article. Jim The difference with Colletes and others seems Cane, Katja Hogendoorn, Terry Houston, Charles to lie with either (1) how the presumed poly- Michener, and Phil Torchio kindly answered var- merase enzyme works on the macrocyclic lac- ious queries. I am greatly indebted to Jim Cane tones, or (2) how the forming polymers are ex- and Charles Michener, for generously providing truded and stretched by the glossa (in the same pictures used to illustrate the cellophane-like lin- way that various plastics, such as polyethy- ing of colletid brood cells and the characteris- tic broad brush-like glossa of Colletidae. Luana lene, take on different physical characteris- Rodrigues kindly drew the figure showing the posi- tics depending on how they are extruded). tion of the Dufour’s gland. I am thankful to Gabriel These features constitute the derived evolu- Melo and Antonio Aguiar for advice with litera- tionary characteristics of colletid sub-groups, ture and assistance with preparing the figures. This and not the lactones themselves. It also points research was partly funded by Systematic Biology to a rewarding direction for future research. Grants to B.N. Danforth (DEB-0211701 and DEB- Chemical analyses of the cell linings of Dipha- 0412176). I am thankful to Fundação de Coorde- glossinae and Stenotritidae would greatly con- nação de Aperfeiçoamento de Pessoal de Nível Su- tribute for elucidating possible differences in perior (CAPES – Brazilian Ministry of Education) their chemical nature as compared to other lin- for a PhD scholarship. eages of Colletidae.

Biologie de la nidification chez les Colleti- ACKNOWLEDGEMENTS dae (Hymenoptera : Apoidea). I particularly grateful to Bryan Danforth for Colletidae / Stenotritidae / glande de Du- his constructive comments and encouragement four / nid / abeille Colletidae nesting biology 27

Zusammenfassung – Nestbiologie der Colle- Batra S.W.T. (1966) The life cycle and behavior of tidae (Hymenoptera: Apoidea). In vielen the primitively social bee, Lasioglossum zephyrum Aspekten ihrer Nestbiologie nehmen die (Halictidae), Univ. Kans. Sci. Bull. 46, 359–423. Colletidae eine Sonderstellung innerhalb der Batra S.W.T. (1970) Behavior of the alkali bee, Nomia melanderi, within the nest (Hymenoptera: Bienen ein. In diesem Übersichtsartikel wer- Halictidae), Ann. Entomol. Soc. Am. 63, 400– den Aspekte der Nestweise der Colletidae un- 406. ter Gesichtspunkten neuer phylogenetischer Batra S.W.T. (1972) Some properties of the nest– Hypothesen für diese Familie diskutiert. Ei- building secretions of Nomia, Anthophora, nige Vorhersagen zur Evolution verschiedener Hylaeus and other bees, J. Kans. Entomol. Soc. Merkmale, wie z.B. das Kokonspinnverhalten 45, 208–218. der Diphaglossinae, wird in Bezug gesetzt zu Batra S.W.T. (1980) Ecology, behavior, pheromones, parasites and management of the sympatric ver- phylogenetischen Befunden. Die Cellophan- nal bees , C. thoracicus and C. ähnliche Auskleidung der Brutzelle der Colle- validus, J. Kans. Entomol. Soc. 53, 509–538. tidae stellte eine Synapomorphie für diese Bie- Benoist R. (1942) Les hyménoptères qui habitent les nenfamilie dar. Diese Auskleidung ist eine tiges de ronce aux environs de Quito (Equateur), dicke, starke Polyestermembran, die wasser- Ann. Soc. Entomol. Fr. 111, 75–90. fest und auch in verschiedenen Lösungsmit- Bergström G. (1974) Studies on natural odoriferous teln unlöslich ist. Ältere Ansichten zum Ver- compounds X. Macrocyclic lactones in the Dufour ständnis der Natur dieser Zellauskleidung der gland secretion of the solitary bees L. and Halictus calceatus Scop. Colletidae und zur Diversität einiger Aspek- (Hymenoptera, Apidae), Chem. Scripta 5, 39–46. te der Nestweise finden sich in dieser Arbeit Brady S.G., Danforth B.N. (2004) Recent intron zusammengestellt. Colletidae sind beispiels- gain in elongation factor-1α of colletid bees weise einzigartig in Hinsicht auf die