Life cycle of Huarpea fallax (Hymenoptera: Sapygidae) in a xeric forest in Argentina

This study describes the life cycle of Huarpea fallax (Hymenoptera: Sapygidae) in a xeric forest in La Pampa province, Argentina. This cleptoparasitic wasp attacks the nests of two species of leaf-cutter bees: Megachile catamarcensis and Anthidium vigintipunctatum, both belonging to the family Megachilidae. Nests of these bee species were obtained during a trap-nesting programme. Adult emergence showed a unimodal pattern indicating a univoltine life cycle. The period from egg-laying to adult emergence lasted for 10–13 months; however, one female took about 2 years to emerge, suggesting parsivoltinism. Most females attack one cell per host nest, the outermost cells being the ones most parasitized. However, the position of the attacked cells was variable. In this paper, although there were insufficient data to prove a correlation, the data suggest a positive trend between body size of sapygid wasps and their host bees.


Introduction
Sapygidae is a small, widely dispersed family, absent only in the Australian Region. This family comprises about 80-90 species, which are divided into two subfamilies: Fedtschenkiinae and Sapyginae (Pate 1947). All species are solitary and their larvae are cleptoparasitic on megachilid and apid bees (Pate 1947). In the Neotropical Region, Sapyginae are represented by three genera: Araucania Pate, Huarpea Pate and Sapyga Latreille (Brothers 2006). The genus Huarpea contains four species confined to southern South America. The strong, well-developed occipital carina and the peculiar conformation of the male antennal flagellum ( Figure 1A) distinguish this genus from all other Sapyginae (Pate 1947).
There is relatively little information about the biology of the species of this genus. Individuals of Huarpea wagneriella (Buysson) have been reared from nests of Megachile sp. in Brazil, and Huarpea fallax (Gerstaecker) from nests of Xylocopa augusti Lepeletier and Xylocopa brasilianorum (Linnaeus) in Argentina and Paraguay (Friese 1923). However, Pate (1947) believed that Friese's identifications were "questionable" and commented that "it is more likely that the Sapygid reared from nests of these Xylocopae were Huarpea wagneriella or perhaps a still undescribed South American species of Sapygid" (p.420). A century ago, Jörgensen (1912a) observed individuals of H. fallax (cited as Sapyga fallax Tasch.) in nests of Xylocopa splendidula Lepeletier and X. brasilianorum in Mendoza Province, Argentina, but Brèthes (1916) corrected the determination of the latter species to Xylocopa ordinaria Smith and/or Xylocopa mendozana Enderlein. Almost 50 years later, Hurd and Moure (1961) reported this sapygid wasp (cited as Polochrum fallax Burmeister) from nests of X. brasilianorum in Paraná (Brazil).
Currently, some authors are examining bees associated with trap nests in agroecosystems in the Pampean Region (Torretta and Durante 2011;Torretta et al. 2012, Torretta et al. forthcoming) and individuals of H. fallax have been reared from nests of native bees in a xeric forest in La Pampa province. Therefore, the objectives of this paper are to provide information about the life cycle of this cleptoparasitic wasp species, including host relationships, time of development, description and measurements of cocoons, and other associated organisms in trap nests.

Study site
The study was conducted in a xeric forest (Caldenal) on Estancia Anquilóo (36°25' S, 64°48' W), Toay, La Pampa province, Argentina. The Caldenal is an ecosystem located in central Argentina (biogeographic province of Espinal, Caldén district), primarily in La Pampa province (Cabrera 1994). This xerophytic open forest system is a transitional ecosystem between the Pampas grasslands, to the east, and the dry Monte shrublands, to the west. It is dominated by the caldén tree (Prosopis caldenia, Fabaceae) with an understorey of perennial grasses.

Collection of parasitized nests
Trap nests were placed in the field during activity periods of bees and wasps (November 2010 to March 2011) and were inspected monthly thereafter. Each trap nest consisted of one hollow bamboo cane, which was cut so that a nodal septum closed one end (Aguiar and Garófalo 2004). Each cane was also cut longitudinally and taped closed, and measurements were taken of both its total length (from the entry to the node) and the inner diameter at the entry. In total, 280 bamboo canes were placed, arranged in 20 blocks of 14 canes. These blocks were located at two forest sites along four transects (two at each site) at intervals of 50-100 m, and 1-2 m above the ground in trees. At each monthly visit the traps with nests were removed and taken to the laboratory, where the cells were separated into plastic vials with cotton plugs, numbered from 1 to n (starting from the innermost) and kept in the laboratory at room temperature (c.15-25°C) until adult eclosion. In those nests where H. fallax adults emerged, host species were determined, and sapygid wasps were sexed; the positions of the attacked cells in the trap nest and emergence dates were also registered. A comparison between the sexes (non-parametric Wilcoxon test) was also performed to determine if differences in development time exist. Cocoons were described and measured.
The material studied is located at the Facultad de Agronomía, Universidad de Buenos Aires, Argentina (FAUBA), and at the Museo Argentino de Ciencias Naturales, Buenos Aires, Argentina (MACN).

Relationship between body size of Huarpea fallax specimens and their host bees
To determine if a relationship exists between the size of individual sapygid wasps and their hosts, the linear correlation between the body sizes of sapygids and bees reared from the same nest was calculated. For wasps and bees the intertegular span was used as an index of adult body size. Intertegular distance was measured as shortest linear distance between inner margins of wing tegulae across the thoracic dorsum (Cane 1987). Measurements were taken using a micrometer to the nearest 0.1 mm. For each nest, at least three adults, randomly chosen, were measured and the mean was used as a metric. If the number of emerged adults was less than three, all adults were measured (one or two). The sapygid which emerged with a perilampid wasp from the same cell (see below) was excluded from the analysis. In addition, a comparison between the sexes (non-parametric Wilcoxon test) was carried out, to determine if differences in body size exist.

