DOI: 10.13102/sociobiology.v61i1.52-59Sociobiology 61(1): 52-59 (March, 2014) Open access journal: http://periodicos.uefs.br/ojs/index.php/sociobiology ISSN: 0361-6525 Studies on an Enigmatic Blepharidatta Wheeler Population (Hymenoptera: Formicidae) from the Brazilian Caatinga JC Pereira1, JHC Delabie3,4, LRS Zanette1, Y Quinet1,2 Introduction The myrmicine ant genus Blepharidatta is a strictly Neotropical group that was described by Wheeler (1915) from workers of Blepharidatta brasiliensis Wheeler collected near Belém (State of Pará, Brazil), in the Amazon Forest. Together with the genus Wasmannia, it forms the monophyletic tribe Blepharidattini (Wheeler & Wheeler, 1991; Bolton, 1995), which is considered to be a close relative of the fungus-growing ant tribe Attini (Schultz & Meier, 1995). Up to seven species are currently recognized (Silva, 2007), but most of them are waiting for a formal taxonomic treatment or confirmation. Based on morphological as well as behavioral data, only three species are formally recognized: B. brasiliensis found in the Amazonian forest, Blepharidatta conops Kempf, an inhabitant of savanna-like formation from central Brazil (Cerrado), and an undescribed species (Blephari- datta sp-ba, hereafter) known from the Atlantic rainforest Abstract Blepharidatta is a rare Neotropical ant genus formed by predatory species whose small colonies nest in soil or leaf-litter. A population of Blepharidatta that presents affinities with Blepharidatta conops Kempf was found in the Caatinga biome, at the “Reserva Particular do Patrimônio Natural Serra das Almas” (RPPNSA), in Crateús (State of Ceará, Brazil). The aim of our study was to obtain data on the nest architecture, size and com- position of colonies, foraging behavior, and female castes morphology for this newly found population, and to compare it with other Blepharidatta species, particularly with B. conops. The results show that Blepharidatta sp. and B. conops share key features of their biology such as their basic nest architecture, diet and foraging behavior, and the presence of a single ergatoid queen with a phragmotic head. However, marked differ- ences were also found in head and mesosoma morphology of the queen, nest architec- ture, colony size, and queen location in the nest. Two alternative hypotheses are pre- sented. The newly found Blepharidatta population represents a new species, possibly endemic to the Caatinga biome or it represents an extreme of the phenotypic variations observed among the populations forming B. conops. Sociobiology An international journal on social insects 1 - Universidade Federal do Ceará, Fortaleza, Ceará, Brazil 2 - Universidade Estadual do Ceará, Fortaleza, Ceará, Brazil 3 - Universidade Estadual de Santa Cruz, Ilhéus-Itabuna, Bahia, Brazil 4 - CEPLAC/CEPEC Centro de Pesquisas do Cacau, Itabuna, Bahia, Brazil Article History Edited by Gilberto M M Santos, UEFS, Brazil Received 02 October 2013 Initial acceptance 28 November 2013 Final acceptance 29 January 2014 Keywords: northeastern Brazil, semi- arid environment, ants, Blepharidat- tini, ergatoid queen Corresponding author: Yves Quinet Laboratório de Entomologia Instituto Superior de Ciências Biomé- dicas – ISCB Universidade Estadual do Ceará Fortaleza, CE, Brazil, 60714-903 Email: yvesq@terra.com.br of the State of Bahia, eastern Brazil (Rabeling et al., 2006; Brandão et al., 2008; Cassano et al., 2009) (Fig 1). All are ground-dwelling predatory species that nest in the ground or in the leaf-litter, with small monogynous or polygynous colonies (Brandão et al., 2001; Rabeling et al., 2006; Silva, 2007). Queens are ergatoid (sensu Peeters, 1991, i.e. perma- nently wingless and worker-like) and, at least in B. conops, it is believed that the foundation of new colonies is by fission of established colonies (Brandão et al., 2001). The best studied species is B. conops (Diniz & Brandão, 1997; Diniz et al., 1998; Brandão et al., 2001, 2008). B. con- ops colonies (up to 250 workers) live in simple nests with a single opening and a vertical tunnel with 2 cm of diame- ter and 20 cm deep that ends in a cone-shaped widening at the bottom. Furthermore, all mature nests have a horizontal subsidiary chamber connected to the vertical tunnel through a narrow tunnel that opens at the mid-length of the verti- cal tunnel. It serves as a refuge for queen and brood when RESEARCH ARTICLE - ANTS Sociobiology 61(1): 52-59 (March, 2014) 53 nests are visited or inhabited by myrmecophiles or predators. Queens of B. conops have a characteristic phragmotic head that