A study of the biology of Epicharis (Epicharoides) picta using emergence-traps

This study investigates the nesting habits of Epicharis picta in a nest aggregation located in a fragment of the Atlantic forest in Southeastern Brazil. Ten emergence-traps were set up in this nest aggregation to standardize data collection of phenology, natural enemies, and sex ratio. Epicharis picta nests were in an area of 160 m2 with a density of 41 nests/m2. Nest and cell architecture are described. Epicharis picta is a protandrous, univoltine species with its emergence in this study occurring between 28 January and 15 April. We provide direct evidence of parasitism on E. picta by Rhathymus friesei, Tetraonyx sexguttata and T. aff. lycoides. The predator Apiomerus lanipes was found to prey Epicharis for the first time. We suggest the use of emergence-traps as tools to support studies of ground-nesting bees. In addition, we compile, update, and discuss data on the nesting biology of all Epicharis subgenera.


Introduction
Solitary bees use a wide variety of nesting substrates, digging their nests in soil or wood, constructing freestanding nest, or using pre-existing cavities (Michener 1974(Michener , 2007. In fact, most of them nest in the ground (Linsley 1958;Batra 1984;Roubik 1989;Michener 2007), which may be a plesiomorphic condition among bees although there are also more generalized cleptoparasites, such as those of the genus Mesoplia Lepeletier, 1841 (Apidae, Ericrocidini) which also attack nests of Centris (Gaglianone 2005;Rocha-Filho et al. 2008;Rocha-Filho et al. 2009;Vivallo 2020a). Besides bees, there are records of Epicharis as a host for both cleptoparasitic species of Meloidae (Coleoptera) and parasitoid species of Conopidae (Diptera) and Mutillidae (Hymenoptera) (Gaglianone 2005;Rocha-Filho et al. 2008;Gaglianone et al. 2015). Despite the scarcity of information in the literature, species of Hemiptera, especially those of the genus Apiomerus (Reduviidae), are known to be predators of adult bees (Silva and Amaral 1973;Cane 1986;Amaral-Filho et al. 1994;Marques et al. 2003Marques et al. , 2006Silva and Gil-Santana 2004).
Studies on the nesting biology, relationships with natural enemies, and phenology of Epicharis species have been performed using direct observations of nest aggregations. Nonetheless, there is a need for methods that provide standardization for data collection. To this aim, emergence-traps have been used in ground-nesting solitary bees and wasps studies and have shown to be effective in answering key questions on the biology of these insects (Hiller and Wittmann 1994;Sardiña and Kremen 2014;Rocha-Filho and Melo 2017;Cope et al. 2019).
Epicharis picta occurs in Uruguay, Paraguay, Argentina, and Brazil (Federal District and the states of Espírito Santo, Minas Gerais, Paraná, Paraíba, Rio de Janeiro, Santa Catarina, and São Paulo - Moure et al. 2012). Its life cycle is univoltine (Gaglianone et al. 2015) and it is oligoletic on Malpighiaceae . However, data on its nesting behavior, nests and cell architecture, and natural enemies are scarce in the literature.
In this study, we investigated the nesting habits of Epicharis picta for two years and provide information on its biology, natural enemies, nest and cell architecture, sex ratio, and phenology, based on direct observations and emergence-traps data. An updated compilation of nesting biology and nest architecture data of the Epicharis genus is also provided.