Verprovi- (Hymenoptera: Colletidae). Mol. Biol. Evol. 21, antierung der Brutzellen, die im allgemeinen 691–696. aus einer halb bis ganz flüssigen Futtermasse Brady S.G., Sipes S., Pearson A., Danforth B.N. besteht, im Unterschied zu den festen Pollen- (2006) Recent and simultaneous origins of euso- ciality in halictid bees, Proc. R. Soc. Lond., B, bällen, wie sie bei den meisten Bienen zu fin- Biol. Sci. 273, 1643–1649. den sind. Ausserdem gibt es bei den Colletidae Cane J.H. (1981) Dufour’s gland secretion in the cell einige Fälle von sozialem und kleptoparasiti- linings of bees (Hymenoptera: Apoidea), J. Chem. schem Verhalten, die hier ebenfalls diskutiert Ecol. 7, 403–410. werden. Cane J.H. (1983) Chemical evolution and chemosys- tematics of the Dufour’s gland secretions of Biene / Nest / Colletidae / Stenotritidae / Du- the lactone–producing bees (Hymenoptera: fourdrüse Colletidae, Halictidae, and Oxaeidae), Evolution 37, 657–674. Cane J.H., Eickwort G.C., Wesley F.R., Spielholz J. (1983a) Foraging, grooming and mate-seeking REFERENCES behaviors of Macropis nuda (Hymenoptera, Albans K.R., Alpin R.T., Brehcist J., Moore J.F., Melittidae) and use of Lysimachia ciliata O’Toole C. (1980) Dufour’s gland and its role (Primulaceae) oils in larval provisions and cell in secretion of cell lining in bees of the genus linings, Am. Midl. Nat. 110, 257–264. Colletes (Hymenoptera: Colletidae), J. Chem. Cane J.H., Gerdin S., Wife G. (1983b) Mandibular Ecol. 6, 549–564. gland secretions of solitary bees (Hymenoptera: Alexander B.A., Michener C.D. (1995) Phylogenetic Apoidea): Potential for nest disinfection, J. Kans. studies of the families of short-tongued bees Entomol. Soc. 56, 199–204. (Hymenoptera: Apoidea), Univ. Kans. Sci. Bull. Claude-Joseph F. [Janvier H.] (1926) Recherches 55, 377–424. biologiques sur les Hyménoptères de Chili Almeida E.A.B. (2007) Systematics and Biogeography (Melliféres), Ann. Sci. Nat. Zool. 10, 113–268. of Colletidae (Hymenoptera: Apoidea), Unpubl- Crespi B.J., Yanega D. (1995) The definition of euso- ished PhD Dissertation, Cornell University, Ithaca. ciality, Behav. Ecol. 6, 109–115. Batra S.W.T. (1964) Behavior of the social bee, Daly H.V., Magnacca K.N. (2003) Insects of Hawaii, Lasioglossum zephyrum, within the nest Vol. 17: Hawaiian Hylaeus (Nesoprosopis) Bees (Hymenoptera: Halictidae), Insectes Soc. 11, (Hymenoptera: Apoidea), University of Hawaii 159–186. Press, Honolulu. 28 E.A.B. Almeida

Danforth B.N. (2007) Bees, Curr. Biol. 17, R156– Houston T.F., Thorp R.W. (1984) Bionomics of the R161. bee Stenotritus greavesi and ethological character- Danforth B.N., Fang J., Sipes S. (2006a). Analysis of istics of Stenotritidae (Hymenoptera), Rec. West. family-level relationships in bees (Hymenoptera: Aust. Mus. 11, 375–385. Apiformes) using 28S and two previously un- Jakobi H. (1964) Über die Löshlichkeit des explored nuclear genes: CAD and RNA poly- Brutzellenwachses von Wildbienen, Z. merase II, Mol. Phylogenet. Evol. 39, 358–372. Bienenforschung 7, 72–76. Danforth B.N., Sipes S., Fang J., Brady S.G. (2006b) Janvier H. (1933) Étude biologique de quelques The history of early bee diversification based on hyménoptères du Chili, Ann. Sci. Nat. Zool. 10, five genes plus morphology, Proc. Natl. Acad. Sci. 209–356. USA 103, 15118–15123. Kirby W. (1802) Monographia Apun Angliae, Vol. 1, Duffield R.M., Fernandes A., McKay S., Wheeler J.W., J. Raw Publ. Ipswich, England. Snelling R.R. (1980) Chemistry of the exocrine Krenn H.W., Plant J.D., Szucsich N.U. (2005) secretions of Hylaeus modestus (Hymenoptera: Mouthparts of flower-visiting insects, Colletidae), Comp. Biochem. Physiol. B 67, 159– Struct. Dev. 34, 1–40. 