The orchid-bee fauna ( Hymenoptera : Apidae ) of ‘ RPPN Feliciano Miguel Abdala ’ revisited : relevant changes in community composition

The orchid-bee fauna of ‘Reserva Particular do Patrimônio Natural Feliciano Miguel Abdala’, a 957-ha preserve of Atlantic Forest in eastern Minas Gerais, southeastern Brazil, was surveyed 12 years after the first inventory in the area. Orchid-bee males were actively collected with insect nets when attracted to seventeen chemical compounds used as scent baits. Three hundred and nineteen males belonging to nine species were collected during 40 hours in late December, 2011, when orchid bees are supposedly more active. Euglossa despecta Moure, 1968, one of the dominant species in the area 12 years ago, was not recorded in the present study. Eulaema nigrita Lepeletier, 1841, on the other hand, represented only 16% of the collected bees in 1999 and 61% in the present study. Possible causes and consequences of these changes are discussed.


Introduction
Orchid bees (Hymenoptera: Apidae: Euglossina) are important Neotropical pollinators (see Dressler, 1982a) and have gained in popularity among field researchers since the late 1960s, when it was realized that their males are easily attracted to aromatic compounds artificially synthesized that mimic floral fragrances (Vogel, 1966;Dodson et al., 1969).
As a consequence, very few studies have been carried out in areas where samplings had already been performed.Since many orchid bees are forest-dependent species (Dressler, 1982a;Morato et al., 1992;Morato, 1994) and at least one biome inhabited by these bees, the Atlantic Forest of eastern Brazil, has been heavily deforested and is still under strong anthropogenic pressure, it is reasonable to suppose that populations of the most sensitive species in these areas may be somehow affected.There is almost no study to show whether and how populations of these bees have been changing in the last years.The only study that addressed this subject (Nemésio, 2013b) suggested that the population of Euglossa marianae Nemésio, 2011b, a highly sensitive species, at a large remnant of Atlantic Forest in Linhares, state of Espírito Santo, has declined in almost 50% in a 12-year interval.
The main goal of the present study was to survey, twelve years after the first preliminary inventory, the orchid-bee fauna of a medium-sized Atlantic Forest remnant in southeastern Brazil (Nemésio, 2003) to assess possible changes in its community composition.

Study sites
This study was conducted at 'Reserva Particular do Patrimônio Natural Feliciano Miguel Abdala' (RFMA), a 957-ha preserve situated in the municipality of Caratinga, in the state of Minas Gerais, southeastern Brazil , in late December, 2011, when orchid bees are most actively foraging in the region (e.g., Nemésio andSilveira, 2007, 2010).Vegetation at RFMA is Tropical Semideciduous Forest, included in the Tropical Atlantic domain (Ab'Saber, 1977), and presents different successional stages (Hatton et al., 1983).Precipitation is highly seasonal, with a rainy season, from October to March, and a well-defined dry season from April to September (Strier, 1986).Elevations range from 400 to 680 m above sea level (Silva, 1993).

Sampling
Twenty hours of active sampling with insect nets were performed exactly in the same sites sampled 12 years before (see Nemésio, 2003), totaling 40 hours, following the methodology proposed by Nemésio (2010aNemésio ( , b, 2011a, b), b).The exact location of each site is: site-1, or the interior of forest in Nemésio (2003) (19°43'54" S -41°49'26" W, ca.500 m above sea level) was sampled on the 21 st (from 09:00 h to 14:00 h), 22 nd (from 08:00 h to 15:00 h) and 24 th of December, 2011 (from 08:00 h to 16:00 h); site-2, or the site close to the edge in Nemésio (2003) (19°44'00" S -41°49'16" W, ca.500 m above sea level) was sampled on the 23 rd and 25 th of December, 2011 (in both dates from 07:00 h to 17:00 h).At each site, 17 scent baits were placed ca.2.0 meters apart from each other at about 1.5 m above the ground.These baits were made of cotton wadding soaked with one of the following substances, known or believed to be attractive to orchid bees: benzyl acetate, benzyl alcohol, r-carvone, 1,8-cineole, p-cresol acetate, dimethoxybenzene, eugenol, b-ionone, methyl benzoate, methyl trans-cinnamate, heneicosane, methyl salicylate, skatole, tricosane, p-tolyl acetate, vanillin, and a mixture (1:1) of methyl transcinnamate and p-tolyl acetate.Baits with cineole, the most volatile compound, were recharged every hour.Bees arriving on the baits during the sampling period were collected with insect nets and killed with ethyl acetate.They were labeled as belonging to the project "Euglossina de Caratinga" and were deposited at the Entomological Collection of Universidade Federal de Uberlândia (UFU), where they were numbered from 22141-65214 to 22286-65603.

