NORTH-WESTERN JOURNAL OF ZOOLOGY 12 (2): 230-238 ©NwjZ, Oradea, Romania, 2016 Article No.: e161202 http://biozoojournals.ro/nwjz/index.html

Orchid bees (, ) from the Brazilian savanna-like ‘Cerrado’: how to adequately survey under low population densities?

André NEMÉSIO

Instituto de Biologia, Universidade Federal de Uberlândia. Rua Ceará, S/N, Campus Umuarama, Uberlândia, MG. 38400-902. Brazil. E-mail: [email protected]

Received: 04. January 2016 / Accepted: 25. May 2016 / Available online: 26. June 2016 / Printed: December 2016

Abstract. Orchid-bee (Hymenoptera: Apidae: Euglossina) faunas from xeric areas have been reported as presenting unusually low abundance, richness and diversity when compared with those from Neotropical forests. Traditional orchid-bee inventories are based on the use of synthetic scents to attract males once or twice a month during a whole year. This protocol presents some potential problems, such as the costs of field sampling, and the uneven distribution of these through the year, particularly when areas under strong precipitation seasonality are involved. Recently, a rapid sampling protocol aimed to minimize the afore- mentioned potential problems has been proposed for densely forested areas. It consists of using as many scents as possible during a 20-hour period during the season when orchid bees are most actively foraging. In the present study this rapid sampling protocol was carried out at ‘Estação Ecológica do Panga’, a patch of the savanna-like ‘Cerrado’ in Uberlândia, Central Brazil, where several samplings, using different protocols, have been conducted before, making direct comparison possible. Four hundred and thirty-one specimens belonging to 11 species were sampled. Abundance, richness, and diversity recorded through the rapid sampling protocol were the highest ever recorded for this single area. The rapid sampling protocol, originally designed to be used in forested areas, revealed to be even a stronger tool in xeric environments.

Key words: bait scents, Cerrado, euglossine bees, , gallery forest, Hexapoda, sampling method.

Introduction nators in Neotropical forests, especially due to their ability of flying long distances in short times Bees present their highest diversity in warm, tem- (e.g., Janzen 1971, Dressler 1982, Roubik & Ac- perate, and xeric environments, showing a de- kerman 1987, Roubik & Hanson 2004, Pokorny et crease towards the humid tropics (Michener 1979). al. 2014), a similar role in xeric areas should also Orchid bees (Hymenoptera: Apidae), however, are be expected. Regrettably, there are only a few in- a particular group of Neotropical bees showing ventories and ecological studies of orchid-bee fau- the opposite trend, i.e., they are typical forest in- nas from xeric environments in Eastern Brazil sects reaching the highest diversity in the Amazon (Nemésio & Faria Jr. 2004, Alvarenga et al. 2007, Basin and the Evergreen forests of Central Amer- Freitas 2009, Silveira 2010, Faria & Silveira 2011, ica (Dressler 1982, Roubik & Hanson 2004, Andrade et al. 2012, Justino & Augusto 2012, Vi- Nemésio & Silveira 2007). They are conspicuous otti et al. 2013). In general, fieldwork is somewhat due to their vivid metallic colours and the re- disappointing to researchers, who spent hundreds markable behaviour of the males, which actively of hours in the field to sample only a few bees collect floral fragrances in hundreds of plant spe- (e.g., Faria & Silveira 2011), whereas a similar cies, notably orchids (Vogel 1966, Dressler 1982, sampling effort in forest areas is enough to sample Roubik & Hanson 2004, Perger 2015). Although several hundreds and even thousands of orchid most species are endemic in forested areas (e.g., bees (e.g., Roubik & Ackerman 1987, Oliveira & Morato et al. 1992, Tonhasca Jr. et al. 2002, Campos 1996, Tonhasca Jr. et al. 1992, Nemésio Nemésio & Silveira 2006b, Nemésio 2011c, 2012c, 2013a-d). Nemésio et al. 2016), a few species seem to suc- In order to adequately study diversity and dis- cessfully exploit xeric environments, but abun- tribution of species it is essential to correctly sur- dance and diversity in these xeric areas have been vey them, in such a way the samples reflect spe- usually reported as low or very low (e.g., Al- cies distributions in nature with the highest fidel- varenga et al. 2007, Faria & Silveira 2011, Andrade ity possible. Unfortunately, this is not a simple et al. 2012, Justino & Augusto 2012, Viotti et al. task, since different organisms are sampled under 2013). different protocols, and some of these protocols As orchid bees play an important role as polli- may present inadequate or biased effectiveness. Rapid sampling of orchid bees in Central Brazil 231

