We describe a new species of the bee genus Xenochlora Engel, Brooks & Yanega, 1997, X. meridionalis sp. nov., based on a single female collected in the coastal forests of southeastern Brazil, in the state of Espírito Santo. The disjunct distribution exhibited by Xenochlora, with species in northern South America and in the Atlantic Forest of eastern Brazil, is discussed in light of current knowledge about other taxa with similar distribution pattern.

Augochlorini, Halictidae, Neotropical, taxonomy, systematics

Xenochlora Engel, Brooks & Yanega, 1997 is a rare genus of halictine bees until recently assumed to be restricted to the Amazon Basin and the Guiana shield (Engel et al. 1997, Santos and Melo 2013). The genus comprises four species: X. ianthina (Smith, 1861), X. nigrofemorata (Smith, 1879), X. chalkeos Engel, Brooks & Yanega, 1997 and X. ochrosterna Engel, Brooks & Yanega, 1997. Specimens of Xenochlora are rarely collected, and the entire genus is known only from a few females and a single male of X. nigrofemorata (Santos and Melo 2013). Nests of Xenochlora are built by females in dead wood, notably small dead branches hanging on the understory vegetation, and can be inhabited by single or multiple females (Smith 1861, Tierney et al. 2008). Multifemale nests are regarded to exhibit parasocial behavior, and a temporal switch between semisocial, eusocial and possibly communal phases was also proposed to occur within the genus (Tierney et al. 2008).

Within the Augochlorini, Xenochlora has been considered as closely related to Megalopta Smith, 1853 (Engel et al. 1997, Engel 2000, Tierney et al. 2012, Gonçalves 2016). In a recent molecular and morphological phylogenetic analysis of the extant Augochlorini, Megalopta was recovered as a paraphyletic group, as Megalopta atra Engel, 2006 appears in a clade with Xenochlora, forming the sister group of the other species of Megalopta (Gonçalves 2016). Reconstructing the phylogenetic history of Xenochlora, Megalopta, Megaloptidia Cockerell, 1900, Megommation Moure, 1943 and related genera is essential to understand the evolution of dim-light foraging in the tribe Augochlorini, which yet is possible to be interpreted in different ways (see Tierney et al. 2012, Gonçalves 2016).

In 2013, during fieldwork of a study on orchid bees at Reserva Natural Vale (Coswosk et al. 2018), one female of Xenochlora was unexpectedly caught in the scent traps. A preliminary comparison with available material suggested that it belonged to an undescribed species. More recently, Nemésio et al. (2016) recorded the genus in the same locality, Reserva Natural Vale, from the study of a female previously deposited in the Coleção Entomológica da Reserva Natural Vale (CERNV). They even considered the possibility of this individual to belong to a new undescribed species.

We herein describe a new species of Xenochlora from a large remnant of Atlantic Forest, collected in 2013. This is the first species occurring outside of Amazonian Formations at northern South America.

The holotype was collected in the Reserva Natural Vale, a large remnant of Brazilian Atlantic forest, encompassing ca. 22,700 ha in northern Espírito Santo state, southeastern Brazil (Lopes and Mello-Silva 2014). This reserve and Reserva Biológica Sooretama (ca. 24,000 ha) are recognized as the last Tabuleiro Forest massif in northern Espírito Santo (Amorim 1984, Silva 2014). Tabuleiro forests (Floresta Ombrófila Densa de Terras Baixas sensu Veloso et al. 1991) have distinctive physiognomic and floristic characteristics (e.g., Peixoto et al. 2008) and share many plant taxa with the Amazon Forest, including several species with disjunct distribution between Amazon and Atlantic forests (reviewed in Rolim et al. 2016).

The terminology for the external morphology follows Eickwort (1969), Engel (2000) and Michener (2007), except for the scutum, referred to as mesoscutum, and the “basal area of propodeum”, here the metapostnotum, following Brothers (1976). The following abbreviations are used: F1–F11 for the flagellomeres; T1–T6, metasomal terga; and S1–S8, metasomal sterna. The punctation density is described in relation to the interspaces between punctures, measured in terms of puncture diameter (pd). The information provided between quotation marks in the type material section is an exact transcription of the label associated with the specimen; the inverted bars (\) indicate different lines in the same label. The holotype is deposited in the Coleção Entomológica Padre Jesus Santiago Moure, Departamento de Zoologia, Universidade Federal do Paraná, Curitiba, Brazil (DZUP).

The occurrence of Xenochlora meridionalis sp. nov. in the Atlantic forest supports the already well-established pattern of disjunct distributions between this forest formation of eastern Brazil and the Amazon basin for some plants and animals (e.g., Vanzolini and Williams 1970, Costa 2003, Martini et al. 2007, Fiaschi and Pirani 2009, Salles et al. 2010, Tonini et al. 2010, Batalha-Filho et al. 2013). Past connections between Amazon and Atlantic forests through central areas of South America are highly supported by independent evidence (e.g., Auler et al. 2004, Wang et al. 2004, Sobral-Souza et al. 2015).