162. ffi Lello E. de (1971) Adnexal glands of the sting Du eld R.M., Wheeler J.M., Eickwort G.C. (1984) apparatus of bees: Anatomy and histology, I Sociochemicals of bees, in: Bell W.J., Cardé R.T. (Hymenoptera: Colletidae and Andrenidae), J. (Eds.), Chemical Ecology of Insects, Chapman & Kans. Entomol. Soc. 44, 5–13. Hall, New York, pp. 387–428. Lello E. de (1976) Adnexal glands of the sting Eickwort G.C. (1967) Aspects of the biology of apparatus of bees: Anatomy and histology, V. Chilicola ashmeadi in Costa Rica (Hymenoptera: (Hymenoptera: Apidae), J. Kans. Entomol. Soc. Colletidae), J. Kans. Entomol. Soc. 40, 42–73. 49, 85–99. Engel M.S. (2005) Family-group names for bees Mader D. (1999) Geologische und biologische (Hymenoptera: Apoidea). Am. Mus. Novit. 3476, Entomoökologie der Rezenten Seidenbiene 1–33. Colletes, Vol. I, Logabook, Köln. Espelie K.E., Cane J.H., Himmelsbach D.S. (1992) Malyshev S.I. (1927) Lebengeschichte des Colletes cu- Nest cell lining of the solitary bee Hylaeus bisin- nicularius L., Z. Morphol. Ökol. Tiere 9, 390–409. uatus (Hymenoptera: Colletidae), Experientia 48, 414–416. Malyshev S.I. (1935) The nesting habits of solitary bees, Eos 11, 201–309, pls. III–XV. Hefetz A., Blum M.S., Eickwort G.C., Wheeler J.W. (1978) Chemistry of the Dufour’s gland secretion Malyshev S.I. (1968) Genesis of the Hymenoptera and of halictine bees, Comp. Biochem. Physiol. B 61, the Phases of Their Evolution, Methuen and Co., 129–132. London. Hefetz A., Fales H.M., Batra S.W.T. (1979) Natural May D.G.K. (1972) Water uptake during larval de- polyesters: Dufour’s gland macrocyclic lactones velopment of a sweat bee, Augochlora pura form brood cell laminesters in Colletes bees, (Hymenoptera: Apoidea), J. Kans. Entomol. Soc. Science 204, 415–417. 45, 439–449. Houston T.F. (1969) Observations of the nests and be- Maynard G.V., Burwell C.J. (1994) Notes on the nests of Leioproctus cristatus (Smith) (Hymenoptera: haviour of some euryglossine bees (Hymenoptera: Colletidae), Aust. Entomol. 21, 61–63. Colletidae), J. Aust. Entomol. Soc. 8, 1–10. McGinley R.J. (1980) Glossal morphology of the Houston T.F. (1975) Nests, behaviour and larvae of Colletidae and recognition of the Stenotritidae at the bee Stenotritus pubescens (Smith) and be- the family level, J. Kans. Entomol. Soc. 53, 539– haviour of some related species (Hymenoptera: 552. Apoidea: Stenotritinae), J. Aust. Entomol. Soc. 14, 145–154. Melo G.A.R. (1997) Silk glands in adult sphecid wasps (Hymenoptera, Sphecidae, ), J. Houston T.F. (1984) Biological observations of Hymenopt. Res. 6, 1–9. bees in the genus Ctenocolletes (Hymenoptera: Stenotritidae), Rec. West. Aust. Mus. 11, 153– Melo G.A.R. (1999) Phylogenetic relationships and 172. classification of the major lineages of Apoidea (Hymenoptera), with emphasis on the crabronid Houston T.F. (1987) A second contribution to the wasps, Sci. Pap. Univ. Kansas Nat. Hist. Mus. 14, biology of Ctenocolletes bees (Hymenoptera: 1–55. Apoidea: Stenotritidae), Rec. West. Aust. Mus. 13, 189–201. Melo G.A.R., Gonçalves R.B. (2005) Revised bee clas- sifications (Hymenoptera, Apoidea, Apidae sensu Houston T.F. (1990) Descriptions of new paracol- lato), Rev. Bras. Zool. 22, 153–159. letine bees associated with flowers of Eremophila (Hymenoptera: Colletidae), Rec. West. Aust. Mus. Michener C.D. (1944) Comparative external morphol- 14, 583–621. ogy, phylogeny, and a classification of the bees Colletidae nesting biology 29

(Hymenoptera), Bull. Am. Mus. Nat. Hist. 82, Réaumur R.A.F. de (1742) Mémoires pour Servir à 151–326. l’Histoire des Insectes, Vol. 6, Imprimerie Royale, Michener C.D. (1960) Notes on the behavior of Paris. Australian colletid bees, J. Kans. Entomol. Soc. Roberts R.B. (1971) Biology of the crepuscular bee 33, 22–31. Ptiloglossa guinnae n.sp. with notes on associated Michener C.D. (1964) Evolution of the nests of bees, bees, mites, and yeasts, J. Kans. Entomol. Soc. 44, Am. Zool. 4, 227–239. 