This is not different concerning orchid bees. The (Nemésio 2012b, Nemésio & Paula 2013), and Pe- most popular sampling protocol used in ecological ruvian Amazon (Nemésio & Rasmussen 2014). It is studies on orchid bees is based on the intensive not known, however, whether this protocol would use of synthetic scents that mimic the floral fra- be useful in more xeric environments of Eastern grances attractive to orchid-bee males (e.g. Ac- Brazil, such as the Brazilian savanna-like ‘Cer- kerman 1983, 1989, Pearson & Dressler 1985, Pow- rado’, the high altitude ‘Campos Rupestres’ (rocky ell & Powell 1987, Armbruster 1993, Milet- fields), the ‘Caatinga’, and patches of gallery for- Pinheiro & Schlindwein 2005, Nemésio & Silveira ests that happen to occur immersed in the previ- 2006a, b, 2010, Abrahamczyk et al. 2011, Nemésio ous vegetation types. & Vasconcelos 2013). Although highly effective, Thus, the main goal of this study was to com- the sole use of synthetic scents may potentially in- pare the results obtained from samplings carried troduce some artifacts, since some groups of or- out exactly at the same site through different bait- chid bees, such as the species of Euglossa (Eu- ing protocols – the traditional year-round inven- glossella) and some species of Eulaema (Eulaema), tory, with monthly samplings – with the 20-hour are poorly or not attracted to the scents commonly sampling protocol proposed by Nemésio (2010b) used in orchid-bee samplings (e.g., Moure 1996, to assess if the latter protocol can be confidently Nemésio & Silveira 2004, 2006c, Nemésio 2012a). used in more xeric areas where orchid-bee abun- Despite the potential pitfalls, the use of artifi- dance is known to be low. cial scents has gained in popularity and, in fact, most of the current information on orchid bees has been obtained through this protocol (see Nemésio Material and Methods 2012a). Nevertheless, there is a strong lack of standardization among studies, as first noticed by Site description This study was conducted at ‘Estação Ecológica do Panga’ Morato (1998). Particularly, the number of scents (EEP), a protected forest remnant with a total area of and the periodicity of collections greatly vary about 404 ha, belonging to the ‘Universidade Federal de among studies, raising questions on how compa- Uberlândia’, in the municipality of Uberlândia, state of rable results obtained from different protocols can Minas Gerais, South-Eastern Brazil (19°10’S, 48°23’W, 800 be (Morato 1998, Nemésio & Silveira 2007, Sydney m a.s.l.). The region is characterized by a subtropical cli- et al. 2010). mate with two well-defined seasons: a dry winter (May to It has been suggested that surveys of orchid- September) and a rainy summer (October to April). The mean annual temperature and precipitation are 22ºC and bee males with scent baits during a single day 1650 mm, respectively. Soils at the site are primarily red have great utility, and may reveal almost as much latosols that vary from moderately to strongly acidic about local community structure as studies lasting (Embrapa 1982). Most of EEP is covered with ‘Cerrado’ a full year (Roubik, 2004). Recently, Nemésio physiognomies – from which ‘Cerrado’ sensu stricto is the (2010a, b) proposed a 20-hour sampling protocol dominant one –, but forest physiognomies such as gallery which basically consists in actively collecting bees forests are also found, especially along the stream valleys (Ribeiro & Walter 1998, Lopes & Vasconcelos 2008, Car- with nets in a given site during 20 hours for doso et al. 2009). For this reason, it has been described as two to four days (consecutive or not) in any period one of the best-preserved ‘Cerrado’ sites in South-Eastern from 7am to 5pm using as many scents as possi- Brazil (Costa & de Araújo 2001, Bruna et al. 2010). ble. This protocol should be preferably used dur- Two sites at EEP were sampled: site-1 (19º10’51”S – ing the rainy summer in the Neotropics, the sea- 48º23’44”W, 800 m a.s.l.), a gallery forest patch entirely son when orchid bees are most actively foraging. surrounded by ‘Cerrado’ sensu stricto situated at the Some of the potential benefits of this method are: southern portion of EEP, and site-2 (19º10’09”S – (i) reducing the costs of fieldworks carried out in 48º23’27”W, 800 m a.s.l.), another patch of gallery forest surrounded by ‘Cerrado’ sensu stricto and pastures at the remote areas; and (ii) avoiding over-killing orchid northern portion of EEP. bees. However, the above protocol was designed taking into consideration the high abundance of Sampling orchid bees usually found in densely forested ar- Samplings were carried out during 20 hours from late eas. It was successfully used in more than twenty November, 2012 to January, 2013 in each site, when or- areas in North-Eastern Brazil (Nemésio 2010b, chid bees are usually most actively foraging in the region. Nemésio & Santos Júnior 2014), southern Bahia At site-1, samplings were carried out from 08:30h to 14:30h on the 29th and 30th of November, 2012, from 9am (Nemésio 2011a, 2013a, c, d), northern Espírito to 2pm on the 4th of December, 2012, and from 9am to Santo (Nemésio 2011b, 2013b), Minas Gerais noon on the 24th of January, 2013. At site-2, samplings 232 A. Nemésio. were carried out from 8:30am to 2:30pm on the 13th and and Justino & Augusto (2012), who sampled EEP simul- 14th of December, 