Study site
The nest aggregation of Epicharis picta studied was in a fragment of semideciduous, montane and submontane Atlantic Forest (Veloso et al. 1991), in the municipality of Viçosa, Minas Gerais-Brazil, on the access road to the Estação de Pesquisa, Treinamento e Educação Ambiental Mata do Paraíso (20°47'56"S, 42°52'07"W) (see Gaglianone et al. 2015). The climate of the region is temperate, rainy (mesothermal), with hot and rainy summers, and cool and dry winters (type Cw, according to the classification of Köppen - Kottek et al. 2006). The climatological data of the region between 2009 and 2011 are shown in Fig. 1. Nesting biology, nest architecture, and brood cells The fieldwork was carried out for two years. In 2010, visits took place monthly; and in 2011, daily, during the period of activity of the adult bees. The nesting habits of E. picta were obtained from direct observations throughout each day from 5:30 h to 19:00 h. To study the brood cells, eight excavations were made in 50 cm × 50 cm plots. The brood cells found in the soil were collected, placed in plastic pots with substrate from the nesting site, and kept in the laboratory to assess their content and dimensions. For nest architecture, as the main burrow remains open after a nest is completed, we injected plaster in five nests. Ten plots of 1 m 2 were randomly set up in the aggregation to estimate nest density.

Emergence-traps, natural enemies, and associated species
To standardize the sampling method when studying phenology, sex ratio, and parasitehost synchrony, 10 emergence-traps were randomly set up at the nesting site from 20 January to 20 May 2011 ( Fig. 2A). The emergence-traps consisted of a pyramidshaped wooden frame covered with thin nylon mesh, with the following dimensions: 50 cm × 50 cm base, 10 cm × 10 cm top end, and 50 cm height (Fig. 2B). On one side, a 12 cm long opening was made in the longitudinal direction to install a Velcro tape, forming a "window" that allowed our access to its interior. (Fig. 2B). A piece of PVC pipe attached to a plastic container was placed at the upper end with its bottom facing the interior of the trap and its top inside the plastic container, which was filled with a 1:1 alcohol and water solution (Fig. 2B). Inspections on emergence-traps were conducted daily.
Data on potential natural enemies and associated species that were seen visiting the nest aggregation area were collected throughout the study period. Vouchers from the specimens studied are deposited at Museu Regional de Entomologia, Departamento de Entomologia-Universidade Federal de Viçosa (MEUFV). This study follows the classification system of Moure et al. (2007), which treats the whole group of bees as a single family (Apidae).

Nesting biology, nest architecture, and brood cells
Epicharis picta nests were aggregated in an area of approximately 160 m² of exposed slopes of about 45°, with an average nesting density of 41 entrances/m². Females began their activities between 6:00 h and 6:30 h and ended between 18:00 h and 18:30 h, daily. The activity peak occurred between 7:30 h and 11:30 h. The females rested inside the nests at night (Fig. 3A). Male activity period was shorter, between 6:00 h and 16:00 h. The mating male behavior is being dealt in a separate study. Every day at the beginning of activities (~6:00 h), females took about five minutes at the nest entrance before making their first flight. After their first trip, their scopae were still clean when they returned, with no evidence of oil or pollen having been collected (Fig. 3B). For digging new nests, they selected a new nest site, excavated the soil vertically, using their mandibles, anterior and middle legs, and deposited the material from the excavation around the nest entrance, forming a circular tumulus (Fig. 3C, D). The excavated nests (N = 8) contained one to two cells each. In nests with only one cell (N = 6), there was a single tunnel ( Fig. 3E, F), whereas in nests containing two cells at the end of the main tunnel (N = 2), a branch with no cell was also found. The diameter of these tunnels ranged from 10 mm to 12 mm (N = 20). The nests were perpendicular to the surface, ranging from 50 cm to 110 cm deep.
The brood cells were slightly curved ( Fig. 4A-E) and built at an angle of about 45° to the ground surface. Their outer walls are rough and rigid, whereas inner walls are smooth, shiny, and hydrophobic ( Fig. 4A-C). The cell cap is inserted below the apex of the cell wall and is slightly inclined (Fig. 4D). The cells size ranged from 20 mm to 27 mm (X̅ = 24.57 mm; N = 27) in length, between 10 mm and 12 mm (X̅ = 11.12 mm; N = 27) in diameter at cell cap height, and the base diameter between 12 mm and 14.5 mm (X̅ = 13.74 mm; N = 27).
Females of Epicharis (Epicharoides) albofasciata were observed founding nests (N = 14) in the aggregation of E. picta. Only one nest of E. albofasciata was excavated, and it consisted of a single 35 cm deep tunnel with one cell at its end. Agonistic behavior among females of E. picta and E. albofasciata was observed when females returned from the field. E. albofasciata males were neither observed nor collected on the nest aggregation during fieldwork.
Epicharis picta first appeared in the emergence-traps on 29 January, and males were the first to emerge (Fig. 5B), indicating protandry. From a total of 211 individuals, 111 were male (52.6%), and 100 female (47.4%). Thus, the sex ratio of E. picta was 1.11 males to 1 female. Rhathymus friesei emerged from 5 February to 18 March, peaking from 26 February to 11 March. Males were the first to emerge, on 5 February, whereas females emerged from 19 February, which also indicates protandry (Fig. 5C). From the 23 emerged individuals, 13 were males (56.53%) and 10 were females (43.47%), resulting in a sex ratio of 1.3 males to 1 female.

