Coelioxys (Allocoelioxys) coturnix, Perez, 1884

Rozen, Jerome G. & Kamel, Soliman M., 2009, Hospicidal Behavior of the Cleptoparasitic Wasp Sapyga luteomaculata And Investigation into Ontogenetic Changes in Its Larval Anatomy (Hymenoptera: Vespoidea: Sapygidae), American Museum Novitates 3644, pp. 1-24 : 1-3

(Allocoelioxys) coturnix Pérez , a cleptoparasitic bee that attacks nests of Megachile (Eutricharaea) minutissima Radoszkowski ,

we discovered females of the wasp Sapyga luteomaculata Pic parasitizing nests of the same host and also of Osmia (Pyrosmia) submicans Morawitz. During the course of the investigation, we recovered immature stages of this Sapyga from nests of both

1 American Museum of Natural History ([e-mail]).

2 Professor, Department of Plant Protection , Faculty of Agriculture, Suez Canal University, Ismailia, Egypt ([e-mail]) .

hosts 3 and made observations on its hospicidal behavior. We present here our understanding of the mode of cleptoparasitism of this species and other information concerning its natural history and, while doing so, compare this species with literature accounts of other Sapygidae . We also provide our analysis of the anatomy of its first and last larval instars and of the anatomical and behavioral changes that take place from one instar to the next. Throughout we compare the anatomical and behavioral similarities and differences between Sapyga luteomaculata and cleptoparasitic bees, which clearly show a strong evolutionary convergence.

Although there are a number of literature references identifying hosts of Sapyga (e.g., Krombein et al., 1979), the most complete investigation of its biology and immature stages to date was presented by Torchio (1972a, 1972b, 1979), pertaining to S. pumila Cresson. The reader is referred to those works for another understanding of the biology of an example of the genus. Malyshev (1968) also gave some biological information pertaining to Sapyga similis Fabricius. Although we were unable to obtain the work of Brechtel (1986), Westrich (1989) recounted much of the information in it, with reference to S. clavicornis (Linnaeus) and Sapygina decemguttata (Jurine) . The biologies of several species of Polochrum (Sapygidae) were offered by Hurd and Moure (1961; see references therein), Parker (1926), who also described the egg and the first and last instars of P. repandum Spinola, and Grandi (1962) , who treated and illustrated the mature larva of the same species.

Although we first detected the presence of Sapyga luteomaculata in September 2006, our field investigation into its biology was initiated in 2007 from April 24 to May 12, primarily on the campus of Suez Canal University, Ismailia, Egypt (N30 ° 37 9 10 0 E32 ° 15 9 58 0), where the second author (S.M.K.) has been

3 We also recovered a single last instar of Sapyga luteomaculata from a nest of an unknown megachilid at the Ismailia Experiment Station on May 12, 2008, that lines its cells only with pollen moistened with nectar. This seems to be evidence that host specificity of Sapyga is a function of nesting ecology rather than the cleptoparasitic female’s ability or need to identify the taxonomic affiliation of the host.

developing a program to rear Megachile minutissima and other possible pollinators of alfalfa. The stock for the rearing program came from several villages near Tel el-Kebir (N30 ° 33 9 30 0 E31 ° 56 9 13 0), about 50 km west of Ismailia, also on the Nile Delta. No doubt the two cleptoparasitic species were introduced with the importation of the hosts’ nests to the university campus. Because of the large array of artificial nesting panels (trap nests) established by S.M.K., we were able to gather considerable study material for this investigation as well as for our study of Coelioxys coturnix ( Rozen and Kamel, 2008) .

Although most of our specimens and observations were from the trap nests on the university campus, we also collected nest samples from some of the panels deployed in the field at Tel el-Kebir, where parasitized nests of Osmia submicans were recovered. Still other nests were collected from panels posted at the Ismailia Experimental Station , Agricultural Research Center, Ismailia. Because this locality is only a few kilometers from the university, observations from there are not differentiated from those at the university .