2012, from 8:30am to 1:30pm on the 18th taneously from December, 2006 to April, 2007, were inde- of December, 2012, and from 9am to noon on the 25th of pendently obtained. For the purpose of this study, they January, 2013. All samplings were carried out under simi- were assumed to be independent data. The similarity in lar climatic conditions, i.e., sunny days. orchid-bee fauna composition among all samplings (2004, Seventeen different scent baits, made of cotton swabs 2006/2007, and 2012/2013) was estimated by the soaked with one of the substances below, were placed ca. Renkonen similarity coefficient, as recommended by two meters apart from each other at about 1.5 m above Wolda (1981) for small samples. The use of similarity in- the ground. The substances used were: benzyl acetate, dexes that take into account not only presence of species, benzyl alcohol, r-carvone, 1,8-cineole, p-cresol acetate, but also their relative abundances, is strongly recom- dimethoxybenzene, eugenol, -ionone, methyl benzoate, mended (Balmer, 2002). methyl trans-cinnamate, heneicosane, linalool, methyl salicylate, skatole, tricosane, p-tolyl acetate, vanillin. Baits with cineole, the most volatile compound, were recharged Taxonomy follows Nemésio & Rasmussen (2011). every hour, whereas the other scents, not highly volatile, were only charged when baits were exposed. Bees at- tracted to the baits during the sampling period were col- Results lected with insect nets, killed with ethyl acetate and pinned for posterior identification. All collected speci- Four hundred and thirty-one orchid-bee males be- mens were deposited at the entomological collection of longing to 11 species were collected from Novem- the ‘Laboratório de Taxonomia de Abelhas’ at the ‘Uni- ber, 2012 to January, 2013 at EEP (Table 1). Eu- versidade Federal de Uberlândia’ and numbered from HY00001 to HY000430 (specimens collected in November laema nigrita Lepeletier was by far the most com- and December, 2012) and from HY000696 to HY000743 mon species at both sampling sites, ranging from (specimens collected in January, 2013). 61% to 67% of the bees collected (Table 1). Euglossa imperialis Cockerell was the second most abundant Data analysis species, representing from 18% (at site-1) to 23% Diversity was estimated through Simpson’s diversity in- (at site-2) of the orchid-bee specimens. Abundance dex (H’), as 1 - D (Smith & Wilson, 1996, Magurran, 2004, Tuomisto, 2012). Evenness (E) was estimated through the was almost the same in both sites (216 specimens formula E = H’/ ln (S), where S is the species richness. at site-1 and 215 specimens at site-2, Table 1), as The results were directly compared to those obtained by: well as diversity (D = 0.43 at site-1 and D = 0.48 at (i) Alvarenga et al. (2007), who actively sampled orchid site-2) and evenness (E = 0.57 at site-1 and E = 0.49 bees with insect nets at EEP from February to May, 2004, at site-2). Richness, on the other hand, was higher on a monthly basis (from 09:00h to 12:00h), using four at site-2 (ten) than at site-1 (seven) (Table 2). Only aromatic scents (cineole, eugenol, methyl salicylate, and one species recorded at site-1 was not present in vanillin); (ii) Freitas (2009), who used bait traps to collect orchid bees from December, 2006 to December, 2007, on a site-2, whereas three species were exclusively re- monthly basis (from 09:00h to 12:00h), using seven aro- corded at site-2; all these species, however, were matic scents (benzyl acetate, -ionone, cineole, eugenol, represented by singletons (Table 1). Thus, the total methyl salicylate, methyl trans-cinnamate, and vanillin); richness observed in the present study reached 11 and (iii) Justino & Augusto (2012), who also used bait species. Overall, similarity among sampling sites traps from November, 2006 to April, 2007, on a monthly was relatively high, ranging from 69% to 88% ac- basis (from 09:00h to 12:00h), using four aromatic com- cording to Renkonen similarity coefficient. Site-2 pounds (cineole, eugenol, methyl salicylate, and vanillin). of the present study grouped with the site sam- Freitas’s (2009) site-1 is exactly the same as the site-1 of the present study, and Freitas’s (2009) site-2 is a patch of pled by Justino & Augusto (2012) with 88% simi- ‘Cerrado’ at the southernmost portion of EEP. Both Al- larity and, then, with site-1 of the present study varenga et al. (2007) and Justino & Augusto (2012) did not with 87% similarity. These three sites grouped provide the correct location of their sampling site. The with site-2 (‘Cerrado’) of Freitas (2009) with 70% geographic coordinates of the latitude provided by them, similarity and, then, with site-1 (gallery forest) of 19°09’20”S, in fact, lies completely outside EEP, in a pas- Freitas (2009) and the site sampled by Alvarenga ture area. It is important to stress because EEP is an island et al. (2007) with 69% similarity. These two latter of well preserved ‘Cerrado’ inserted in a matrix of pas- ture. Thus, considering a landscape perspective, the coor- sites shared 87% similarity with each other. Inter- dinates provided by the latter authors places their site in estingly, site-1 of the present study and site-1 of another environment. It is most probably an error, since Freitas (2009), which are the same, shared 75% from their text it is possible to infer that the sampling sites similarity.