In addition to the distributional pattern of Xenochlora, other bees also provide good examples of disjunct taxa where the Atlantic morphs closely resemble their Amazonian counterparts. Orchid bees (Apidae, Euglossini) deserve special attention in this scenario when taking into consideration the following reasons (i) they are closely related to forested areas (e.g., Dressler 1982), (ii) they are well sampled when compared to other groups of bees, mainly in the Atlantic forest (e.g., Sydney et al. 2010), and (iii) there have been recent revisions at species level for the fauna (e.g., Nemésio 2009, Faria and Melo 2012). Some examples where a species inhabits the Amazon Basin and its putative sibling is found in the Atlantic forest include (i) Euglossa parvula Dressler, 1982 and Eg. pepei Nemésio & Engel, 2012, (ii) Eufriesea ornata (Mocsáry, 1896) and Ef. atlantica Nemésio, 2008, and (iii) Exaerete trochanterica (Friese, 1900) and Ex. salsai Nemésio, 2011. It is important to highlight that phylogenetic analyses and divergence times of these pairs of species are not available, so their closer relationship is hypothesized based only in the taxonomic interpretations of the authors. But phylogenetically supported evidence also exists. When studying the phylogeography of some species of Eulaema Lepeletier, 1841, López-Uribe et al. (2014) found that the individuals from Atlantic forest are closely related to their counterparts in the Amazon Basin from both species complexes involving respectively Eulaema meriana (Olivier, 1789) and Eulaema bombiformis (Packard, 1869). They also found that the separation between the Amazonian and Atlantic forms are somewhat recent, ca. 1.0–2.0 Mya in both species complexes.

Nonetheless, unlike the examples presented above, the distribution patterns of some other bee taxa raise the following question: why some Amazonian groups have never crossed the forest bridges? It is remarkable in this context of faunal exchange that some particular genera of stingless bees as Aparatrigona Moure, 1951, Duckeola Moure, 1944, Nogueirapis Moure, 1953 and Ptilotrigona Moure, 1951, cleptoparasites as Aglae Lepeletier & Serville, 1825 and Aglaomelissa Snelling & Brooks, 1985 and other solitary bees as Xylocopa (Xylocopina) Hurd & Moure, 1963 have not been found in the Atlantic forest. Finding an answer to this question is certainly not a simple task, but the bionomics and phylogeny of taxa with disjunct or restricted distribution, both in Amazon basin or Atlantic forest, may shed light on the relation between past environmental filters and the functional traits of species with these distributional patterns. In this sense, the pattern exhibited by the stingless bees, whose newly founded nests depend of a nearby mother nest during their early beginning, suggests that the bridges between the Atlantic and Amazonian forests probably formed a fragmented landscape and not broadly continuous forested areas.

The southeastern-northwestern bridge (SE–NW; Bigarella et al. 1975) was proposed based on the distribution of taxa that inhabit western Amazonia and southeastern Atlantic Forest but not northeastern areas of this latter formation (e.g., Costa 2003) and have been regarded as the most important route between these forested biomes (Por 1992). This route seems to be the most ancient connection between southeastern Atlantic and Amazon forests and was established since the Andean uplift and the formation of the South American monsoon system (Ledo and Colli 2017). Batalha-Filho et al. (2013) presented data supporting two distinct pathways connecting both biomes, (i) more ancient connections (middle to late Miocene) through the current southern Cerrado and Mato Grosso and the transition towards the Chaco and palm savannas of Bolivia and Paraguay, and (ii) more recent connections (Pliocene and Pleistocene) through drier formations (Cerrado and Caatinga) in northeastern Brazil. Molecular dating of splits between sister lineages (one in Amazon and one in Atlantic forest) suggests the events related to the SE-NW bridge are more frequent and older than those related to the NE route (Ledo and Colli 2017).

Xenochlora meridionalis sp. nov. brings a new piece to the Amazon-Atlantic forest connection puzzle, and we hypothesize that Xenochlora reached the Atlantic forest through the SE–NW bridge, since so far it has been collected only in eastern Brazil. Considering the recent discovery of Aglae in forest fragments in Goiás, central Brazil (Silva et al. 2015), it is possible that more intensive sampling in this region may reveal the southernmost records for Amazonian species of Xenochlora. Furthermore, the general morphological similarity between Xenochlora meridionalis sp. nov. and their Amazon counterparts (mainly X. nigrofemorata and X. ianthina) suggests that the occurrence of this bee genus in the Atlantic Forest derives from a more recent faunal exchange, in the same way as with several pairs of species of orchid bees, dated from the late Pliocene-Pleistocene (see molecular dating of euglossine evolution in Ramírez et al. 2010 and López-Uribe et al. 2014).

We acknowledge Maria Cecília Martins Kierulff for all the support during fieldwork, and Vale company/Reserva Natural Vale administration for allowing us to sample the bees of the preserve. J.A. Coswosk, E.D.G. Soares, R.A. Ferreira, A. Storari and F.F. Salles are acknowledged for all the help with fieldwork. We also thank Eduardo A.B. Almeida and Ângelo P. Pinto for reviewing the manuscript and for suggestions that helped improve it. SISBIO/ICMBio is acknowledged for the provided collecting permit (#21803). Partial support has been provided by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq 309641/2016-0) to GARM. All authors revised the text and are solely responsible for accuracy in use of the English language.