together with the anterior slope of pronotum forms a frontal disk whose shape and dimensions fit the subsidiary chamber entrance and is used by the queen to block that en- trance (Brandão et al., 2001). The foragers patrol a roughly circular area around the nest opening, where they collect live and dead arthropods (mainly other ants) to feed their larvae (Diniz & Brandão, 1997). In the bottom area of the nest, prey is dismembered, chewed and fed to larvae via trophallaxis. The discarded remains are arranged in a ring around the nest opening and workers concealed in the nest entrance ambush arthropods, including other ants, attracted by the carcasses ring (Diniz et al., 1998). The other known species of Blepharidatta have dif- ferent nesting habits: their smaller colonies (mean number of workers varying from 112 to 132) are found within the leaf- litter, in natural cavities between leaves or in rotting branches (Rabeling et al., 2006; Silva, 2007). Contrary to B. conops, the queens of B. brasiliensis and Blepharidatta sp-ba do not have a phragmotic head (Silva, 2007). Colonies of B. conops occur in locally dense but widely scattered populations, with large areas of Cerrado devoided of B. conops (Diniz & Brandão, 1997; Brandão et al., 2001). According to Brandão et al. (2001), this distribution pattern may be explained by the limited dispersal mode of ergatoid queens and by the type of nest foundation (fission of established colonies). Such nest distribution may be characteristic of all Blepharidatta species and probably explains why ant species of this genus are considered rare (Brandão et al., 2001). Here we present data on nest architecture, size and com- position of colonies, foraging behavior, morphology (female castes) of a Caatinga population of an unidentified species of Blepharidatta (Quinet & Tavares, 2005). We compare our findings to the data available for other species of Blephari- datta, particularly B. conops and discuss the identity of this potential new taxon Material and methods Study site The study was conducted from November 2011 to June 2012 in a 4.5ha area of the “Reserva Particular do Patrimônio Natural Serra das Almas - RPPNSA” (5º08’S, 40º51’W), a 6146ha protected area of deciduous thorny woodland vegetation (Caatinga) in Crateús (State of Ceará, northeastern Brazil, 5º10’S 40º40’W) (Fig 1). Nest architecture, size and composition of colonies Twenty nine nests of Blepharidatta sp. were located in the study area, using freshly killed termites as baits. Any Blepharidatta sp. worker that picked a termite was then fol- lowed back to its nest. Nineteen nests were excavated to study nest architec- ture as well as the size and composition of colonies. Before initiating a nest excavation, the maximum diameter of the nest opening and of the carcasses ring around it was recorded (see Diniz & Brandão, 1997). The carcasses forming the ring were collected, and a 30-cm deep trench was dug in order to obtain a 20-cm side cube with the nest opening in the middle of the Fig. 1. Distribution of known Blepharidatta populations and study site location in the state of Ceará (Brazil) (RPPNSA: Reserva Particular do Patrimônio Natural Serra das Almas). () Blepharidatta brasiliensis; () Blepharidatta conops; () Blepharidatta sp-ba;  studied Blepharidatta new population CEARÁ Crateús RPPNSA CAATINGA CERRADO AMAZON ATLANTIC FOREST 0 500 1000 1500 km Fortaleza JC Pereira et al - An enigmatic population of Blepharidatta from northeastern Brazil54 upper face. Starting from one lateral side, the cube was sliced with a spatula until a nest chamber or a tunnel was found. The depth, maximum height and maximum diameter of each chamber were recorded, as well as the direction, diameter and length of each tunnel leading to a chamber. All biological material (workers, queen(s), male(s) and brood; invertebrate and vegetal fragments; myrmecophiles) found in chambers or tunnels was collected, and, whenever possible, its exact location in the nest was recorded. Workers, queen(s), male(s), brood and myrmecophile organisms found in each nest were counted and fixed in ethanol 90%. Diet and foraging behavior In order to obtain information on Blepharidatta sp. diet, all invertebrate and vegetal fragments found in the car- casses ring and chambers of 10 excavated nests were ana- lyzed. Fragments were first separated in three categories: ants, other invertebrates and seeds. Ant fragments were identified at least to genus level. Fragments of other invertebrates were identified to order level. The diel foraging activity pattern was