were located in gallery forests inside EEP. It is also not

clear whether the overlapping data from Freitas (2009)

Rapid sampling of orchid bees in Central Brazil 233

Table 1. Number of specimens of each orchid-bee species collected under different sampling protocols at ‘Estação Ecológica do Panga’, Uberlândia, state of Minas Gerais. A and B: 20-hour sampling protocol with insect nets with 17 scents (this study), C and D: year-round sampling protocol with baited traps with seven scents (Freitas 2009), E: samplings during four months with insect nets with four scents (Alvarenga et al. 2007), F: samplings during six months with insect nets with four scents (Justino & Augusto 2012).

Alvarenga Justino and This study Freitas (2009) et al. (2009) Augusto (2012) A1 B C1 D E F surinamensis (Linnaeus, 1758) 0 1 0 0 0 0 Euglossa amazonica2 Dressler, 1982 1 0 0 0 1 0 E. carolina3 Nemésio, 2009 0 0 0 2 0 0 E. decorata Smith, 1874 0 3 0 0 0 0 E. despecta Moure, 1968 0 1 0 0 0 0 E. fimbriata Moure, 1968 2 10 0 0 0 0 E. imperialis Cockerell, 1922 50 38 85 2 13 8 E. jacquelynae Nemésio, 2007 14 9 16 0 1 0 E. melanotricha Moure, 1967 0 0 2 2 0 0 E. pleosticta Dressler, 1982 15 7 3 1 3 1 Eulaema marcii4 Nemésio, 2009 0 1 0 1 0 1 E. nigrita Lepeletier, 1841 132 144 79 82 11 26 2 1 2 0 0 0 Guérin-Méneville, 1844 Total 216 215 187 90 29 36

1Sites A (site-1 of this study) and C (site-1 of Freitas 2009) are the same site. 2This species was listed as Euglossa truncata Rebêlo & Moure by Alvarenga et al. (2007). 3This species was listed as Euglossa cordata (Linnaeus) by Freitas (2009). 4This species was listed as Eulaema cingulata (Fabricius) by Freitas (2009) and Justino & Augusto (2012).

Discussion An additional female specimen of El. helvola was collected at the campus of the ‘Universidade Fed- Faunistics, richness and diversity eral de Uberlândia’ in January, 2013, by the pre- The orchid-bee fauna of EEP was sampled with sent author. Thus, at least 15 orchid-bee species bait scents three times before the present study are known to occur in Uberlândia. (Alvarenga et al. 2007, Freitas 2009, Justino & Au- Although containing species with obvious af- gusto 2012), besides fortuitous data gathered from finities with the Amazon Forest, such as Euglossa bee samplings on flowers (e.g., Carvalho & Bego decorata Smith (see Nemésio et al. 2007) and Eg. 1996, Barbosa 1997), making its orchid-bee assem- amazonica (see Nemésio, 2012c), the orchid-bee blage well known. These previous samplings re- fauna of the region is not particularly rich, show- corded a total number of 11 species, and only two ing a low number of species when compared to of them, Euglossa carolina Nemésio and Eg. other biomes in Eastern Brazil, such as the “Hileia melanotricha Moure, were not recorded in the pre- Baiana”, the Atlantic Forest of southern Bahia, sent study. On the other hand, two species are with more than 40 recorded species (Nemésio here reported for EEP for the first time: Eg. despecta 2013d). On the other hand, if compared to other Moure and Eg. fimbriata Moure. These two records xeric areas, such as nearby ‘Cerrado’ areas (e.g., raise to 13 the number of orchid-bee species Nemésio & Faria Jr. 2004, Faria & Silveira 2011), known from EEP. It should be stressed that after high elevation ‘campos rupestres’ (Nemésio & studying specimens directly, I reached the conclu- Faria Jr. 2004, Viotti et al. 2013), or ‘Caatinga’ sion that the record of Eg. truncata Rebêlo & (Andrade-Silva et al. 2012), the orchid-bee fauna of Moure by Alvarenga et al. (2007) was a misidenti- EEP is fairly rich. Diversity, as measured by Simp- fication of Eg. amazonica Dressler (see Nemésio son’s diversity index, on the other hand, is rea- 2012c). Moreover, two other orchid-bee species are sonably low, even when compared to poorer fau- known from nearby areas in Uberlândia: Eufriesea nas of ‘Cerrado’ (e.g., Faria & Silveira 2011). This auriceps (Friese) and Eulaema helvola Moure were is at least partially explained by the strong domi- collected in a ‘Cerrado’ preserve ca. 20 km far nance of El. nigrita and Eg. imperialis which, to- from EEP (Alvarenga et al. 2007, Freitas 2009: 37). gether, represented more than 80% of orchid-bee