283–294. Michener C.D. (1974) The Social Behavior of the Roubik D.W., Michener C.D. (1984) Nesting biol- Bees: A Comparative Study, The Belknap Press of ogy of Crawfordapis in Panama (Hymenoptera, Harvard University Press, Cambridge, MA. Colletidae), J. Kans. Entomol. Soc. 57, 662–671. Michener C.D. (1992) Sexual dimorphism in the glossa Rozen J.G. Jr. (1984) Nesting biology of diphaglossine of Colletidae, J. Kans. Entomol. Soc. 65, 1–9. bees (Hymenoptera: Colletidae), Am. Mus. Novit. 2786, 1–33. Michener C.D. (1995) A classification of the bees of the subfamily Xeromelissinae (Hymenoptera: Rozen J.G. Jr. (1987) Nesting biology and imma- Colletidae), J. Kans. Entomol. Soc. 68, 332–345. ture stages of a new species in the bee genus Hesperapis (Hymenoptera: Apoidea: Melittidae: Michener C.D. (2000) The Bees of the World, John Dasypodinae), Am. Mus. Novit. 2887, 1–13. Hopkins University Press, Baltimore. Rozen J.G. Jr., Favreau M.S. (1968) Biological notes Michener C.D. (2002) The bee genus Chilicola in the on compactus and its cuckoo tropical Andes, with observations on nesting biol- bee, pusillus (Hymenoptera: Colletidae ogy and a phylogenetic analysis of the subgenera and Anthophoridae), J. N.Y. Entomol. Soc. 76, (Hymenoptera: Colletidae, Xeromelissinae), Sci. 106–111. Pap. Univ. Kansas Nat. Hist. Mus. 26, 1–47. Rozen J.G. Jr., Michener C.D. (1968) The biology Michener C.D., Brooks R.W. (1984) A comparative of Scrapter and its , Pseudodichroa study of the glossae of bees (Apoidea), Contrib. (Hymenoptera: Colletidae and Anthophoridae), Am. Entomol. Inst. 22, 1–73. Am. Mus. Novit. 2335, 1–13. Michener C.D., Lange R.B. (1957) Observations on Sakagami S.F., Zucchi R. (1978) Nests of Hylaeus the ethology of some Brazilian colletid bees, J. (Hylaeopsis) tricolor: the first record of non– Kans. Entomol. Soc. 30, 72–80. solitary life in colletid bees, with notes in com- Michener C.D., Rozen J.G. Jr. (1999) A new munal and quasi–social colonies (Hymenoptera: ground–nesting genus of xeromelissine bees from Colletidae), J. Kans. Entomol. Soc. 51, 597–614. Argentina and the tribal classification of the Silveira F.A., Melo G.A.R., Almeida E.A.B. (2002) subfamily (Hymenoptera: Colletidae), Am. Mus. Abelhas Brasileiras – Sistemática e Identificação, Novit. 3281, 1–10. Published by the authors, Belo Horizonte, Brazil. Packer L. (2004) Taxonomic and behavioural notes on Spessa A., Schwarz M.P., Adams M. (2000) Patagonian Xeromelissinae with the description of Sociality in Amphylaeus morosus (Hymenoptera: a new species (Hymenoptera: Colletidae), J. Kans. Colletidae: Hylaeinae), Ann. Entomol. Soc. Am. Entomol. Soc. 77, 805–820. 93, 684–692. Packer L. (2005) A new species of Geodiscelis Torchio P.F. (1984) The nesting biology of Hylaeus (Hymenoptera: Colletidae: Xeromelissinae) from bisinuatus Forster and development of its imma- the Atacama Desert of Chile, J. Hymenopt. Res. ture forms (Hymenoptera: Colletidae), J. Kans. 14, 84–91. Entomol. Soc. 57, 276–297. Perkins R.C.L. (1912) Notes, with descriptions of Torchio P.F., Burwell B. (1987) Notes on the bi- new species, on aculeate Hymenoptera of the ology of Cadeguala occidentalis (Hymenoptera: Australian region, Ann. Mag. Nat. Hist. 8, 96–121. Colletidae), and a review of colletid pupae, Ann. Radchenko V.G., Pesenko Y.A. (1996) “Protobee” and Entomol. Soc. Am. 80, 781–789. its nests: A new hypothesis concerning the early Torchio P.F., Trostle G.E., Burdick D.J. (1988) The evolution of Apoidea (Hymenoptera), Entomol. nesting biology of Cockerell Rev. 75, 140–162. (Hymenoptera: Colletidae) and development of Rayment T. (1935) A Cluster of Bees: Sixty Essays its immature forms, Ann. Entomol. Soc. Am. 81, on the Life Histories of Australian Bees, with 605–625. Specific Descriptions of Over 100 New Species Vinson S.B., Williams H.J., Frankie G.W., Shrum G. and an Introduction by Professor E.F. Phillips, The (1997) Floral lipid chemistry of Byrsonima crassi- Endeavour Press, Sydney. folia (Malpigheaceae) and a use of floral lipids by Rayment T. (1948) Notes on remarkable wasps and Centris bees (Hymenoptera: Apidae), Biotropica bees, Aust. Zool. 11, 238–254. 29, 76–83.