Results
In total, 11 individuals of H. fallax ( Figure 1B, C) were obtained (eight females and three males) from 10 different nests. Only in one nest (#384) were two cells parasitized by H. fallax (Table 1). Nests were grouped in six different blocks of canes (Table 1). Eight nests were built by females of Megachile catamarcensis Schrottky ( Figure 1D) and two by Anthidium vigintipunctatum Friese (Table 1). Most trap nests consisted of one nest, except for one in which two nests of M. catamarcensis had been built, separated by a space of 75 mm (#555 and #555B). Seven nests were obtained in December 2010, whereas the others were collected in January 2011. The positioning of parasitized cells by females of H. fallax was variable among nests, although the attacked cells were mainly the outermost ones in the traps (Table 1).

Discussion
This paper provides biological information about the poorly known cleptoparasitic wasp species H. fallax in relation to its life cycle, host-parasite relationships, other parasitic species associated, and development time. Biological information on species of Sapygidae is known mainly for species of the northern hemisphere. Parker (1926) described the egg and preimaginal stages of Polochrum repandum Spinola reared from nests of Xylocopa violacea L. Pate (1947) summarized the host species for numerous species of Sapyginae. Later, Krombein (1967) comprehensively described the life cycle of Sapyga centrata Say, and Torchio (1972Torchio ( , 1979 studied Sapyga pumila Cresson parasitizing two economically important bee species, the alfalfa pollinators Megachile rotundata (Fabricius) and Megachile pacifica (Panzer). Recently, Rozen and Kamel (2009)  described the pre-imaginal stages, and discussed the behaviour and morphological adaptations of early larval stages to kill host larvae. As mentioned above, there are few data for Neotropical species. Hurd and Moure (1961) reported on parasitism by H. fallax on different species of carpenter bees of the genus Xylocopa and reviewed the limited knowledge of the life cycle of this sapygid wasp. In this work, I reared adults of H. fallax from the nests of two species of Megachilidae bees, instead of carpenter bees. In our study site, X. ordinaria and X. splendidula were observed throughout the sampling period (November-March) foraging on flowers (Marrero 2013), but these carpenter bees did not use trap-nests. These two species of Xylocopa have been cited as hosts of H. fallax in Mendoza province (500 km northwest of our study site) (Jörgensen 1912a(Jörgensen , 1912b. The current work is the first report of this species of Sapygidae attacking nests of Megachilidae. Emergence of adults of H. fallax demonstrates a clear unimodal pattern, indicating a univoltine life cycle with adults emerging in late spring and parasitizing nests between late spring and early summer. The period between egg laying and adult emergence was similar between years and lasted for 10-13 months. At our study site, M. catamarcensis (host of H. fallax) is a specialist on the pollen of Prosopis caldenia (Torretta et al. forthcoming), flowering of which occurs in December-January (Burkart 1976). A high synchronicity between flowering of host plant and activity of specialist bee and cleptoparasitic wasp would be expected. One female took about 2 years to emerge, suggesting that H. fallax exhibits parsivoltinism. In several bee and wasp species, some individuals of each age cohort emerge in the following year (Torchio and Tepedino 1982;Wcislo and Cane 1996) and this may be a bet-hedging strategy related to unpredictable resources or to natural enemies or host availability (Neff and Simpson 1992;Wcislo and Cane 1996  Sapygid cocoons studied in our population were similar in appearance to those reported by Jörgensen (1912a) from X. brasilianorum and X.splendidula in Mendoza, but smaller (9.2 ± 0.3 × 3.9 ± 0.2 mm in the La Pampa population versus 13 × 7 mm in the Mendoza population). The quantity of food supplied has been considered as an explanation for the variations in size that occur within the same species (Mickel 1924). Species of Xylocopa cited by Jörgensen (1912b) as hosts for H. fallax are bigger than M. catamarcensis and A.vigintipunctatum. If the provisioned pollen masses of females of Xylocopa spp. were larger than Megachile/Anthidium, the difference of body size in individuals of adults of H. fallax could be due to this fact. Within cleptoparasitic wasps, this phenomenon is well known in Mutillidae (Mickel 1924). In this study, although there were insufficient data to prove a correlation, the data suggest a positive trend between body size of sapygid wasps and their host bees. A greater number of individuals could help to confirm or reject this hypothesis. Another possible explanation could be differences in inherent body size among populations of H. fallax; obtaining individuals from nests of Xylocopa spp. and Megachilidae (or other possible host) in the same population could help to solve this issue.
In most nests attacked by H. fallax females, only one cell was parasitized. This is different for Sapyga centrata females, which have been shown to parasitize several cells in consecutive series in nests of Osmia bucephala Cresson and Osmia pumila Cresson (Krombein 1967). The same is apparently true for Brazilian populations of H. fallax according to Hurd and Moure (1961). Regarding the position within the nest, most attacked cells were the outermost ones, as reported for S. centrata (Krombein 1967) and the Brazilian population of H. fallax. The females of H. fallax parasitized nests that were also parasitized by other species (Hymenoptera and Diptera). The most remarkable fact is that in one cell of M.catamarcensis both a sapygid wasp and an individual of Perilampus sp. successfully developed. Perilampus species are very diverse in terms of their hosts, but with restrictions on particular species groups (Darling 2006); however, none is cited as a parasite of Megachilidae.