234 A. Nemésio.

Table 2. Orchid-bee inventories carried out at ‘Estação Ecológica do Panga’, Uberlândia, state of Minas Gerais, Eastern Brazil. A = Abundance (number of bees collected per hour), R = richness, D = diversity, measured as Simpson’s in- dex, E = evenness. Vegetation types are Ce = ‘Cerrado’, GF = gallery forest. EEP = ‘Estação Ecológica do Panga’.

Study Site Vegetation Protocol A R D E This study EEP-1 GF Net 10.8 7 0.43 0.57 This study EEP-2 GF Net 10.75 10 0.48 0.49 Alvarenga et al. (2007) EEP ? Net 3.77 6 0.39 0.60 Freitas (2009) EEP-1 GF Trap 4.79 6 0.83 0.61 Freitas (2009) EEP-2 Ce Trap 2.31 5 0.33 0.24 Justino and Augusto (2012) EEP ? Trap 2.00 4 0.56 0.55

specimens sampled in all studies carried out at cern because it introduces some level of uncer- EEP (Alvarenga et al. 2007, Freitas 2009), including tainty regarding how comparable results from dif- the present one (see Table 1). ferent studies can be. One of the first approached Diversity of bees on a global scale is higher in issues concerning alternative methodologies was temperate, warm and xeric regions (see Michener the differences in effectiveness between active col- 1979). Orchid bees, thus, are an exception to the lecting with insect nets and passive collecting with rule for presenting a higher diversity, abundance bait traps. Nemésio & Morato (2004, 2006), and richness in forested areas in the Neotropics Matozzo et al. (2011), Justino & Augusto (2012), (Dressler 1982, Roubik & Hanson 2004, Nemésio & Nemésio & Santos Júnior (2014), and Nemésio & Silveira 2007). Although the low diversity and Vasconcelos (2014) demonstrated that effective- abundance of orchid bees in xeric areas is a known ness of bait traps is quite questionable, since it in- phenomenon, only recently a reasonable number troduces a serious bias as a consequence of a of studies in these areas were performed (e.g. higher proportion of specimens belonging to Eu- Nemésio & Faria Jr. 2004, Faria & Silveira 2011, laema being trapped when compared to samplings Viotti et al. 2013), and only now their richness, di- with insect nets. It does not mean that bait traps versity and abundance can be quantified more are more effective than insect nets when it comes precisely. Storck-Tonon et al. (2009: 700-701) pre- to sampling Eulaema. Bait traps just allow that sented a summary of diversity of orchid-bee fau- other orchid bees escape more easily than Eulaema nas along the Neotropical region, which was species (Nemésio & Vasconcelos 2014). shown to range from very low in Rio Grande do Thus, comparisons involving the data ob- Sul, Southern Brazil, close to the southernmost tained by Freitas (2009) and Justino & Augusto geographic limits for orchid bees, to very high in (2012), through bait traps, and those obtained by southwestern Amazon and Peruvian Amazon Alvarenga et al. (2007) and the present study, (Nemésio & Rasmussen 2014). In general, in per- through active sampling with insect nets, should ennial tropical forests, orchid-bee diversity is con- be interpreted with great care. Moreover, other siderably higher than that observed in most xeric considerable differences should be taken into ac- areas (Storck-Tonon et al. 2009, Nemésio & Ras- count, as for example the season when orchid bees mussen 2014). In fact, only two species seem to be were sampled. Alvarenga et al. (2007) sampled endemic in more xeric areas: Eufriesea nigrohirta EEP at the end of the rainy season and the begin- (Friese), a species only recorded in rocky fields ning of the dry season, whereas most of Justino & over 1,000 m a.s.l. (see Nemésio 2005), and Eu- Augusto’s (2012) and the present study were car- laema helvola, a species only found in ‘Cerrado’ ar- ried out only during the rainy season, and eas of central and Eastern Brazil (see Nemésio & Freitas’s (2009) study was conducted during a Ferrari 2012). All the remaining species are typical whole year. Even taking these inconsistencies into of semideciduous and/or tropical rain forests, al- consideration, though, similarity (69% to 88%) though some of them, such as Eg. carolina and El. among all samplings was surprisingly high. A re- nigrita are also highly tolerant to open and even cent study in a large remnant of the Atlantic Forest urban areas. in eastern Minas Gerais showed only 38% similar- ity between two samplings at the same sites, un- Comparing sampling protocols der similar protocols (see Nemésio 2003, Nemésio The lack of standardization in orchid-bee sam- & Paula 2013). The high similarity observed in the plings noted by Morato (1998) is a matter of con- present study is mainly explained because the two Rapid sampling of orchid bees in Central Brazil 235 most common species at the area, Eulaema nigrita during the season when orchid bees are most ac- and Euglossa imperialis, were the most common in tive. Furthermore, it reduces the costs of field all samplings, except the site located at the ‘Cer- work by reducing the needed number of field rado’ (site-2 of Freitas, 2009), where Eulaema nigrita trips. Of course, this protocol is not adequate to alone represented more than 90% of the collected study phenology of orchid-bee populations, as bees. In fact, the orchid-bee fauna of EEP is basi- previously discussed (Nemésio 2010b, 2011a). cally characterized by the prevalence of El. nigrita and Eg. imperialis, moderate abundance of Eg. jac- Towards a standardized sampling protocol quelynae Nemésio and Eg. pleosticta Dressler, and A standard sampling protocol would be desirable some rare species, some of them represented by for many reasons, but there is one reason above all singletons in all samplings. This predominance of that would be enough to justify Morato’s (1998) El. nigrita in Eastern Brazil has already been ob- plea for a standard protocol: the possibility of di- served by previous authors (e.g. Knoll et al. 2012, rect comparisons in ecological and biogeographic Silva & De Marco Jr., 2013, Knoll, 2016). This is a studies. Orchid-bees are exclusive Neotropical in- potential explanation for the absence of Eg. caro- sects and raising funds for field trips to sample in- lina and Eg. melanotricha in the present study, spe- sects is not an easy task in most, if not all, coun- cies rarely recorded at EEP. tries in the Neotropics. If only the traditional year- This study is the one which recorded the high- round methodology is to be encouraged, it be- est number of specimens in a single site and, also, comes difficult, from a financial point of view, to the only study where the highly seasonal Ef. suri- support studies far from the cities where the few namensis and the rarely attracted Eg. decorata were universities where orchid-bee researchers are situ- recorded regularly (see Table 1). Previous records ated. Of course, year-round and even long term of these species in the area (Carvalho & Bego 1996, studies should always be encouraged (see Barbosa 1997) were based on sporadic samplings Nemésio et al. 2015). Rapid inventories cannot re- on flowers. The results presented here, thus, veal important aspects of species’ natural history strongly suggest that the 20-hour sampling proto- and other biological features, such as seasonal col is a valuable tool for obtaining reliable data on population fluctuations, assessment of possible orchid-bee faunas, even in xeric environments. declines in species’ populations and consequences The high abundance recorded in the present study of deforestation and other human-made impacts is outstanding, only rivaled by a single site at a in natural environments on the assemblages of ‘Cerrado’ area at ‘Parque Estadual do Rio Preto’ these bees. A rapid sampling protocol, however, (see Nemésio & Faria Jr. 2004). It has been sug- may be also of great utility, as previously shown gested that undersampling, i.e., not recording all in the literature, being the most evident of all the taxa present at a site, is a common situation in species composition of a given site (e.g. Braga fauna inventories (Coddington et al. 2009, Beck & 1976, Nemésio 2003, Nemésio & Faria Jr. 2004, Schwanghart 2010, Beck et al. 2013). Some orchid- Abrahamczyk et al. 2011, Storck-Tonon et al. bee species are simply not attracted to the syn- 2011), and a rapid inventory which results can also thetic scents and even those which are may be dif- be reasonably comparable to the traditional inven- ferently attracted to them (reviewed by Nemésio tories should be pursued. In order to improve the 2012a). For example, it is highly likely that El. hel- potential of these rapid inventories, the following vola occurs at EEP, but since their males are usu- suggestions are presented, based on present re- ally not attracted to the commonly used synthetic sults and on those available from previous studies. scents, this species was not recorded in the present study, even though, as pointed out above, it was Concerning baits recorded inside the urban area of Uberlândia. Traditional orchid-bee inventories normally use The 20-hour sampling protocol, originally de- three to six scents, although usually no specific signed to be used in forested areas, revealed to be explanation is given by the authors of each study a promising methodology from sampling in xeric for their particular choices (discussed by Nemésio areas, where the abundance of orchid bees is very 2012a). A larger number of scents may potentially low during most of the year (e.g., Freitas 2009, attract more species, since those species attracted Silveira 2010, Faria & Silveira 2011). The protocol to only uncommon scents have a higher chance of used here maximizes the possibility of sampling being sampled. For example, in the present study the rarest species by raising the sampling effort the few specimens of Exaerete smaragdina (Guérin- 236 A. Nemésio.

