A new species of Centris ( Centris) (Fabricius) from northeastern Brazil, with taxonomic notes on C. ( C.) pulchra Moure, Oliveira & Viana (Hymenoptera, Apidae)

Abstract We describe a new species of the bee genus Centris, Centris (Centris) byrsonimae Mahlmann & Oliveira sp. n., whose name has appeared as a nomen nudum in the literature since 1985. Further, a new species group of Centris s.str. is proposed, the pulchra group, based on morphological characters, which comprises the species Centris pulchra Moure, Oliveira & Viana, 2003 and Centris byrsonimae sp. n..Based on information from specimen labels studied and data from the literature, a list of plant species visited by the pulchra group is presented. The male genitalia and hidden metasomal sterna 7 and 8 of Centris pulchra are described for the first time. Typographic errors pertaining to the paratype labels reported in the original description of Centris pulchra are corrected. One female paratype of Centris pulchra is designated herein as a paratype of Centris byrsonimae sp. n. An updated list of species of Centris s.str. from northeastern Brazil is provided including references about geographic distributions as well as an identification key to the pulchra species group.


introduction
The bee tribe Centridini (Apidae: Apinae) contains numerous robust, large to moderate-sized species, most of which are Neotropical and collect floral oils, principally from Malpighiaceae and Krameriaceae (Ayala 2002). A subgeneric classification based on a cladistic analysis was proposed by Ayala (1998), who recognized 12 subgenera in Centris Fabricius, two of which were described subsequently by Ayala (2002). According to Moure et al. (2012), the subgenus Centris is one of the most diverse containing 34 described species. According to Michener (2007), this subgenus occurs from Baja California, Mexico and southern Arizona to southern Florida, USA, and the Bahamas, and south through the Antilles and the continental tropics to Santa Catarina, Brazil. Only one species of the subgenus is exclusive to the Nearctic region, while six occur in both the Nearctic and Neotropical regions, and the remaining 27 species are exclusively distributed in the neotropics (Moure et al. 2012). In northeastern Brazilian nine nominal valid species are known for the subgenus (Table 1).  C. flavifrons (Fabricius, 1775) MA Albuquerque (1986) BA Vivallo and Zanella (2012) ‡ SE Silveira and Mendonça (2005) C. nitens Lepeletier, 1841 BA Viana and Santos (2002) SE Silveira and Mendonça (2005) C. pulchra Moure, Oliveira & Viana, 2003BA Viana (1999; Moure et al. (2003) § ; Moure et al. (2012) The present paper describes a new species of Centris s.str. from northeastern Brazil. Centris byrsonimae sp. n., whose epithet has appeared as a nomen nudum in the literature since 1985, is described and figured herein, finally validating the name for this species. Based on morphological characters, a new species group of Centris s.str. is proposed, the pulchra group, which currently comprises C. pulchra and C. byrsonimae sp. n. In addition, we provide an identification key to this group. A list of plant species visited by the pulchra group is provided based on information from specimen labels and a survey of the literature. The male terminalia of C. pulchra is described for the first time. Typographic errors reported from the paratype labels of C. pulchra in its original description were discovered during the course of this work and herein they are corrected. One female paratype of C. pulchra is designated as a paratype of C. byrsonimae sp. n.

Methods
We examined 30 paratypes (24♀♀, 6♂♂) and additional specimens of C. pulchra deposited in the Entomological Collection, Zoological Museum, Federal University of Bahia (MZUFBA), Salvador, Bahia, Brazil. In addition we studied the specimen originally examined by Pe. Moure in 1985 and labeled by him as holotype of C. byrsonimae (DZUP: Federal University of Paraná, Curitiba, Paraná, Brazil) but until now undescribed. Additional material was borrowed from the Entomological Collection of the Studies on Bees Laboratory (LEACOL), Federal University of Maranhão, São Luís, Maranhão, Brazil. General morphological terminology follows Engel (2001) and Michener (2007), with the standard abbreviations as follows: F1, F2, etc.-antennal flagellomeres; T1 to T7 and S1 to S8 -metasomal terga and sterna, respectively. In addition we have used the abbreviation SL for scape length. The upper and lower interocular distances were measured using the shortest distance between the compound eyes in frontal view. The mandibular teeth were numbered from the apex to the base of the mandible. All measurements are given in millimeters (mm). Label information from separate labels are segregared by double slashes,"//". Typographic errors from paratype labels as reported in the original description of C. pulchra were corrected and the corrections identified with brackets, "[ ]". Floral records for pulchra species group were based on information from specimen labels and the literature. Additionally, a label reporting the host plant of the holotype of C. byrsonimae sp. n. was included. Photomicrographs were prepared using a Leica M165C stereomicroscope coupled with a Leica DFC295 and a Leica Application Suite V4. Pulchra species group Diagnosis. Herein we propose the pulchra species group as a distinct lineage within Centris s.str., comprising for the moment two species: C. pulchra and C. byrsonimae sp. n. This species group is characterized by the clypeus largely yellow with two longitudinal dark brown streaks located in the upper half; a narrow band of whitish and dense bristles bordering the posterior border of T2-T4, extended to the sides and narrowed in the middle, but not interrupted; by the maxillary palpus reduced to four palpomeres; the mandibles with five teeth, the most basal substantially smaller and located in the lower inner edge of the mandible; and the secondary basitibial plate yellow, but only slightly elevated (merely inflated), not projecting over the basal plate.
