Sites
Carajás national forest
The Carajás National Forest (05°52′S–06°33′S, 49°53′W–50°45′W) is located in the southeastern portion of Pará State, Brazil (Fig. 3a). Carajás is an Amazonian domain area, mainly covered by forest formations38,39. It is located at an altitudinal range of 700 to 800 meters above sea level. The climate in this region is characterized as rainy tropical with winter drought (AWi), according to Koppen’s classification, with an annual precipitation range of more than 2000 mm and a well-defined drought period from June to September. The average temperature ranges from 25° to 26 °C, but the absolute recorded values range from 15 °C, from July to October, to 38 °C in the remaining months of the year. The predominant vegetation cover is evergreen ombrophilous forests; however, there are also areas of stationary vegetation with different degrees of deciduousness39. The largest Brazilian mineral extraction project is located in Carajás and deals primarily with the extraction of iron ore but also other minerals. It was initiated in the late 1980s and remains active to the present date.
Study sites. (a) Serra dos Carajás in Pará State, eastern Amazon, northern Brazil. (b) Nova Lima in Minas Gerais State, Iron Quadrangle, Southeastern Brazil.
Nova lima
Nova Lima is a Brazilian municipality in the state of Minas Gerais (Fig. 3b). Located in a mountainous landscape surrounded by the Serra do Curral and Serra da Moeda (altitude range: 900 to 1400 m), it is part of the region denominated as the “Plateaus and Mountain ranges of the East-Southern Atlantic” and is mainly composed of metamorphic rocks. The climate is hot and temperate and is classified as Cfa (subtropical humid climate). The average rainfall is 1390 mm per year, with August being the driest month and December the month with the highest precipitation. The average temperature is 23 °C; January is the warmest month of the year, and the lowest temperature occurs in June, i.e., an average temperature of 17.6 °C. The area of the municipality is 430 km2, and the population is estimated to be 92,000 inhabitants with a 0.8 HDI (human development index) (Brazilian Institute of Geography and Statistics – IBGE). The region has intense mineral extraction activities, including the mines of Morro Velho, Mostardas and Rio de Peixe, in addition to the recently closed mine of Águas Claras; the most important minerals are iron and gold. The city of Nova Lima integrates the metropolitan region of the state’s capital, Belo Horizonte (with approximately 2.5 billion inhabitants).
Experimental and sampling design
The list of bee species names was obtained from two Brazilian entomological collections as the main repositories for specimens collected in both Carajás and Nova Lima, i.e., the Museu Paraense Emilio Goeldi (MPEG) entomological collection and the Universidade Federal de Minas Gerais (UFMG) entomological collection. Neither of these collections have online databases. Specimens of both collections were validated by specialists (for specimens’ IDs see37). Therefore, our dataset contains those records whose specimens were located and certified by specialists for both collections37.
At MPEG, we found bees from Carajás that were collected in the Serra Norte area (Fig. 3a) from 1983 to 2018. At UFMG, we found bees from Carajás that were collected in the Serra Norte and Serra Sul areas (Fig. 3a) during the 2008 to 2017 period. Bees from Nova Lima (Fig. 3b), collected from 1998 to 2017, were found only at UFMG. Bees were mainly collected with the use of entomological nets (active search), odoriferous traps (for male orchid bees, tribe Euglossini) and flight interception traps (malaise). Sampling efforts were not standardized, as a number of different researchers conducted their field work at these locations to answer several unrelated questions.
A total of 328 bee species have been recorded in these two areas, corresponding to more than 20% of the estimated 1500 bee species recorded in Brazil40. A total of 222 species were recorded in Carajás, representing nearly 80% of the bee fauna in Pará State, eastern Amazon (the 2nd largest Brazilian state and the world’s 13th largest state)36. More than 30% of these species are social (mostly the Meliponini tribe, but also the Bombini and Apini tribes), nesting mainly in pre-existing cavities (47%) or in the ground (22%), and 33% of these bees (at least at the generic level) have been identified as crop pollinators. A total of 144 species were recorded in Nova Lima, representing 65% of the bee fauna in Minas Gerais State (the 4th largest Brazilian state)36. Nearly 80% of these species are solitary, more than 50% nest in the ground and 37% have been reported to be pollinators of Brazilian crops.