Méneville) were attracted to p-tolyl acetate, one of Ackerman, J.D. (1989): Geographic and seasonal variation in  fragrance choices and preferences of male euglossine bees. the Eg. decorata specimens was collected at a - Biotropica 21: 340–347. ionone bait, r-carvone was attractive to Eg. jacque- Alvarenga, P.E.F., Freitas, R..F, Augusto, S.C. (2007): Diversidade lynae, and skatole was attractive to many species, de Euglossini (Hymenoptera: Apidae) em áreas de Cerrado do Triângulo Mineiro, MG. Bioscience Journal 23: 30–37. including Eg. amazonica, Eg. fimbriata, and Eg. jac- Andrade-Silva, A.C.R., Nemésio, A., Oliveira, F.F., Nascimento, F.S. quelynae. The use of these scents may account for (2012): Spatial-temporal variation in orchid bee communities the higher abundance and richness observed in (Hymenoptera: Apidae) in remnants of arboreal Caatinga in the Chapada Diamantina region, state of Bahia, Brazil. Neotropical this study when compared to previous studies at Entomology 41: 296–305. EEP (Alvarenga et al. 2007, Freitas 2009, Justino & Armbruster, W.C. (1993): Within-habitat heterogeneity in baiting Augusto 2012). Based on the present results and samples of male euglossine bees: possible causes and implications. Biotropica 25: 122–128. previous studies using the same scents employed Balmer, O. (2002): Species lists in ecology and conservation: here both in the Atlantic and in the Amazon for- abundances matter. Conservation Biology 16: 1160–1161. ests, I strongly suggest that future samplings use Barbosa, A.A.A. (1997): Biologia reprodutiva de uma comunidade de campo sujo, Uberlândia/MG. Ph.D. Dissertation. at least ten scents, instead of the traditional five or Universidade Estadual de Campinas, Campinas, 189p. six. These are: benzyl acetate, r-carvone, 1,8- Beck, J., Holloway, J.D., Schwanghart, W. (2013): Undersampling cineole, eugenol, -ionone, trans methyl- and the measurement of beta diversity. Methods in Ecology & Evolution 4: 370–382. cinnamate, methyl salicylate, skatole, p-tolyl ace- Beck, J., Schwanghart, W. (2010): Comparing measures of species tate, and vanillin. Some of them, as eugenol, trans diversity from incomplete inventories: an update. Methods in methyl-cinnamate and p-tolyl acetate, usually do Ecology & Evolution 1: 38–44. Braga, P.I.S. (1976): Atração de abelhas polinizadoras de not attract many bees, but are strongly attractive Orchidaceae com auxílio de iscas-odores na campina, to particular species that fail to attend to other campinarana e floresta tropical úmida da região de Manaus. baits or are weakly attracted to them (see Nemésio Ciência e Cultura 28: 767–773. Bruna, E.M., Guimarães, J.F., Lopes, C.T., Duarte, P., Gomes, et al. 2012, Viotti et al. 2013, Nemésio & Rasmus- A.C.L., Belentani, S.C.S., Pacheco, R., Facure, K.G., Lemos, F.G., sen 2014). Vasconcelos, H.L. (2010): Mammalia, Estação Ecológica do Panga, a Cerrado protected area in Minas Gerais state, Brazil. Check List 6: 668–675. Concerning season to be sampled Cardoso, E., Moreno, M.I.C.B., Vasconcelos, H.L. (2009): Mudanças Even though rapid samplings recorded a high fitofisionômicas no Cerrado: 18 anos de sucessão ecológica na abundance in some studies conducted in less fa- Estação Ecológica do Panga, Uberlândia - MG. Caminhos de Geografia 10: 254–268. vourable seasons (e.g. Nemésio 2003, Nemésio & Carvalho, A.M.C., Bego, L.R. (1996): Studies on Apoidea fauna of Faria Jr. 2004, Abrahamczyk et al. 2011, Nemésio Cerrado vegetation at the Panga Ecological Reserve, Uberlândia, & Santos Júnior 2014), I strongly recommend that MG, Brazil. Revista Brasileira de Entomologia 40: 147–156. Coddington, J.A., Agnarsson, I., Miller, J.A., Kuntner, M., Hormiga, rapid inventories be carried out during the season G. (2009): Undersampling bias: the null hypothesis for singleton when orchid bees are most actively foraging, i.e., species in tropical surveys. Journal of the rainy season. In these circumstances, not only Ecology 78: 573–584. Costa, A.A., Araújo, G.M. (2001): Comparação da vegetação arbórea the overall community composition can be sam- de cerradão e cerrado na Reserva do Panga, Uberlândia, Minas pled with higher fidelity, but also the rarest spe- Gerais. Acta Botanica Brasilica 15: 63–72. cies, including specimens of Eufriesea spp., Dressler, R.L. (1982): Biology of the orchid bees (Euglossini). Annual Review of Ecology and Systematics 13: 373–394. univoltine species highly seasonal and usually Embrapa (1982): Levantamento de reconhecimento de média only active during one or two months during the intensidade dos solos e avaliação da aptidão agrícola das terras rainy season (see Kimsey 1982), are most probably do Triângulo Mineiro. Rio de Janeiro: Serviço Nacional de Levantamento e Conservação de Solos. 526p. active and richness can, thus, be more confidently Faria, L.R.R., Silveira, F.A. (2011): The orchid bee fauna recorded. (Hymenoptera, Apidae) of a core area of the Cerrado, Brazil: the role of riparian forests as corridors for Forest-associated bees. Biota Neotropica 11: 87–94. Freitas, R.F. (2009): Diversidade e Sazonalidade de Abelhas Euglossini Latreille (Hymenoptera: Apidae) em Fitofisionomias References do Bioma Cerrado em Uberlândia, MG. M.Sc. Thesis. Universidade Federal de Uberlândia, Uberlândia, xi+65p. Abrahamczyk, S., Gottleuber, P., Matauschek, C., Kessler, M. Janzen, D.H. (1971): Euglossine bees as long distance pollinators of (2011): Diversity and community composition of euglossine bee tropical plants. Science 171: 203–205. assemblages (Hymenoptera: Apidae) in western Amazonia. Justino, D.G., Augusto, S.C. (2012) [2010]: Avaliação da eficiência Biodiversity Conservation 20: 2981–3001. de coleta utilizando armadilhas aromáticas e riqueza de Ackerman, J.D. (1983): Diversity and seasonality of male euglossine Euglossini (Hymenoptera, Apidae) em áreas de Cerrado do bees (Hymenoptera: Apidae) in Central Panama. Ecology 64: Triângulo Mineiro. Revista Brasileira de Zoociências 12: 227– 274–283. 239. Rapid sampling of orchid bees in Central Brazil 237