Natural enemies and associated species sampled from direct observations
In addition to species sampled from the emergence-traps, we collected another 24 species of insects found in the area, which were then identified and classified according to their association with the nesting aggregation (Table 1). Rhathymus friesei was the most abundant natural enemy (Fig. 6A); in some moments, up to four individuals could be seen inspecting the nest aggregation at the same time. Tetraonyx sexguttata was observed walking on the aggregation, landing on vegetation, and emerging from E. picta nests (Fig. 6B). Individuals of Apiomerus lanipes were observed five times near the entrances of the nests capturing females of E. picta when those entered or left their nest. The predator attacked the prey with its forelegs, inserting the stylet between the thorax and the head (Fig. 6C). Twenty-seven individuals of Mutillidae were collected in the aggregation. However, only one female of Traumatomutilla sp. was observed directly inspecting E. picta nests (Fig. 6D). Physocephala sp. specimens were also observed flying over the nest site between 5 February and 22 March.
Females of Augochloropsis cf. cupreola (Apidae, Augochlorini), Hypanthidium nigritulum (Apidae, Anthidiini), and Colletes petropolitanus, (Apidae, Colletini) were observed performing inspection flights over the soil and branches of vegetation in the aggregation, but they neither nested nor interacted with E. picta females. Trigona spinipes (Apidae, Meliponini) workers landed on the nest site and collected soil material removed by E. picta females during the excavation of their nests. A female of Mesoplia rufipes was collected on 20 March 2010. This bee flew over the aggregation and periodically approached some entrances of E. picta nests. However, it was not seen entering any nest.

Brood cells collected during excavations
In the two years of studies, a total of 121 cells were collected, ranging between 30 cm and 110 cm deep. From these, 45 were already open, containing only soil in their interior. Six cells were taken by fungi, one of which contained a dead E. picta female pupae (Fig. 7A). Another six cells containing only fungi on the food (Fig. 7B) with no evidence of dead larvae or egg, nor parasitic traces. In one cell there was an exoskeleton of T. sexguttata (Fig. 7C). In the 46 cells, there were 38 mature larvae (Fig.  4B) and eight pupae of E. picta. From the 17 cells kept in the laboratory throughout  7D) and six females of E. picta, two T. sexguttata, and one T. aff. lycoides emerged. In the remaining four cells, there were mature larvae of R. friesei. Larvae of R. friesei were easy to identify due to the presence of their cocoon (see Rozen 1969;Camargo et al. 1975;Werneck et al. 2012).

Discussion
Studies reporting biological data about E. picta are recent in the literature (Werneck et al. 2012;Werneck et al. 2015;Gaglianone et al. 2015). These, however, do not address aspects related to their nesting habits, natural enemies, and associated species. Therefore, this present study is the first to bring such data, in addition to using emergence-traps as a model.