investigated by monitoring three colonies for a 24h period: the first from 12/16/2011 (10 a.m.) to 12/17/2011 (9 a.m.), the two others from 06/16/2012 (9 a.m.) to 06/17/2012 (8 a.m.). The nest opening of each colony was observed for 10min every hour and all ants that left or entered the nest were counted. In total, four hours of foraging activity were monitored per nest. Soil surface temperature was recorded each time ant activity was measured. The density and foraging area size of Blepharidatta sp. nests were assessed using a 144 m2 area (12 x 12 m) with a grid of 1m2 quadrats (Fig 6). Each quadrat was baited with sardine and checked 40 min later. Baiting was repeated three times in three successive days. Each quadrat whose bait was visited by Blepharidatta sp. workers at least one time was then baited four times, on three consecutive days, with dead workers (or soldiers) of Nasutitermes sp. All workers carrying termites back to the nest were followed and their path marked with pieces of plastic straw. Body measurements The maximum transverse diameter of the frontal disk (phragmotic head plus anterior slope of the pronotum) of 10 Blepharidatta sp. queens (one per nest) and the maximum head width of 54 workers (from 18 nests) were measured under a stereomicroscope with an ocular micrometer. A Petri dish filled with fine white sand was used to correctly position the ants under a microscope. Total body length of 10 queens (from the foremost part of the frontal disk to the tip of gaster, in individuals with outstretched body) and 20 workers (from the middle part of the clypeus to the tip of the gaster, in individuals with out- stretched body) were also measured. Voucher specimens of Blepharidatta sp. are deposited at the Myrmecological Collection of the Laboratório de Ento- mologia, Universidade Estadual do Ceará, in Fortaleza, CE, Brazil, at the Myrmecological Collection of the Museu de Zoologia of the Universidade de São Paulo [MZSP] in São Paulo, SP, Brazil, and at the Myrmecological Collection of the Centro de Pesquisas do Cacau [CPDC], CEPLAC, in Itabuna, BA, Brazil. Results Nest architecture, size and composition of colonies All excavated nests (n=19) had a carcasses ring that surrounded the nest opening, with invertebrate fragments sometimes partially obstructing it. In three nests, however, the carcasses ring was inconspicuous, with very few invertebrate or vegetal fragments. Carcasses rings were circular or elliptic, with maximum diameter varying from 5 to 20cm. Mean frag- ment density in the carcasses ring was 8 (± 6.3) fragments/ cm2 (n=10; range = 1.7-19.1 fragments/cm2). All nests had only one circular (sometimes elliptic) nest opening whose mean diameter was 0.73 (± 0.17) cm (n=14; range: 0.5-1 cm). Each nest entrance was connected to a tunnel with approximately the same mean diameter (0.74 ± 0.23 cm; n=15; range = 0.5–1.3 cm), that led to a first cham- ber (carcasses chamber), and ended in a bottom chamber (Fig 2). In six nests, the tunnel was a straight and vertical structure connecting the nest opening to the carcasses and the bottom chambers (Fig 2). In other seven nests, the tunnel had sloped parts, with, sometimes, a branching pattern (Fig 2). The bottom chamber was located at a mean depth of 26.5 (± 6.3) cm (n=19; range = 15-40 cm). It housed the Fig 2. Schematic drawing of four Blepharidatta sp. nests excavated from 11/2011 to 06//2012, at the “Reserva Particular do Patrimônio Natural Serra das Almas”, CE, Brazil. R: carcasses ring; O: nest opening; T: tunnel; CC: carcasses chamber; BC: bottom chamber; UC: undefined chamber. 2 cm O CC BC R R T O CC BC R R 2 cm CC T T T 2 cm R R O CC T BC CC O CC BC R R 2 cm UC UC UC T T Sociobiology 61(1): 52-59 (March, 2014) 55 queen and brood in most nests (n=8) where a queen was found (n=12). Its mean maximum height and diameter were 2.3 (± 1.54) cm (n=12; range: 0.7-2.5 cm) and 3.3 (± 1.07) cm (n=12; range: 2-4 cm) respectively. A carcasses chamber, so called because it was full of carcasses, was found in most (14 out 19) nests. It was located very close to the nest opening (Fig 2), at a mean depth of 4.9 (± 2.4) cm (n=14, range: 1-11 cm), and its mean maximum height and diameter were 1.9 (± 0.95) cm (n=12; range: 0.5- 2.5 cm) and 2.8 (± 0.8) cm (n=12; range: 2-4 cm), respec- tively. In six nests, one to two additional carcasses chambers were found at a depth varying from 10 to 25 cm (Fig 2). In one nest, up to five chambers were found (a carcasses chamber, a bottom chamber and three extra chambers), in addition to branching tunnels (Fig 2). A