Kimsey, L.S. (1982): Systematics of bees of the genus Eufriesea Gerais, eastern Brazil (Hymenoptera, Apidae, Euglossina). (Hymenoptera, Apidae). University of California Publications in Spixiana 35: 109–116. Entomology 95: 1–125. Nemésio, A. (2013a): The orchid-bee fauna (Hymenoptera: Apidae) Knoll, F.R.N. (2016): Variation in the Abundance of Neotropical of ‘Reserva Biológica de Una’, a hotspot in the Atlantic Forest of Bees in an Unpredictable Seasonal Environment. Neotropical southern Bahia, eastern Brazil. Brazilian Journal of Biology 73: Entomology 45: 129–138. 347–352. Knoll, F.R.N., Penatti, N.C. (2012): Habitat fragmentation effects on Nemésio, A. (2013b): Are orchid bees at risk? First comparative the orchid bee communities in remnant forests of southeastern survey suggests declining populations of forest-dependent Brazil. Neotropical Entomology 41: 355–65. species. Brazilian Journal of Biology 73: 367–374. Lopes, C.T., Vasconcelos, H.L. (2008): Evaluation of three methods Nemésio, A. (2013c): The orchid-bee faunas (Hymenoptera: for sampling ground-dwelling ants in the Brazilian Cerrado. Apidae) of two Atlantic Forest remnants in southern Bahia, Neotropical Entomology 37: 399–405. eastern Brazil. Brazilian Journal of Biology 73: 375–381. Magurran, A.E. (2004): Measuring Biological Diversity. Blackwell Nemésio, A. (2013d): The orchid-bee faunas (Hymenoptera: Publishing, Oxford. Apidae) of ‘Parque Nacional do Monte Pascoal’, ‘Parque Mattozo, V.C., Faria, L.R.R., Melo, G.A.R. (2011): Orchid bees Nacional do Descobrimento’ and three other forest remnants of (Hymenoptera: Apidae) in the coastal forests of southern Brazil: Atlantic Forest in southern Bahia, eastern Brazil. Brazilian diversity, efficiency of sampling methods and comparison with Journal of Biology 73: 437–446. other Atlantic Forest surveys. Papéis Avulsos de Zoologia 51: Nemésio, A., Augusto, S.C., Almeida, E.A.B. (2007): Euglossa 505–515. decorata Smith (Hymenoptera: Apidae) in central Brazil – Michener, C.D. (1979): Biogeography of the bees. Annals of the biogeographic implications. Lundiana 8: 57–61. Missouri Botanical Garden 66: 277–342. Nemésio, A., Cerântola, N.C.M., Vasconcelos, H.L., Nabout, J.C., Milet-Pinheiro, P., Schlindwein, C. (2005): Do euglossine males Silveira, F.A., Del Lama, M.A. (2012): Searching for Euglossa (Apidae, Euglossini) leave tropical rainforest to collect cyanochlora Moure, 1996 (Hymenoptera: Apidae), one of the fragrances in sugarcane monocultures? Revista Brasileira de rarest bees in the world. Journal of Insect Conservation 16: 745– Zoologia 22: 853–858. 755. Morato, E.F. (1998): Estudos sobre comunidades de abelhas Nemésio, A., Faria Jr., L.R.R. (2004): First assessment of orchid bee Euglossini. Anais do Encontro sobre Abelhas 3: 135–143. fauna (Hymenoptera: Apidae: Apini: Euglossina) of Parque Morato, E.F., Campos, L.A.O., Moure, J.S. (1992): Abelhas Estadual do Rio Preto, a cerrado area in southeastern Brazil. Euglossini (Hymenoptera, Apidae) coletadas na Amazônia Lundiana 5: 113–117. Central. Revista Brasileira de Entomologia 36: 767–771. Nemésio, A., Ferrari, R.R. (2012): The species of Eulaema (Eulaema) Moure, J.S. (1996) [1995]: Notas sobre algumas espécies de abelhas Lepeletier, 1841 (Hymenoptera: Apidae: Euglossina) from da Bahia, Brasil (Hymenoptera, Apoidea). Revista Brasileira de eastern Brazil, with description of Eulaema quadragintanovem sp. Zoologia 12: 467–470. n. from the state of Ceará. Zootaxa 3478: 123–132. Nemésio, A. (2003): Preliminary sampling of Euglossina Nemésio, A., Morato, E.F. (2004): Euglossina (Hymenoptera: (Hymenoptera: Apidae: Apini) of Reserva Particular do Apidae: Apini) of the Humaitá Reserve, Acre state, Brazilian Patrimônio Natural Feliciano Miguel Abdala, Caratinga, Minas Amazon, with comments on bait trap efficiency. Revista de Gerais state, Brazil. Lundiana 4: 121–124. Tecnologia e Ambiente 10: 71–80. Nemésio, A. (2005): Description of the male Eufriesea nigrohirta Nemésio, A., Morato, E.F. (2006): The orchid-bee fauna (Friese, 1899) (Hymenoptera: Apidae: Apini: Euglossina) with (Hymenoptera: Apidae) of Acre state (northwestern Brazil) and comments on the holotype, species biology and distribution. a re-evaluation of euglossine bait-trapping. Lundiana. 7: 59–64. Lundiana 6: 41–45. Nemésio, A., Paula, I.R.C. (2013): The orchid-bee fauna Nemésio, A. (2009): Orchid bees (Hymenoptera: Apidae) of the (Hymenoptera: Apidae) of 'RPPN Feliciano Miguel Abdala' Brazilian Atlantic Forest. Zootaxa 2041: 1–242. revisited: relevant changes in community composition. Brazilian Nemésio, A. (2010a): Eulaema (Apeulaema) felipei sp. n. Journal of Biology 73: 515–520. (Hymenoptera: Apidae: Euglossina): a new forest-dependent Nemésio, A., Rasmussen, C. (2011): Taxonomic issues in the orchid orchid bee found at the brink of extinction in northeastern bees (Hymenoptera: Apidae: Euglossina), and an updated Brazil. Zootaxa 2424: 51–62. catalogue. Zootaxa 3006: 1–42. Nemésio, A. (2010b): The orchid-bee fauna (Hymenoptera: Apidae) Nemésio, A., Rasmussen, C. (2014): Sampling a biodiversity of a forest remnant in northeastern Brazil, with new geographic hotspot: the orchid-bee fauna (Hymenoptera: Apidae) of records and an identification key to the known species of the Tarapoto, northeastern Peru, the richest and most diverse site of Atlantic Forest of northeastern Brazil. Zootaxa 2656: 55–66. the Neotropics. Brazilian Journal of Biology 74: S033-S044. Nemésio, A. (2011a): The orchid-bee fauna (Hymenoptera: Apidae) Nemésio, A., Santos, L.M., Vasconcelos, H.L. (2015): Long-term of a forest remnant in southern Bahia, Brazil, with new ecology of orchid bees in an urban forest remnant. Apidologie geographic records and an identification key to the known 46: 359–368. species of the area. Zootaxa 2821: 47–54. Nemésio, A., Santos Júnior, J.E. (2014): Is the “Centro de Nemésio, A. (2011b): Euglossa marianae sp. n. (Hymenoptera: Endemismo Pernambuco” a biodiversity hotspot for orchid Apidae): a new orchid bee from the Brazilian Atlantic Forest and bees? Brazilian Journal of Biology 74: S078-S092. the possible first documented local extinction of a forest Nemésio, A., Silva, D.P., Nabout, J.C., Varela, S. (2016): Effects of dependent orchid bee. Zootaxa 2892: 59–68. climate change and habitat loss on a forest-dependent bee Nemésio, A. (2011c): Rediscovered after forty-two years: the rare species in a tropical fragmented landscape. Insect Conservation orchid bee Eufriesea brasilianorum (Hymenoptera: Apidae) of and Diversity 9(2): 149-160. eastern Brazil. North-Western Journal of Zoology 7: 356–359. Nemésio, A., Silveira, F.A. (2004): Biogeographic notes on rare Nemésio, A. (2012a): Methodological concerns and challenges in species of Euglossina (Hymenoptera: Apidae: Apini) occurring in ecological studies with orchid bees (Hymenoptera: Apidae: the Brazilian Atlantic Rain Forest. Neotropical Entomology 33: Euglossina). Bioscience Journal 28: 118–134. 117–120. Nemésio, A. (2012b): Species of Euglossa Latreille, 1802 Nemésio, A., Silveira, F.A. (2006a): Deriving ecological (Hymenoptera: Apidae: Euglossina) belonging to the purpurea relationships from geographical correlations between host and species group occurring in eastern Brazil, with description of parasitic species: an example with orchid bees. Journal of Euglossa monnei sp. n. Zootaxa 3151: 35-52. Biogeography 33: 91–97. Nemésio, A. (2012c): The western limits of the “Hileia Baiana” for Nemésio, A., Silveira, F.A. (2006b): Edge effects on the orchid-bee orchid bees, including seven new records for the state of Minas fauna (Hymenoptera: Apidae) at a large remnant of Atlantic 238 A. Nemésio.