MATERIALS, METHODS, AND TERMINOLOGY

Nest panels are oversized trap nests designed to attract and rear large numbers of cavity-nesting bees ( Rozen and Kamel, 2008: figs. 1, 2). Although varying somewhat in outside dimensions, a typical nesting panel is built from a series of polystyrene boards (approximately 90 cm X 15 cm X 1.8 cm), each of which has a series of channels 12 cm long (shorter than the width of polystyrene board) spaced 15 mm apart on center, cut into the upper and lower board edges on one side of the board so that when boards are stackedin succession, two channels of opposing boards when properly aligned form receptacles for the nest straws ( Rozen and Kamel, 2008: fig. 2). The straws, about 12 cm long, with a 5.1 mm inside diameter, are constructed from rolled paper and provide the preformed cavity in which the host bees can build nests. Panels when fully constructed are framed to hold the polystyrene boards together and mounted on a firm piece of plywood, which is then hung vertically either near the alfalfa field at Suez Canal University or at appropriate sites near clover ( Trifolium alexandrinum L.) or alfalfa ( Medicago sativa L.) fields in the vicinity of Tel el-Kebir, so that the straws become horizontal. See Rozen and Kamel (2008: figs. 1–3) for images. The trap nests are described elsewhere in greater detail (http://www.polinatorparadise.com/ Egypt.htm).

Most straws that we examined had been used in one or more previous seasons and thus often contained remnants of nests and/or cocoons of preceding generations of Megachile minutissima , other Megachile , one or more species of Osmia , Sapyga luteomaculata , Stelis (Stelis) murina Pérez , and Coelioxys coturnix . The most abundant nester in the panels on campus, M. minutissima , apparently never removed accumulated debris from previous generations but merely mined through the cocoons and leaf litter, selecting a space and installing fresh leaf snippets to construct a new cell within the existing materials. Rarely were more than two freshly provisioned cells found end-to-end, generally toward the rear end of a straw. With straws not used previously, cells were almost invariably found only at the far rear end. Entrances to straws containing fresh cells were closed with leaf snippets loosely packed against one another within the first several cm of the front end, and the passage between the stopper and the fresh and old cells was left open. It is not surprising that Dermestidae often occupied nest straws, where they fed on the dead bodies of failed immature or adult bees and other debris, although live immature bees there appeared unharmed for the most part by these scavengers.

To study the nests, we simply pulled individual straws from the polystyrene panel with forceps or pliers and inserted them in an improvised carrying container made from pieces of polystyrene bound together by strapping tape, so that horizontal orientation of the straws could be maintained during transport. In the lab, each straw was cut lengthwise with a utility knife, and the two cut edges were spread to reveal the cells, which we then opened with forceps under a stereomicroscope. Specimens were preserved, reared in plastic rearing dishes, or discarded, depending upon their perceived utility.

Where we describe the egg of Sapyga luteomaculata , we also present statistics on the number of ovarioles, egg size, and number of mature oocytes carried by cleptoparasites. This area of investigation was pioneered by Iwata (1965) and Iwata and Sakagami (1966) and added to by a number of more recent studies (e.g., Alexander, 1996; Rozen, 2003). For the study of egg size in bees, Iwata and Sakagami (1966) developed an index of egg size relative to the size of the female bee by dividing the length of the largest mature oocyte or egg by the distance between the outer rims of the female tegulae. An egg index when applied to such a distant relative as a Sapyga , in which the adult anatomy differs markedly from that of a bee, has doubtful value and is not employed here.

The terminology regarding anatomy follows that commonly used in descriptions of bee larvae ( Rozen, 2001). It corresponds closely with that employed by Kojima (1998) for social Vespidae , except for the following (Kojima’s terms italicized parenthetically): epistomal area (palate); epistomal ridge (or suture) below (or laterad of) anterior tentorial pit (parafrontal suture); epistomal ridge (or suture) between anterior tentorial pits (frontclypeal [sic] suture); postoccipital ridge (posterior thickening of head capsule). All head ridges are internal and are best observed on head capsules that have been cleared of tissue by brief boiling in an aqueous solution of sodium hydroxide or on cast larval exoskeletons (euphemistically termed ‘‘skins’’ hereafter).