Although extensive sampling efforts were previously carried out in these areas, our dataset does not represent a final list but rather the consolidation of past sampling efforts together with species traits that are of great importance for ecological analysis and pollination services valuation10. New field sampling strategies are currently being carried out, following standardized methodology, to equalize sampling efforts and provide future comparison. Further steps will include a better understanding of the ecological functioning in these areas, with practical implications for the ecological restoration of mine-land rehabilitated areas in the Amazon Forest. Additionally, bee taxonomy for the Neotropical region is not complete, and recent taxonomic reviews continue to reveal new species, mainly from the less studied and most speciose areas, such as tropical rain forests36,41,42.
From this large dataset, it will be possible to answer more complex questions on the functioning of these ecosystems, such as the importance of the bees in each of these environments, the value of the ecosystem services they may provide and how these species will respond to ongoing global changes. Our database adds a robust set of trait-based information for Brazilian wild bees. We apply techniques that were previously reported in the literature9,10,24,26,27 and that are necessary for the advancement of functional ecology and the understanding of wild bees and their role in crop pollination and the provision of ecosystem services28,29.
Traits acquisition and laboratory research methods
Body size, size classes and flight range
Bees from the two abovementioned entomological collections were analyzed under a stereomicroscope coupled with a calibrated micrometric ocular. Body size measurements were based on the bees’ intertegular distance (hereafter ITD) (Fig. 4i). We measured up to five specimens of each species and the average value was calculated37. ITD measures represent the mesoscutum width of each bee species, which is the body section where the alary muscles are located. Bees with a larger ITD have been shown to be able to fly longer distances across landscapes26. The estimation of the flight distance of each bee species was calculated from the ITD measurement and taxonomic position using equations presented in27. Flight range estimations were based on models generated from previous on-site experiments for both social43,44 and solitary bees45,46. Recorded flight range experiments used two standard methodologies: (1) the release of marked bees at known distances from their nests and their recapture at the nest entrance (homing distance), considering the maximum homing distance (mhd – 90% return rate) and typical homing distance (thd – 50% return rate) and (2) feeder training techniques, which record the maximum energetically profitable foraging distance for a bee to forage at an artificial feeder (maximum foraging distance – mfd) and the distance to which bees stop recruiting nest mates to an artificial feeder (maximum communication distance – mcd)26.
Bees body size and their relation to flight range. (a) Photos of bees showing their different body size classes. (a) Lateral view of Hylaeus tricolor (Schrottky, 1906) (b) Augochloropsis callichroa (Cockerell, 1900) (c) Melipona seminigra Moure & Kerr, 1950 (d) Megachile orba Schrottky, 1913 (e) Euglossa amazonica Dressler, 1982 (f) Bombus transversalis (Olivier, 1789) (g) Eulaema cingulata (Fabricius, 1804) (h) Xylocopa frontalis (Olivier, 1789). Photos by Fernanda Trancoso. (i) Dorsal view of a Euglossini bee with ITD (intertegular distance) measure. Photo by Rafael Borges. (b) Non-linear correlation between bees body size and flight range (estimated as the highest measurement obtained), dots represent bees in our data set and colors indicate body size categories (r² = 0.88).