possible subsidiary chamber, as defined by Brandão et al. (2001) for B. conops, was found in one nest. This cham- ber was located at the mid-length of the vertical tunnel leading to the bottom chamber, at a depth of 13 cm, and contained a queen, brood and 15 males. Queens were missing in seven nests, possibly because they escaped during the excavation process. All other nests (n=12) had a single queen, except one that had two queens. Most queens (8 out 13) were found in the bottom chamber together with brood. In three remaining nests, the queen was found in the carcasses chamber (n=2) or in the subsidiary chamber (sensu Brandão et al., 2001) (n=1). Two queens could not be assigned to a specific nest location. A total of twenty five males were found, in chambers of four nests, together with the queen and brood. Fifteen of them were found in the subsidiary chamber of one nest; in the three other nests, from one to five males were found in the bottom chamber. Three myrmecophiles were frequently found in Blepharidatta sp. nests, almost always in the bottom cham- ber: two species of the crustacean genus Trichorhina Budde- Lund (Oniscidea, Platyarthridae) and one cockroach species (Corydiidae); we also noticed the frequent occurrence of the pseudoscorpion Petterchernes brasiliensis Heurtault (Cher- netidae), up to now recorded only from burrows of small mam- mals in the state of Pernambuco, Brazil (Heurtault, 1986). Average colony size (number of workers) was 193 (± 107.4) (n=19; range: 30-437). Most colonies (13 out 19) ranged from 110 and 220 workers; a few (n=4) reached more than 300 workers (one of them had more than 400 workers) (Fig 3). Queen and worker morphology Blepharidatta sp. queens are ergatoid and have an en- larged phragmotic head that, together with the anterior slope of the pronotum, form a nearly circular disk (Fig 4) whose mean maximum transverse diameter is 1.68 (± 0.04) mm (n=10; range: 1.60-1.76 mm) (Table 1). Frontal disk cuticle is discretely reticulate-punctate, nearly smooth; head margin is strongly curved upwards, and stiff hairs protrude laterally from the perimeter of the disk (Fig 4). Total mean body length of queens is 4.96 (± 0.12) mm (n=10; range: 4.8-5.20 mm). Blepharidatta sp. workers are monomorphic (mean body length: 3.79 ± 0.22 mm; n=20; range: 3.50-4.48 mm), and their mean maximum head width is 0.96 (± 0.05) mm (n=54; range: 0.84-1.04 mm) (Table 1). Fig 3. Distribution (%) of colony size categories for 19 nests of Blepharidatta sp. excavated from 11/2011 to 06/2012, at the “Reserva Particular do Patrimônio Natural Serra das Almas”, CE, Brazil. Fig 4. Phragmotic head of Blepharidatta conops (a) (photo by April Nobile – www.antweb.org) and Blepharidatta sp. (b) queen. 0.5 mm 0.5 mm b a Table 1. Mean maximum transverse diameter (mm) of queen (Q) frontal disk and mean workers (W) maximum head width (mm) in different Blepharidatta conops populations (SR, BA, SEL, CH-G) and in the Blepharidatta population found at the “Reserva Particular do Patrimônio Natural da Serra das Almas” (RPPNSA), in Crateús (State of Ceará). SR: Serranópolis (State of Goiás); BA: Balsas (State of Ma- ranhão); SEL: Selvíria (State of Mato Grosso do Sul); CH-G: Chapada dos Guimarães (State of Mato Grosso). Data within parenthesis refer to standard errors of the means. Data source for B. conops: Brandão et al. (2001). Blepharidatta conops Blepharidatta sp. SR BA SEL CH-G RPPNSA Q 1.55 (± 0.03) n=12 1.2 (± 0.06) n=5 1.45 n =1 1.51 n=1 1.68 (± 0.04) n=10 W 0.87 (± 0.04) n=5 0.75 (± 0.05) n=5 0.78 n=1 --- 0.96 (± 0.05) n=54 JC Pereira et al - An enigmatic population of Blepharidatta from northeastern Brazil56 Diet and foraging behavior In total, 13,576 fragments were found in the carcasses ring and chambers of 10 Blepharidatta sp. nests. Of these, 8,385 (62%) were from ants, 4,660 (34%) from others inver- tebrates, and 522 (4%) were whole seeds or pieces of seeds. Most invertebrate and seed items were found in the carcasses ring (77% and 85% respectively); the rest was found in nest chambers, mostly in carcasses chambers. Most ant fragments (93%) could be identified; 61% of the other invertebrate frag- ments and none of the seeds could be identified. Ant carcasses included 41 species from 18 (or 19) genera belonging to seven subfamilies (Table 2). Three species accounted for 50% of all ant carcasses. The most common species was Ectatomma muticum Mayr (Ectatomminae) (23%), followed by Camponotus crassus Mayr (Formicinae) (16%) and Labidus sp. nr. coecus (Ecitoninae) (11%). Other eight species (Acromyrmex