Comments. In the identification keys of Ayala (in Michener 2007) and Silveira et al. (2002), the species of the pulchra group do not fit well into Centris s.str. based on four characters: (1) the maxillary palpus is 5-or 6-segmented, (2) the mandible 3-or 4-toothed, (3) the basitibial plate with a defined secondary plate with sharp projecting margins, and (4) the clypeus with yellow or white markings in the form of an inverted T or Y. Although the pulchra group differs in these four characters and indicated in the key as diagnostic for the subgenus Centris, C. pulchra and C. byrsonimae sp. n. can be retained there based on the unique character of the subgenus: the long, slender, apical projection of the male gonocoxite, extending parallel to the gonostylus (Michener 2007); this gonocoxal projection bearing giant branched setae. In regard to parasitism by Mesonychium asteria (Smith, 1854) (Apinae: Ericrocidini) in nests of C. pulchra reported by M.C. Ramos (pers. comm. in Rocha-Filho et al. 2009), this was probably a mistake. According to Marina C. Ramos (pers. comm. 2012) the parasitism observed was of an unidentified species of Mesoplia Lepeletier de Saint Fargeau (Apinae: Ericrocidini) in nests of C. byrsonimae sp. n. She observed adults of Mesoplia sp. emerging from brood cells of C. byrsonimae sp. n. nesting in the ground.  Diagnosis. This species is quite similar to C. (C.) pulchra but differs by its smaller body size (ca. 10.6); the largely honey-brown integument (Figs. 2, 7); metallic blue iridescence almost imperceptible on metasomal terga; T1-T5 interlay covered by pale pubescence including the yellowish pubescence, largely golden on apical margin of T5 (Fig. 7); inner surface of probasitarsus with the combs for collecting floral oil, on distal half of posterior margin with three distinct strongly spatulate setae, curved on the inner surface and wider on its apex (Fig. 6); and the gonocoxal projection covered by dense, giant branched setae on the inner surface, forming a distinct fringe of plumose hairs longer than the gonocoxal projection itself .
Coloration: Integument predominantly honey brown with metallic blue highlights almost imperceptible on metasoma (Fig. 7); head and mesosoma dark brown, mesoscutellum brownish. Yellow marks: clypeus, except two divergent wide dark brown stripes, located on upper half of clypeus and above separated from each other by distance equivalent to lateral ocellar diameter; labrum; basal 2/3 of mandibles; malar area; very short narrow triangle on supraclypeal area; paraocular stripe ventrally wider, widest at level of tentorial fovea, narrowing upward and gradually ending at median level of compound eyes; wide stripe on lateroposterior surface of scape; small yellow marks on base of tibiae, most developed on secondary basitibial plate on metatibia; on protibia yellow marks followed by long dorsolateral stripe. Tegulae translucent honey colored; wing membrane slightly yellowish, with little brownish streak in basal third of marginal cell, with very thin arched transverse vein on apical third of first submarginal cell; brown venation slightly yellowish at pterostigma and base of wings; primary basitibial plate honey translucent; metallic blue reflections almost imperceptible on T2-T5.