We classified bees into three body size categories according to their ITD measures: small, medium and large (Fig. 4a). As body size and flight range are positively correlated, our body size categories also indicate maximum flight range, being up to 1 km for small bees, up to 5 km for medium bees and up to 30 km for large bees (Fig. 4b). Body size categories were established using the body size of bees in the genus Melipona Illiger as a standard for the medium-size class. Thus, ITD measures for the medium-size class ranged from 2.2 to 3.9 mm, with 2.2 mm being the ITD of Melipona amazonica Schulz, 1905, the smaller Melipona species in our dataset, and 3.9 mm being the ITD of Melipona fuliginosa Lepeletier, 1836, the larger Melipona species in our dataset. Bees with an ITD smaller than 2.2 mm were classified as small (e.g., Trigonisca, Plebeia and Augochlora), and bees with an ITD larger than 3.9 mm were classified as large (e.g., Xylocopa, Centris and Eulaema). Although arbitrary, our body-size classes are related to the fact that Melipona bees are commonly known as medium-sized bees.
Additionally, our classes agree with those previously established in other papers24,47,48,49,50 although some of those papers did not include Melipona bees. Therefore, we believe our classification is an appropriate standard for size classification of bees worldwide. To date, ours is the most species-rich dataset, which represents five bee families (Apidae, Andrenidae, Colletidae, Halictidae, Megachille) and includes the widest range of body size measurements reported. For example, the dataset includes the very large species Xylocopa frontalis (Olivier, 1789) (ITD = 8.4 mm) and also the minute Trigonisca variegatifrons Albuquerque and Camargo, 2007 (ITD = 0.6 mm).
Taxonomic ranks, known distribution, distribution area, new record, locality and location of measured specimens
Information on the taxonomic ranks (family, tribe, genus, subgenus, specific epithet, scientific name authorship) and known Neotropical distribution of each species were acquired using the online version of Moure’s bee catalog (available at http://moure.cria.org.br). Additionally, we searched and included updated information from the literature (after 2014, which was the last catalog update) when available. We chose to use Michener’s32 family classification instead of that provided in the catalog, as this classification is mostly used in other regions of the world besides Brazil. We included new occurrence records for all the specimens collected out of the known distribution provided in the Moure’s bee catalog.
Locality represents where each species was collected in our study sites. Localities from Carajás National Forest are Bocaina and Canaã dos Carajás from Serra Sul area and Carajás and Parauapebas from Serra Norte area. Remaining bee species were collected in Nova Lima.
Location of measured specimens indicate in which entomological collection (MPEG or UFMG) we found and measured the specimens collected at our study sites. For each of the specimens measured at the entomological collections (MPEG and UFMG) we provide the complete label information as well as specimen ITD measure37.
Crop pollination
Information on crop pollinator species was obtained from the database published by Giannini et al.51, which includes an extensive assessment of crop pollinators in Brazil. This database contains all the published interactions recorded between Brazilian crops and their pollinators, including the information source and the type of interaction recorded (effective, occasional or potential pollinator or simply a visitor).
In our dataset, we specify which species are known crop pollinators and also, for each bee species we specify which crop they were reported to pollinate37. Knowledge of crop pollinators is of great importance, not only in ecology and ecosystem functioning, but mainly for directing conservation strategies and policy decision making for ensuring food security. This knowledge can also be used as a means to evaluate ecosystem services provided by bees, to understand landscape population structure and to establish management and conservation strategies for individual species.
Sociality and nest location
Sociality and nest location data for each species were acquired by consulting previous natural history or review articles on these subjects. Primarily, a search for natural history data was performed for each species. For those species that completely lack natural history information, we searched for subgeneric information (when available), and when this information was not available, we included generic and tribe natural history information for the particular species. In all cases, before using generic, subgeneric and tribe information, a search in the specific literature was performed to verify the variation level in natural history traits for the specific clade.
We found sociality and nesting information for all species except some Euglossa Latreille, 1802 species. For 23 of the 43 Euglossa species, we did not find natural history data. Natural history traits vary widely within this genus, and the use of subgeneric or generic classification is not applicable for Euglossa species.
We provide the accuracy level of the information provided (i.e. tribe, genus, subgenus, species) and the reference papers from which natural history information was gathered for each species. The references are in the following format: authors, year, article title, journal, journal volume and pages37.
Source: Ecology - nature.com