rugosus (Smith), Gnamptogenys striatula Mayr, Camponotus sp., Cephalotes pusillus (Klug), Eciton sp., Crematogaster sp., Pheidole sp., Odontomachus bauri Emery) were frequent, accounting each for 3 to 7% of all ant carcasses. It is also worth mentioning the Ecitoninae group since 18% of all ant carcasses were from Eciton (3 species), Labidus or Neivamyrmex (4 species), No- mamyrmex (1 species) and an unidentified genus (1 species). Other invertebrate carcasses were from eight insect groups (Blattodea [including termites], Coleoptera, Dermaptera, Hemiptera, Hymenoptera, Lepidoptera, Neuroptera, Orthoptera), and two arachnid groups (Araneae, Scorpiones). However, the vast majori- ty (92%) of carcasses were from four groups: Hemiptera (37.5%), Coleoptera (22.4%), Araneae (19.4%) and termites (13%) (all termite fragments were heads of Nasutitermes sp. soldiers). Foraging activities of Blepharidatta sp. are predomi- nantly crepuscular, with two peaks of activity, the first one cor- responding to the night/day transition period (5 a.m.-9 a.m.), the second one to the day/night transition period (4 p.m.-7 p.m.) (Fig 5). At other periods, especially during the warmest period of the day (± 10 a.m.-3 p.m.), foraging activity stops almost entirely, at least at the nest opening (Fig 5). However, at night, from 7 p.m. to 2 a.m., intense cleaning activity was observed in all nests, with workers carrying carcasses or sand particles out of the nest and depositing them on the carcasses ring, while others removed carcasses obstructing the nest opening and organized them on the carcasses ring. During the period when nests foraging activity was monitored (up to 12 hours of nest opening observation for the three nests), it was observed that the carcasses ring was fre- quently visited by invertebrates, including by ants that robbed carcasses and sometimes inspected Blepharidatta sp. nest opening. However, none of these visitors was ambushed by Blepharidatta sp. workers, contrary to what was observed in B. conops, whose workers, concealed inside the nest entrance, frequently ambush visitors (Diniz et al., 1998). Collective transport of large prey items was observed in Blepharidatta sp. in one occasion. More than 12 workers Table 2. List of subfamilies and genera identified in ant carcasses found in the carcasses ring and nest chambers of 10 Blepharidatta sp. nests, with number of species found in each genus. Subfamily Genus Species number Dolichoderinae Dolichoderus 1 Dorymyrmex 1 Ectatomminae Ectatomma 1 Gnamptogenys 1 Ecitoninae Eciton 3 Labidus/Neivamyrmex 4 Nomamyrmex 1 unidentified genus 1 Formicinae Brachymyrmex 1 Camponotus 5 Myrmicinae Acromyrmex 1 Cephalotes 1 Crematogaster 4 Cyphomyrmex 1 Pheidole 10 Solenopsis 2 Strumigenys 1 Ponerinae Odontomachus 1 Pseudomyrmecinae Pseudomyrmex 1 a b Figure 5. Daily cycle of activity of two Blepharidatta sp. colonies (a, b) at the “Reserva Particular do Patrimônio Natural Serra das Almas”, CE, Brazil. All ants exiting/entering the nest were counted during 10 minutes, during a 24-hour observation period (a: 16.12-17 .12.2011; b: 16.06-17.06.2012). Bars: foraging ants/10 min; line with dots: soil surface temperature. Sunrise: 5.30 h (onset of daylight: 6.00 h); sunset: 17.30 h (beginning of darkness: 18.00 h). The daily cycle of activity of the third monitored nest (16.06-17.06.2012), not shown here, had similar pattern. Sociobiology 61(1): 52-59 (March, 2014) 57 were observed grasping and pulling the appendages of two still alive and fighting Ectatomma muticum workers, trying to transport them to the nest. Only two nests were found in the 144 m2 area mapped (three if a nest close to the area border is included) (Fig 6), giving a nest density of 0.020 nest/m2. Although some foragers were observed foraging at a distance of seven meters from the nest, the most common foraging distances observed were two to three meters (Fig 6). Blepharidatta sp. nest foraging area can therefore be described as a nearly circular area with a radius of ± 2.5m around the nest opening. Discussion There is no doubt that B. conops and Blepharidatta sp. belong to a group of closely related taxa since they share key features of their behavior, ecology and morphology, while at the same time both show fundamental differences with the other Blepharidatta species living in tropical humid environ- ments. B. conops and Blepharidatta sp. are soil-dwelling spe- cies with nest opening surrounded by a carcasses ring (Diniz et al., 1998), while B. brasiliensis and Blepharidatta sp-ba nest in leaf-litter, and do not