Forest in southeastern Brazil. Neotropical Entomology 35:313– Roubik, D.W., Ackerman, J.D. (1987): Long-term ecology of 323. euglossine orchid-bees (Apidae: Euglossini) in Panama. Nemésio, A., Silveira, F.A. (2006c): First record of Eulaema helvola Oecologia 73: 321–333. Moure (Hymenoptera: Apidae: Euglossina) for the state of Roubik, D.W., Hanson, P.E. (2004): Orchid bees: biology and field Minas Gerais: biogeographic and taxonomic implications. guide. San Jose. INBIO. 370p. Neotropical Entomology 35: 418–420. Silva, D.P., De Marco, Jr, P. (2014): No evidence of habitat loss Nemésio, A., Silveira, F.A. (2007): Diversity and distribution of affecting the orchid bees Eulaema nigrita Lepeletier and Eufriesea orchid bees (Hymenoptera: Apidae) with a revised checklist of auriceps Friese (Apidae: Euglossini) in the Brazilian Cerrado species. Neotropical Entomology 36: 874–888. Savanna. Neotropical Entomology 43: 509–518. Nemésio, A., Silveira, F.A. (2010): Forest fragments with larger core Silveira, G.C. (2010): Diversidade e sazonalidade de abelhas areas better sustain diverse orchid bee faunas (Hymenoptera: Euglossini Latreille (Hymenoptera: Apidae) em duas áreas de Apidae). Neotropical Entomology 39: 555–561. mata estacional semidecidual no domínio do Cerrado no Nemésio, A., Vasconcelos, H.L. (2013): Beta diversity of orchid bees Triângulo Mineiro. M.Sc. Thesis. Universidade Federal de in a tropical biodiversity hotspot. Biodiversity and Conservation Uberlândia, Uberlândia, xi+57p. 22: 1647–1661. Smith, B., Wilson, J.B. (1996): A consumer’s guide to evenness Nemésio, A., Vasconcelos, H.L. (2014): Effectiveness of two indices. Oikos 76: 70–82. sampling protocols to survey orchid bees (Hymenoptera: Storck-Tonon, D., Morato, E.F., Oliveira, M.L. (2009): Fauna de Apidae) in the Neotropics. Journal of Insect Conservation 18: Euglossina (Hymenoptera: Apidae) da Amazônia Sul-Ocidental, 197–202. Acre, Brasil. Acta Amazonica 39: 693–706. Oliveira, M.L., Campos, L.A.O. (1996): Preferências por estratos Storck-Tonon, D., Silva, M.V., Morato, E.F. (2011): Checklist of florestais e por substâncias odoríferas em abelhas Euglossinae orchid bees (Hymenoptera: Apidae) of “Lago do Silêncio” area, (Hymenoptera, Apidae). Revista Brasileira de Zoologia 13: 1075– Boca do Acre, Amazonas, Brazil. Check List, 7: 648–651. 1085. Sydney, N.V., Gonçalves, R.B., Faria, L.R.R. (2010): Padrões Pearson, D.L., Dressler, R.L. (1985): Two-year study of male orchid espaciais na distribuição de abelhas Euglossina (Hymenoptera, bee (Hymenoptera: Apidae: Euglossini) attraction to chemical Apidae) da região Neotropical. Papéis Avulsos de Zoologia 50: baits in lowland south-eastern Perú. Journal of Tropical Ecology 667–679. 1: 37–54. Tonhasca Jr., A., Blackmer, J.L., Albuquerque, G.S. (2002): Perger, R. (2015): The highest known egulossine bee community Abundance and diversity of euglossine bees in the fragmented from a garden in the Bolivian Andes (Hymeoptera, Apidae, landscape of the Brazilian Atlantic Forest. Biotropica 34: 416– Euglossini). Journal of Hymenoptera Research 45: 65–73. 422. Pielou, E.C. (1975): Ecological Diversity. New York, John Wiley & Tuomisto, H. (2012): An updated consumer’s guide to evenness and Sons. 165p. related indices. Oikos, 121: 1203–1218. Pokorny, T., Loose, D., Dyker, G., Quezada-Euan, J.J.G., Eltz, T. Viotti, M.A., Moura, F.R., Lourenço, A.P. (2013): Species diversity (2014): Dispersal ability of male orchid bees and direct evidence and temporal variation of the orchid-bee fauna (Hymenoptera, for long-range flights. Apidologie 46: 224–237. Apidae) in a conservation gradient of a rocky Field area in the Powell, A.H., Powell, G.V.N. (1987): Population dynamics of male Espinhaço Range, state of Minas Gerais, southeastern Brazil. euglossine bees in Amazonian forest fragments. Biotropica 19: Neotropical Entomology 42: 565–575. 176–179. Vogel, S. (1966): Parfümsammelnde Bienen als Bestäuber von Ribeiro, J.F., Walter, B.M.T. (1998): Fitofisionomias do Bioma Orchidaceen und Gloxinia. Österreiches Bot. Zeitschrift 113: 302– Cerrado. In: Sano, S.M., Almeida, S.P. (Eds.), Cerrado: Ambiente 361. e Flora. Embrapa, Planaltina, DF, 89–168. Wolda, H. (1981): Similarity indices, sample sizes and diversity. Roubik, D.W. (2004): Long-term studies of solitary bees: what the Oecologia 50: 296–302. orchid bees are telling us. In: Freitas, BM, and Pereira, JO, (Eds) Solitary bees? Conservation, rearing, management for pollination. Imprensa Universitária, Fortaleza, pp. 97–103.