Nesting biology and notes on patterns among Epicharis subgenera
Our data on E. picta reinforce the hypothesis that all species of Epicharis nest gregariously in the soil (Vesey-FitzGerald 1939; Michener and Lange 1958; Rozen 1965;  In spite of the low quantity of species studied, it is possible to point to a pattern on the exclusive use of sandy soil for nesting by the subgenera Epicharoides, Epicharis, Triepicharis and Anepicharis (Table 2). Although Parepicharis also uses sandy soil, there are records of E. metatarsalis nesting in clay (Thiele and Inouye 2007). For Epicharana, there seems to be some preference over nesting in low light places, as recorded for E. flava (Camargo et al. 1975), E. rustica, and E. elegans (Michener and Lange 1958). Despite that, it is not possible to point to any patterns regarding nest depth, which can vary according to the characteristics of each site used for nesting (Cane 1991), nest density, and the number of brood cells per nest. A study of phylogenetic reconstruction based on nesting behavior characters encompassing the whole Centridini tribe is under way (Werneck HA, unpublished data). As a result, we expect a better understanding of the patterns and evolution of these characteristics. For a comparison of the characteristics regarding nesting habits amongst the species of Epicharis studied, see Table 2.   (2007); g Roubik and Michener (1980); h Raw (1992); i Hiller and Wittmann (1994) The data on E.
(Triepicharis) analis were joined with those on E.
Here, we joined the data from Raw (1992) with the available data for E.

dejeanii;
Species that nest in the soil in aggregations might also build their nests in the nest aggregations of other species (Michener 1974). For instance, there are reports of E. albofasciata building their nests in E. nigrita (Gaglianone 2005) and E. picta (this study) nest aggregations. However, these bees can build their own nest aggregation, as described by Rozen (2016). The characteristics described by Rozen (2016), such as nest depth, presence of a tumulus around the entrance of each nest, female preference for inclined sites (about 45°), and shape and composition of the brood cells, corroborate the findings and allow us to point out that these characteristics may be diverse within Epicharoides.
The development of immature stages seems to be more constant within Epicharis. Both E. picta and E. albofasciata present the same hatching pattern, with the presence of a pharate first instar, which is also recorded for E. flava and E. nigrita (Camargo et al. 1975;Gaglianone et al. 2015). Pharate first instar larvae were also observed for Centris flavofasciata Friese, 1899 (Rozen et al. 2011), in addition to other groups of solitary bees that nest in the soil, such as Monoeca haemorrhoidalis (Smith, 1854) (Apidae: Tapinotaspidini) (Rozen et al. 2006).

Emergence-traps and phenology
Emergence-traps have been effective in collecting data on species of bees and wasps nesting in the soil (Hiller and Wittmann 1994;Sardiña and Kremen 2014;Rocha-Filho and Melo 2017;Cope et al. 2019;Martins et al. 2019). In the present study, these emergence-traps were useful for sampling data on natural enemies that emerged in the nest aggregation as well as in measuring the phenology and sex ratio of E. picta. We were able to measure the emergence patterns of both the host and its natural enemies.
Our data from emergence-traps corroborate the hypothesis that most Epicharis species are univoltine (see Gaglianone et al. 2015). The beginning of the emergence of E. picta occurred during the period of high temperatures and rainy season. The adults remained active until the middle of April, when the temperatures decreased and the rainfall declined drastically (see Fig. 1). These abiotic factors combined with other biotic features may influence in the diapause of Epicharis. A discussion about the biotic and abiotic factors that can influence the diapause process, and, consequently, the phenology of Epicharis species, can be seen in Gaglianone et al. (2015).
The emergence of Rhathymus friesei occurred about one week after the beginning of the E. picta adult activity. As cleptoparasitic species require provisioned brood cells from their hosts to oviposit (Wcislo 1987;Rozen 2001;Michener 2007), this emergence pattern was expected. Another fact that reinforces this pattern is that Rhathymus females lay their eggs in cells closed by the host female (Camargo et al. 1975;Rozen 1991). The strategy of parasitizing cells closed by the host is a plesiomorphic feature among the cleptoparasitic species that might have evolved from nest-building species, which could have five independent origins within bees (Litman et al. 2013).
Augochlora thalia and Acamptopoeum prinii, and Epinysson sp. were collected in emergence-traps. These species nest in the soil and presumably were just opportunistically using the Epicharis nest site. The issue of nesting in nest aggregations of other species may be related to the strategy of inhibiting parasite attacks (Wcislo 1987;Wcislo and Cane 1996). Physocephala sp., also collected in the emergence-traps, is a parasitoid of bees and is discussed in the next section of this study.
Some natural enemies do not emerge in the same nest aggregation in which they attack their hosts. This can be a problem when the inference about the relationships between natural enemies and hosts is conducted by direct observations. Therefore, the use of emergence-traps allowed us to determine which species of natural enemies actually emerged from the nest aggregation. On the other hand, as in this study we have a mixed species nest aggregation, there were limitations to determine the type of association among some species that emerged in the emergence-traps (see Table 1).
In summary, we suggest that emergence-traps are tools that can aid in studies of solitary bees nesting in the soil. This method allows the measurement of phenology, parasite-host synchrony, parasitism rate, and sex ratio.