form carcasses ring (Rabeling et al., 2006; Silva, 2007). B. conops and Blepharidatta sp. are monogynous species whose ergatoid queens have a phrag- motic head (Brandão et al., 2001), while B. brasiliensis and Blepharidatta sp-ba have polygynous colonies, and queens do not have a phragmotic head (Rabeling et al., 2006; Silva, 2007). It is worth mentioning, however, that polygyny or mo- nogyny of Blepharidatta sp-ba remains a controversial ques- tion (Silva, 2007). Frontal disk diameter in queens, head width in work- ers and body length in queens and workers are almost identi- cal in both species (Table 1) (Diniz et al., 1998; Silva, 2007). On the other hand, body length in queens and workers of B. conops and Blepharidatta sp. (± 5 mm for queens; ± 3 mm for workers) is almost three times that observed in B. brasiliensis and Blepharidatta sp-ba (± 2mm and ± 1mm for queens and workers, respectively) (Silva, 2007). Blepharidatta sp. and B. conops have similar diet and foraging activity patterns, with two daily and mostly diurnal peaks of foraging activity separated by periods of inactivity, and a protein diet predominantly made of other ant species (Diniz et al., 1998; Brandão et al., 2008). In B. brasiliensis, foraging activity is predominantly nocturnal (Rabeling et al., 2006). Finally, the two groups of populations are found in completely different environments: rainforests (Amazon and Atlantic Rainforest) for B. brasiliensis and Blepharidatta sp- ba (Rabeling et al., 2006; Silva, 2007); drier and savanna- like environments (Cerrado and Caatinga) for B. conops and Blepharidatta sp. (Diniz et al., 1998; Brandão et al., 2001). According to Silva (2007), B. conops possesses more derived character states (e.g. phragmotic head of queens) when compared to other Blephadatta species. Considering the number of key behavioral and morphological features shared by B. conops and Blepharidatta sp., it can be stated that Blepharidatta sp. shares with B. conops this apomorphic condition. However, our study also shows the existence of many differences between the populations living in the Cerrado biome (B. conops) and the population found in the Caatinga biome (Blepharidatta sp.). The most noticeable are those re- lated to queen morphology. In Blepharidatta sp., frontal disk cuticle is discretely reticulate-punctate, nearly smooth, while it is rugose with a strong relief composed of closely packed polygonal units delimited by ridges in B. conops (Brandão et al., 2001) (Fig 4). Furthermore, other morphological differences, not detailed above, can be seen in the mesosoma sculpture, much stronger in the B. conops queen than in Blepharidatta sp., in the humeral angle of the pronotum which is sharply an- gulated in Blepharidatta sp. but rounded in B. conops. Finally, the pilosity on the first gastral tergite of the queen is erect and short in Blepharidatta sp. while it is subdecumbent and a little longer in B. conops. Blepharidatta sp. and B. conops also differ in relation to nest architecture, colony size, location of queens and brood in nests and foraging behavior. Blepharidatta sp. nests are deeper (± 26.5 cm; up to 40 cm in some nests) than in B. conops (up to 20 cm in mature nests) (Diniz et al., 1998). Conversely, the mean diameter of tunnels and nest opening is much greater in B. conops (2cm versus 0.7cm in Blepharidatta sp.) (Brandão et al., 2001). In most excavated nests of Blepharidatta sp., 12 m 12 m : trail back to the nest with a bait (dead termite) : nest of Blepharidatta sp. Fig 6. Mapping of Blepharidatta sp. nests and of the trails used by foragers to go back to the nest with a bait (dead termite) in a 144 m2 (12 x 12 m) area with a grid of 1 m2 quadrats, at the “Reserva Particular do Patrimônio Natural Serra das Almas”, CE, Brazil, in 01/2012. JC Pereira et al - An enigmatic population of Blepharidatta from northeastern Brazil58 there was one, or more, chambers located close to the nest opening and full of carcasses. The function of those carcasses chambers in Blepharidatta sp. is probably to store prey car- casses before discarding them on the carcasses ring. No such chamber was found in B. conops nests. One of the most im- portant differences in nest architecture between B. conops and Blepharidatta sp. is the so-called subsidiary chamber. In B. conops, all mature nests have a subsidiary chambers used as refuge by the queen and the brood (Diniz et al., 1998; Brandão et al., 2001). Only one Blepharidatta sp. nest had a chamber full with brood, queen and males and whose features suggest a subsidiary