Data analysis
Diversity was estimated through Shannon-Wiener diversity index (H'), as H' = -Sp i ln (p i ), where p i is the proportion of total number of species made up of the ith species (Pielou, 1975).Evenness (E) was estimated through the formulaE = H'/ ln (S), where S is the species richness.Diversity and evenness were estimated both to the samplings carried out in the present study and to the samplings carried out in 1999 (Nemésio, 2003) for comparison.Similarity in faunistic composition between collections carried out in 1999 (Nemésio, 2003) and in the present study was estimated by the percent similarity index of Renkonen, recommended by Wolda (1981) and Balmer (2002) for small samples.

Results
Three hundred and nineteen orchid-bee males belonging to eight species in three genera were collected in the present study, 172 (in 6 species) at site-1 (interior of forest) and 147 (in 6 species) at site-2 (close to the edge) (Table 1).Additionally, a single specimen of Euglossa carolina Nemésio, 2009 was collected at a third site not included in the analysis, because it was sampled during only five hours.One specimen of a species belonging to the Euglossa analis Westwood, 1840 species group visited the skatole bait at site-2, but was not collected.Although it is impossible to know the exact species it belongs to, it is not Eg.marianae because the last three terga were reddish, not green.Eulaema nigrita Lepeletier, 1841 (61% of the total specimens, and the dominant species at both sites) was the most common species at RFMA, followed by Eulaema marcii Nemésio, 2009 (17%) and Eg.fimbriata Moure, 1968 (14%).These three species together represented 92% of the orchid-bee community during the sampling period.The only species represented by a singleton was Eg. truncata Rebêlo and Moure, 1996. Abundance (ca. 8.6 specimens/hour), diversity (H' = 1.26) and evenness (E = 0.65) were higher at site-1 than at site-2 (ca.7.3 specimens/hour; H' = 1.04;E = 0.58).Renkonen similarity index between samplings carried out in 1999 and in the present study was only 38%.