Natural enemies and species associated with the nest aggregation of E. picta
Many natural enemies are reported for Epicharis species, but there is direct evidence only for Rhathymus spp. and Tetraonyx spp. (Werneck et al. 2012;Gaglianone et al. 2015; this study). Indirect evidence, however, is reported for species of Mesoplia and Mesonychium (Apidae, Ericrocidini), Physocephala spp. and several species of Mutillidae (Camargo et al. 1975;Hiller and Wittmann 1993;Gaglianone 2005;Rocha-Filho et al. 2008;Luz et al. 2016). Regarding cleptoparasitism on E. picta, our data show that there is direct evidence only for R.friesei, T. sexguttata and T. aff. lycoides. Rhathymus friesei was the most abundant cleptoparasitic species observed in this study. Even though it is not possible to determine specificity relationships between Rhathymus species and Epicharis subgenera, there is specificity in the cleptoparasite-host relationship between the genera Rhathymus and Epicharis (Werneck et al. 2012). On the meloid beetles, there are data on the relationship of Tetraonyx spp. to E. dejeanii Lepeletier, 1841 (Hiller and Wittmann 1994), E. nigrita (Gaglianone 2005;Martins et al. 2019), E. bicolor (Rocha-Filho et al. 2008), and E. picta (Gaglianone et al. 2015).
The Mutillidae is composed of parasitic wasps that attack Hymenoptera in general, with records for bees as hosts (Brothers et al. 2000;Luz et al. 2016). Although some studies report the parasite-host relationship between these wasps and Epicharis, there is no direct evidence of this relationship (Rocha-Filho et al. 2008). Luz et al. (2016) compiled the data known for host bees of these wasps and these authors consider Apidae to be the main hosts of Mutillidae in the Neotropical region. For Centris, nine records of Mutillidae species as parasitoids are known (see Luz et al. 2016), whereas for Epicharis, the only record is for Hoplomutilla myops myops (Burmeister, 1854), considered as a potential parasitoid of E. bicolor (Rocha-Filho et al. 2008). In the present study, it was only possible to observe Traumatomutilla sp. approaching the nests of E. picta, although it was not seen entering nor leaving the nests.
Apiomerus are predators, some species being reported as common predators of bee species, such as stingless bees (Apidae, Meliponini) (Silva and Gil-Santana 2004), and Apis mellifera Linnaeus, 1758 (Apidae, Apini) (Silva and Amaral 1973;Amaral-Filho et al. 1994;Marques et al. 2003Marques et al. , 2006. In this study, we show direct evidence that this hemipteran preys on E. picta. Until now, no species of Epicharis had been associated with this predator. Data on the biology of A. lanipes and its mode of predation are still scarce in the literature and it will be necessary to study whether it is a predator specialized in bees.