chamber. More generally, Blepharidatta sp. nests are more vari- able in structure than B. conops nests. In B. conops nests, there is only one tunnel, always vertical (Diniz et al., 1998). In Blepharidatta sp., the main tunnel varies from vertical to inclined. Furthermore, the main tunnel can have branchings, the result being nests three-dimensionally more complex that those of B. conops. It cannot be excluded that the characteris- tics of the soil where the nests of Blepharidatta sp. were built have a strong influence on nest architecture. However, such characteristics were not analyzed in the present study, nor in studies with B. conops (Diniz et al., 1998). As in other Blepharidatta ants, Blepharidatta sp. colo- nies are small. However, the average number of workers in Blepharidatta sp. colonies (193 ± 107, n=19) is higher than in B. conops (142 ± 57, n=19) (Diniz et al., 1998), B. brasiliensis (132 ± 96, n=13) (Rabeling et al., 2006) and Blepharidatta sp- ba (112 ± 13, n=2) (Silva, 2007). One of the Blepharidatta sp. nests had up to 437 workers. In B. conops, the largest known colony size is 248 workers (Diniz et al., 1998). Interestingly, two queens were found in a single Blepharidatta sp. nest. This observation supports the hypothesis that in B. conops, the foundation of new colonies is through fission of established colonies (Brandão et al., 2001). If this hypothesis is correct, there must be a time when two queens are present in the nest: the old resident queen and a young virgin queen. The location of queen and brood in nests is another difference between B. conops and Blepharidatta sp. In most nests of Blepharidatta sp., the queen and brood were found in the bottom chamber. In B. conops, they are generally found in the subsidiary chamber, but never in the bottom chamber of the nest, which is used as a place for prey dismembring and to temporarily store carcasses, before they are taken out of the nest (Diniz et al., 1998; Brandão et al., 2001). Foraging behavior of Blepharidatta sp. and B. conops is very similar. However, some differences were observed. According to Diniz et al. (1998), the ants and other arthropods visiting the carcasses ring of nests “are frequently ambushed by B. conops workers concealed immediately inside the nest entrance”. Such behavior was never observed in Blepharidatta sp. Another significant difference is the size of foraging area. In Blepharidatta sp. the mean radius of foraging area (2.5 m) is nearly twice that observed in B. conops (1.5 m), while nest density seems to be much lower in Blepharidatta sp. (0.02 nests/m2 against 0.2 nests/m2 in B. conops), at least in the investigated area (Diniz et al., 1998). However, the influence of local factors on nest density, such as prey density or com- petition with other ant species that use similar resources can not be excluded. Although it cannot be excluded that Blepharidatta sp. is a species different from B. conops, it is obvious that both taxa are closely related and constitute a single evolutionary unit, with a branch that spread into the Cerrado biome (sa- vanna) and another into the Caatinga biome (semi-arid). The traits that strongly suggest the occurrence of two distinct spe- cies are seen in the queen morphology, mainly on the frontal disk and mesosoma. On the other hand, it has been suggested that B. conops distribution in the Cerrado biome is fragmented in a range of local populations, with size and sculpture of queen frontal disk varying from one population to the other (Brandão et al., 2001). Therefore, Blepharidatta sp. could be a local variation of B. conops adapted to semi-arid conditions and consequently represent one extreme of the phenotypic variations observed in populations of B. conops. Ant popula- tions in species that have lost mating flight have low levels of gene flow and can evolve independently, as it could happen in fragmented populations of B. conops or its ancestor since the early Quaternary. Future studies should use cytogenetic and molecular tools to compare these Blepharidatta populations (e.g. Mari- ano et al., 2008; Resende et al., 2010 with Dinoponera lucida Emery 1901). Finally, our study extends the known geographic range of the Blepharidatta group and the range of biomes where it is found, as it is the first time that the genus, until now found only in wet forests and savannas (Cerrado), is reported for the Neotropical semi-arid region. Acknowledgments The authors are grateful to Rodrigo dos Santos Macha- do Feitosa (Universidade Federal do Paraná) for the identifi- cation of ant