Pubescence: In general pale and yellowish, with mostly whitish hairs as follows: lower surface of face and gena, ventral portion of mesepisterna, metepisterna, propodeum, base of metasoma and sterna; reddish bristles only on inner surface of tarsi, mostly visible on basitarsi; mesepisterna with plumose hairs restricted to upper half, dorsoventral surface covered by simple setae more widely spaced; lower 2/3 of ventral surface of procoxae covered by extremely thick, curved, yellow-whitish setae, decreasing in length toward posterior border, upper portion of procoxae covered by plumose hairs; outer surface of mesocoxae covered by simple bristles, upper margins with plumose hairs 0.5x smaller; probasitarsus with long line of bristles (2× SL) on outer dorsolateral surface, shortening toward apex and interspersed by a dense band of short plumose hairs, inner surface with few short bristles sparsely distributed; inner surface of probasitarsus, on distal half of posterior margin, with combs for collecting floral oils with three distinct and strongly spatulate setae, curved toward inner surface and wider at apex (Fig. 6); outer surface of mesobasitarsus with dense, short, plumose pubescence interspersed with simple bristles longer and sparser; metatibiae and metabasitarsus with thicker, milky white bristles; T2-T4 entirely covered by pale yellow, short and sparse setae, longer and denser on sides of terga and near posterior border, contrasting with hairiness of discal terga but not forming distinct band of yellowish bristles (Fig. 7); apical margin of T5 with long fringe of golden branched bristles (0.5× SL); S2-S5 with dense apical fringe, with hairs of central stripe slightly longer; fringe of S5 curved and with rachis relatively longer.
Sculpturing: Dense but of moderate size: on vertex spacing 1/2 of puncture width, with cariniform spaces between punctures, slightly smoother on supraclypeal area; midline of clypeus with integument smooth and polished, interspersed by very few punctures; punctures most evident on the side slopes of clypeus and labrum. Punctures obscured by pilosity on mesosoma and metasoma, mostly fine and spaced on sides of propodeum, and finer and piligerous on terga.
♂: Structure: Total body length 10.8; forewing length 7.7; head length 2.7, width 4.1; clypeus length 1.2, width 1.7; labrum length 0.8, width 1.2; scape length 0.6; F1 length 0.7; F2 length 0.2; F3 length 0.3; diameter of the anterior ocellus 0.3; ocellocular distance 0.3 (1.3× lateral ocellar diameter); upper interocular distance 2.0; lower interocular distance 2.3; metasoma width 4.2 (measured on T2). Similar to female except as follows: mandibles with three normal teeth and one basal denticle on inner margin; basitibial and pygidial plates absent. S7 and S8 as in figure 22: S7 about as long as wide as measured at base, approximately 2/3 of S8 length and covered by setae on its lateral areas and apex; apical margin of S7 indented; median projection of S8 slightly uniformly tapering to apex; dorsal surface and apical 1/3 of S8 covered by long branched setae, apex covered by simple, shorter erect setae. Genitalia as in figures 19-21: dorsal surface of gonocoxite with small basal edge, internal surface covered by long, dense pilosity; gonocoxal projection covered by dense, long branched setae on inside surface forming distinct fringe of plumose hairs longer than gonocoxal projection itself; gonostylus covered by short setae.
Pubescence: Similar to female but mesepisterna fully covered by dense plumose hairs; ventral surface of pro-and mesocoxae covered by plumose hairs only, procoxae with plumose hairs markedly longer, denser and recurved; probasitarsus without long line of long bristles; inner surface of probasitarsus without combs for collecting floral oils; outer surface of mesobasitarsus without velvet, dense, short plumose pubescence; external surface of metabasitarsus covered by dark-brown setae; base of metapretarsal claws with pair of thin, simple setae almost as long as pretarsal claws; pilosity of T1 denser and longer than remaining terga; T2-T5 with band of yellowish setae more distinctive and wider than in female (Fig. 12); margin of T5 without long fringe of golden branched bristles.
Sculpturing: Midline of clypeus with integument mostly smooth and polished, interspersed by very few punctures.
Etymology. The specific epithet was originally given by Dr. Jesus S. Moure in 1985 when he identified the material collected by Dra. Patrícia Maia C. de Albuquerque for her dissertation, using the generic name of the most common host plant for this bee species (Byrsonima crassifolia L. Rich: Malpighiaceae). Unfortunately, while Moure placed the name on labels of specimens he never published a description of the species or put into writing those characters from which he based his conclusions. In 2003, after the publication of C. pulchra, Moure invited the second author of the present paper to describe with him this new species after a restudy of all paratypes of C. pulchra and other specimens from Maranhão deposited in LEACOL, including the male specimens, are re-examined. However, Moure died in June 2010 before the paper was completed. Thus, despite the fact that he recognized the novelty of the species, his name is not included as a coauthor given that he did not review or approve any version of this manuscript. Nonetheless, we retained the name he intended as well as the specimen he selected as the holotype. Such an epithet also seems to be a good choice given that it has been used before in the literature.