Discussion
The efficiency of the sampling methodology used here was previously discussed in depth (Nemésio 2011a;Nemésio et al., 2012).The relatively low abundance found in the present study (seven to nine specimens/hour) compared to other studies using the same methodology (Nemésio, 2010a(Nemésio, , b, 2011a(Nemésio, , b, 2013a-d)-d), may be partially understood as natural of the area, since the previous studies were carried out at Centro de Endemismo Pernambuco (Nemésio 2010a, b) and 'Hileia Baiana' (Nemésio 2011a, b), two regions well known for their high abundance and diversity (Bonilla-Gómez, 1999; Bezerra and Martins, 2001).On the other hand, it must be noticed that abundance at RFMA in 1999 was higher (ca.12 specimens/hour, or 50% higher) than in the present study.It also should be stressed that collections in 1999 were carried out during only five hours in each site (vs.20 hours in each site in the present study); as abundance is here given as the number of specimens collected per hour, this difference in sampling effort is, thus, reduced.
Seven species were collected in 1999 and only Euglossa despecta Moure, 1968 was collected then (Nemésio, 2003) and not collected in the present study.However, this species was one of the dominant species in 1999 and its absence in the present study is remarkable, especially because the bait that most strongly attracted it (b-ionone) was also used in the present study.On the other hand, four species collected in the present study were not recorded in 1999: the highly seasonal Eufriesea surinamensis (Linnaeus, 1758) and Ef.violacea (Blanchard, 1840), and Euglossa carolina (collected in a third site not included in the present analysis), besides one specimen belonging to the Euglossa analis species group, not collected, but attracted to skatole at site-2, and presenting the last three terga reddish.In the nearby 'Parque Estadual do Rio Doce' (see Figure 1), two spe-cies belonging to the Eg.analis species group were collected, Euglossa cognata Moure, 1970 and Eg.marianae (see Nemésio and Silveira, 2006 -Eg.marianae listed as Eg.analis), but Eg. marianae should be discarded since this species presents the last three terga green-colored.A third species, Eg. mixta Friese, 1899 was recorded deep in the interior of Minas Gerais (Nemésio, 2012), almost at the same longitude as RFMA, but at a lower latitude.It is impossible, thus, to know to which species the observed specimen belongs to.The number of species known to occur at RFMA, consequently, rises to eleven.
Although being a relatively large forest remnant of the Atlantic Forest in the state of Minas Gerais, and although considered to be of very high to extreme importance for conservation for many taxonomic groups (Fonseca and Rylands, 1998;Marini and Lamas, 1998;Silva and Mendonça, 1998;Haddad, 2000), including invertebrates (Silveira, 1998), RFMA apparently does not hold a rich and diverse orchid-bee fauna.In fact, the dominant species, especially Eulaema nigrita and El.marcii, are among the most common orchid-bee species in the Neotropics and considered to be tolerant to disturbed areas (Morato, 1998;Tonhasca Jr. et al., 2002;Nemésio and Silveira, 2006).Even smaller forest remnants, immersed in an urban matrix, and located in drier portions of the Atlantic Forest (Nemésio andSilveira, 2007, 2010) seem to present a more diverse orchid-bee fauna than RFMA.
The fluctuation in the orchid-bee community composition between both samplings (Nemésio, 2003 and the present study) is also remarkable.Although diversity and abundance did not change greatly, and almost all species collected in 1999 were also collected in the present study, their relative abundances changed dramatically and the only missing species from 1999 to 2011 (Eg.despecta) was an important part of the orchid-bee community twelve years before.The absence of Eg. despecta, and the Orchid bees of Caratinga, Minas Gerais, Brazil enormous rise of the relative contributions of Eg. fimbriata and El.nigrita to the orchid-bee community of RFMA in the present study made the similarity of both "faunas" to be only 38%, a very low value.Seasonality may respond for at least part of the observed change: the first sampling was carried out in August, 1999, in the middle of the dry season, while the present sampling was conducted in late December, exactly when precipitation is the highest at RFMA.Some species, as El.nigrita, present their lowest abundance during the middle to the end of the dry season in the region (Nemésio, unpub. data), and it may account for the observed differences.
On the other hand, the absence of Eg. despecta is not easily explained.It is known that populations of many species of orchid bees largely fluctuates (Roubik, 2001).However, this species occurred in relatively high abundances during the rainy season in recent collections in the 'Hileia Baiana' (Nemésio, 2011a, b).Moreover, its population did not "fluctuate", but entirely "disappeared" and there is no apparent explanation for its absence from RFMA.Euglossa despecta was also an important part of the orchid-bee fauna of 'Parque Estadual do Rio Doce', distant only 60 km from RFMA, in 1999 (Nemésio and Silveira, 2006).Nemésio and Silveira (2007) argued that deforestation and consequent fragmentation in many parts of the Atlantic Forest are relatively recent events, and populations of some species could be declining towards extinction in some of these forest remnants.The observed decline of populations of Eg. marianae at a large remnant in the state of Espírito Santo (Nemésio, 2013b) in a 12-year interval is suggestive of this hypothesis.Furthermore, except for this study and Nemésio (2013b), there is virtually no study comparing changes in orchid-bee community compositions in mid to long terms, and we largely ignore what is happening to populations of these forest-dependent species in most areas.Studies focusing on this approach are, thus, urgently needed.

Figure 1 -
Figure 1 -Map showing the exact locations of 'Reserva Particular do Patrimônio Natural Feliciano Miguel Abdala' (RFMA) and the nearby 'Parque Estadual do Rio Doce' (PERD), also mentioned in the text.

Table 1 -
Diversity, evenness, species richness and number of specimens of each orchid-bee species collected at sites 1 (forest) and 2 (edge) at RPPN Feliciano Miguel Abdala in late December, 2011.