carcasses. We are also grateful to Leila Apare- cida Souza (Universidade Estadual do Ceará [UECE]), Sonia Maria Lopes (Museu Nacional – Universidade Federal do Rio de Janeiro) and Mark Harvey (Western Australian Museum – Australia) for the identification of the myrmecophile fauna (crustaceans, cockroaches and pseudoscorpions, respectively) found in Blepharidatta sp. nests. We also thank UECE for transport facilities as well as the “Associação Caatinga” and the staff of the RPPN “Serra das Almas” for allowing us excel- lent field work conditions. JCP acknowledges her CAPES fel- lowship during her MSc, and JHC his CNPq research grant. References Bolton, B. (1995). Taxonomic and zoogeographical census of the extant ant taxa (Hymenoptera). Journal of Natural History, 29: 1037-1056. doi: 10.1080/00222939500770411. Sociobiology 61(1): 52-59 (March, 2014) 59 Brandão, C.R.F., Diniz, J.L.M., Silva, P.R., Albuquerque, N.L. & Silvestre, R. (2001). The first case of intranidal phragmosis in ants: the ergatoid queen of Blepharidatta conops (Formici- dae, Myrmicinae) blocks the entrance of the brood chamber. Insectes Sociaux, 48: 251-258. doi: 10.1007/PL00001774. Brandão, C.R.F., Silva, P.R. & Diniz, J.L.M. (2008). O “mis-2008). O “mis-O “mis- tério” da formiga lenta Blepharidatta. In E.F. Vilela, I.A. Santos, J.H. Shoereder, J.E. Serrão, L.A.O. Campos & J.L. Lino-Neto (Eds.), Insetos sociais: da biologia à aplicação (pp. 38-46). Viçosa: Editora da UFV. Cassano, C.R., Schroth, G., Faria, D., Delabie, J.H.C. & Bede, L. (2009). Landscape and farm scale management to enhance biodiversity conservation in the cocoa producing region of southern Bahia, Brazil. Biodiversity and Conservation, 18: 577-603. doi: 10.1007/s10531-008-9526-x. Diniz, J.L.M. & Brandão, C.R.F. (1997). Competition for car-Competition for car- casses and the evolution of fungus-ants symbiosis. In J.H.C. Delabie, S. Campiolo, I.C. Nascimento & C.S. Ferreira Ma- riano (Eds.), Anais do VI International Pest Ant Symposium & XIII Encontro de Mirmecologia (pp. 81-87). Ilhéus: Uni- versidade Estadual de Santa Cruz, Ilhéus. Diniz, J.L.M., Brandão, C.R.F. & Yamamoto, C.I. (1998). Bi-Bi- ology of Blepharidatta ants, the sister group of the Attini: a possible origin of fungus-ant symbiosis. Naturwissenschaften, 85: 270-274. doi: 10.1007/s001140050497. Heurtault, J. (1986). Petterchernes brasilienses, genre et espèce nouveaux de Pseudoscorpions du Brésil (Arachnides, Pseu- doscorpionida, Chernetidae), Bulletin du Museum National d'Histoire Naturelle, Paris, 2: 351-355. Mariano, C.S.F., Pompolo, C.G., Barros, L.A.C., Mariano- Neto, E., Campiolo, S. & Delabie, J.H.C., (2008). A bio-A bio- geographical study of the threatened ant Dinoponera lucida Emery (Hymenoptera: Formicidae: Ponerinae) using a cyto- genetic approach. Insect Conservation and Diversity, 1: 161- 168. doi: 10.1111/j.1752-4598.2008.00022.x. Peeters, C. (1991). Ergatoid queens and interacaste in ants: two distinct adult forms, which look morphologically inter- mediate between workers and winged queens. Insectes Sociaux., 38: 1-15. Quinet, Y.P. & Tavares, A.A. (2005). Formigas (Hymenopte- ra: Formicidae) da área Reserva Serra das Almas, Ceará. In F.S. Araújo, M.J.N. Rodal & M.R.V. Barbosa (Eds.), Análise das variações da Biodiversidade do Bioma Caatinga (pp. 329- 349). Brasília: Ministério do Meio Ambiente. Rabeling, C., Verhaagh, M. & Mueller, U.G. (2006). Behav-Behav- ioral ecology and natural history of Blepharidatta brasiliensis (Formicidae, Blepharidattini). Insectes Sociaux, 53: 300-306. doi: 10.1007/s00040-006-0872-y. Resende, H.C., Yotoko, K.S.C., Delabie, J.H.C., Costa, M.A., Campiolo, S., Tavares, M.G., Campos, L.A.O. & Fernandes- Salomão, T.M. (2010). Pliocene and Pleistocene events shap-Pliocene and Pleistocene events shap- ing the genetic diversity within the Central Corridor of Brazil- ian Atlantic Forest. Biological Journal of the Linnean Society, 101(4): 949-960. doi: 10.1111/j.1095-8312.2010.01534.x. Schultz, T.R. & Meier, R. (1995). A phylogenetic analysis of the fungus-growing ants based on morphological charac- ters of the larvae. Systematic Entomology, 20: 337-370. doi: 10.1111/j.1365-3113.1995.tb00100.x. Silva, P.R. (2007). Biologia de algumas espécies de Blephari- datta. Biológico, 69 (suplemento): 161-164. Wheeler, W.M. (1915). Two new genera of myrmicine ants from Brazil. Bulletin of the Museum of Comparative Zoology, 59: 483-491. Wheeler, C.G. & Wheeler, J. (1991). The larva of Blephari- datta (Hymenoptera: Formicidae). Journal of the NewYork Entomological Society, 99: 132-137.