Comments. Although C. byrsonimae sp. n. is quite similar to C. pulchra, it can be differentiated from that species by the integument predominantly honey-brown (mostly dark brown in C. pulchra); C. byrsonimae with metallic blue highlights almost imperceptible on metasomal terga, marked in C. pulchra. The pubescence in general is yellowish, especially on the disc of the terga and the apex of T5, with an apical fringe of golden bristles at the apex of T5; in C. byrsonimae sp. n. T2-T4 are entirely covered by pale, short yellow and sparse setae on the disk of the terga but not forming a distinct band of bristles on the posterior border, while in C. pulchra the discs of T2-T4 are covered by dark-brown setae contrasting with the narrow and dense bands of whitish bristles bordering the posterior edge of the terga extending to the sides and narrowed in the middle, but not interrupted, and contrasting also with the hairiness of the disc of the terga. On the mesepisterna of C. byrsonimae sp. n. the plumose hairs are restricted to the upper half, while in C. pulchra they are distributed on the dorsoventral surface; the lower 2/3 of the ventral surface of the procoxae are covered by extremely thick, curved yellow-whitish setae (the upper portion covered by plumose hairs) in C. byrsonimae sp. n. while in C. pulchra this structure has only plumose hairs; the inner surface of the probasitarsus with secondary combs for collecting floral oils is modified, with three distinct, strongly spatulate setae in C. byrsonimae sp. n. while in C. pulchra these setae are thick but not modified (Figs. 6, 16); the pygidial plate of both species is doubled but in C. byrsonimae sp. n. the upper plate is relatively longer (Fig. 8); the lower plate of the pygidial plate is slightly depressed from the sides to the middle and flat on the apical third contrasting with C. pulchra in which this depression is notably more pronounced and convex in the apical third (Fig.18). The male of both species are quite similar but in C. byrsonimae sp. n. the pubescence is markedly yellowish while in C. pulchra it is whitish; the metallic highlights are most visible and bluish in C. pulchra and almost imperceptible and greenish in C. byrsonimae sp. n.; the base of the metapretarsal claws have a pair of thin simple setae in C. byrsonimae sp. n. while in C. pulchra there is a tuft of long dense plumose hairs. In addition, the male terminalia is quite different as shown in figures 19-26.
Remarks. One female specimen originally designated by Moure et al. (2003) as paratype of C. pulchra is actually an individual of the species described above as C. byrsonimae sp. n. and is selected as a paratype of that species ( Moure, Oliveira & Viana, 2003). Diagnosis. Total body length 13.9. Integument brown-blackish, except legs and metasoma ferruginous; metallic blue highlights on T2 -T5, mostly marked on T5 (Fig. 17); distal edge of terga with a band of whitish setae; most part of basal area of T2-T4 with black bristles. Inner surface of probasitarsus with combs for collecting floral oil, on distal half of posterior margin, with setae unmodified (Fig. 16). Male geni-talia with pubescence relatively spaced, apical projection of gonocoxite with a fringe of short, unbranched bristles, about 1/2 length of projection itself (23-25).
Male S7, S8 and genitalia as in figures 23-26: S7 notably wider than long, approximately 1/5 of length of S8; edge of S7 strongly angular, forming two distinct lobes, these covered with erect bristles; median projection of S8 with a strong median strangulation, apical 1/3 enlarged; dorsal surface of S8 covered by erect bristles, apical 1/3 covered by longer branched bristles curved laterally, apex with shorter simple erect bristles. Dorsal surface of gonocoxite with a large basal edge strongly keeled, internal border covered by short dense pilosity; apical projection of gonocoxite relatively short, its apex distant from gonostylus apex and covered internally by a fringe of simple setae shorter than projection; gonostylus covered by short bristles sparsely distributed (Figures 23-25).
Floral records. Integument predominantly dark brown to black (Figs. 14, 17); inferior 2/3 of ventral surface of procoxae with tufts of long plumose hairs, interspersed by very thin, long, simple hair, superior portion of coxae practically glabrous; discs of T2-T4 covered by dark brown to black setae, contrasting with whitish fringe of hairs on posterior border; probasitarsus with secondary unmodified combs for collecting floral oil (Fig. 16)  -Integument predominantly honey brown (Figs. 2, 7); inferior 2/3 of ventral surface of procoxae covered by extremely thick, curved, yellow-whitish setae, decreasing in length toward posterior border, superior portion of coxae covered by plumose hairs; T2-T4 entirely covered by yellowish short and sparse setae most longer and denser on sides of terga and near posterior border, contrasting with hairiness of disc of terga but not forming a distinct band of whitish bristles; probasitarsus with secondary combs for collecting floral oil modified, with three distinct strongly spatulate setae, curved toward inner surface and with enlarged apex (Fig. 6)