Nesting biology of three Megachile (Hymenoptera: Megachilidae) species from Eastern Amazonia, Brazil

Abstract

Megachile Latreille is a conspicuous genus of solitary bees distributed worldwide. However, the biology of tropical species is still little known. We present data on biology of Megachile brasiliensis Data Torre, Megachile sejuncta Cockerell and Megachile stilbonotaspis Moure found in two remnants of eastern Amazonian forest in northeastern Brazil. The study was conducted using the trap-nest methodology in two different areas during four periods. We collected a total of 24 nests of M. brasiliensis, 26 of M. sejuncta and 28 of M. stilbonotaspis. The differential abundance of collected nests may reflect the population size in each sampled place. The nesting activity was concentrated mainly between July and January and species presented a multivoltine pattern, except for M. sejuncta, which was partly univoltine. Assessed pollen use showed a predominant use of Attalea sp. (Arecaceae) and, for M. stilbonotaspis, Tylesia sp. and Lepidaploa sp. (Asteraceae). Babassu is a very common palm in the studied areas and the studied species seem to have a strong link with it. We also reported change of pollen use by M. sejuncta, probably due to competition with M. brasiliensis, which may have influenced the biased sex ratio observed in M. sejuncta toward males. Parasites reported here were also recorded for other Megachile species, such as Coelioxys, Brachymeria, Meloidae and Pyralidae species. Mites were observed in association with M. stilbonotaspis. The data presented here set up a background that encourages new studies on the ecology of these three Amazonian species, providing tools for proper biodiversity management and conservation.

Keywords
Leaf-cutter bee; Megachilini; Pollination; Solitary bee; Trap-nest

Introduction

Megachile Latreille (Hymenoptera: Megachilidae) is a large genus of solitary bees, with about 2645 species distributed around the world (^{Michener, 2007}; ^{Raw, 2007}). Approximately 524 species are distributed in the Neotropics, but this number may surpass 600 species (^{Raw, 2007}). Due to the group diversity, many problems exist relative to species position among the several subgenera, hindering identification of specimens collected on specific level, and thus the available data on their nesting biology (^{Raw, 2007}).

Nesting habits of Megachile are conspicuous among solitary bees, mainly due to how they build their nests: using leaves, leaf fragments and/or petals to line their brood cells (^{Eickwort et al., 1981}; ^{Michener, 2007}; ^{Torretta et al., 2014}). However, there is great variability of how such nests are built, and several types of materials may be employed such as clay, sand, chewed leaves, vegetal matter, man-made cavities, etc. (^{Krombein, 1967}; ^{Michener, 2007}; ^{Sheffield, 2017}). This knowledge is important since the material used to build nests may provide clues about female morphological functionality, especially their mandibles (^{Laroca et al., 1992}). Therefore, nest architecture may provide useful characters for taxonomy and phylogeny of those species, contributing to a better understanding about the evolution of the genus (^{Gonzalez, 2008}).

Substrates used to build nests may be cavities in the ground, pre-existing cavities in wood or even living plant stems (^{Eickwort et al., 1981}; ^{Michener, 2007}; ^{Orr et al., 2015}). The behavior of nesting in pre-existing cavities associated with the use of the trap-nest methodology (^{Krombein, 1967}) facilitates obtaining solitary bee specimens. Trap-nests also allow research on biology, life cycle and associated organisms. Moreover, we can infer on plant hosts used to provision for immature larvae (^{Buschini et al., 2009}; ^{Torretta et al., 2014}).

The trap-nest methodology also allows research on pollination, since some solitary bee species that nest in pre-existing cavities are pollinators of native plants, and in some cases are the foundation for bee management on economically important crops, such as Megachile rotundata Fabricius (^{Pitts-Singer and Cane, 2011}), although few studies report pollen content used by bees, especially in the Neotropical region (^{Buschini et al., 2009}; ^{Strickler et al., 1996}; ^{Rocha-Filho and Garófalo, 2016}; ^{Torretta and Durante, 2011}; ^{Torretta et al., 2014}). In Brazil, Megachile seems to prefer Fabaceae and Asteraceae (^{Garófalo et al., 2004}) and has great potential for management in crops of economic interest (^{Pitts-Singer and Cane, 2011}).

Megachile species are better studied in the south of Brazil, whereas few studies were conducted in the Amazonian region (^{Morato, 2003}; ^{Pinheiro-Machado et al., 2002}). Regarding this knowledge gap, the aim of this study was to describe nesting biology of three Megachile species in Amazonian forest remnants and Restinga from Maranhão Island and forest areas at “Baixada Maranhense”, Maranhão State, Northeastern Brazil.

Material and methods

The study was conducted in four different areas in the Maranhão State, Brazil: in Maranhão Island, at São José de Ribamar municipality, in “Sítio Aguahy” (2º 38' S 44º 08'O) belonging to “Quercegen Agronegócios I LTDA” (henceforth referred to as “SA”); and in “Área de Proteção Ambiental da Baixada Maranhense”(henceforth referred to as “APABM”), in three municipalities: “Peri-Mirim” (“Parque Agroecológico Buritirana” - 2º38'S 44º50'O), “Penalva” (“Canadá Farm” - 3º17'S 45º07'O) and “Vitória do Mearim” (“Coque Farm” - 3º32'S 44º57'O) (Fig. 1). The three areas sampled in the APABM will be henceforth referred to as “PM”, “PE” and “VM”, respectively. All areas are part of the Amazonian biome.

Fig. 1
Study areas and sampling locations in the Maranhão state, Brazil.

SA comprises about 600 ha of transitional vegetation, merging remnants of Amazonian forest, mangrove and restinga. Some forest remnants are better preserved, being called primary forest, whereas other forest remnants are called secondary forest, due to the high number of trails and crops, constituting a mosaic of environments. Restinga area is characterized by vegetation adapted to high salinity, with plants found in herbaceous, shrubby, and arborous layers (^{Serra et al., 2016}).

APABM is characterized by a transition zone between “Cocais forest” to the south, Amazonian forest to the west and marine systems to the north, all of these with Attalea speciosa Mart. Ex Spreng palms (Arecaceae). It is located in the largest set of lake basins of the Brazilian Northeastern region and is subjected to seasonal flooding (^{Pinheiro et al., 2010}). PM harbors secondary forests of A. speciosa mainly and herbaceous fields that flood only partially. PE also has secondary forests, but these are surrounded by flooded fields of herbaceous plants during the rainy season. VM is the best preserved area, and has primary forests with low human interference, besides gallery forests typical of those with springs and small water streams.

Following ^{Peel et al. (2007)}, climate at SA is characterized as type Awh, with high temperatures throughout year and a rainy season concentrated during January to July; the annual average rainfall reaches 1900 mm. At BM, the climate varies between Am and Aw and the annual rainfall regime varies from 1000 to 2000 mm, concentrated from February to May. Rainfall data from each area during all periods were provided by “Laboratório de Meteorologia do Núcleo Geoambiental da Universidade Estadual do Maranhão”.

We used the trap-nest methodology (henceforth “traps”) to sample the study areas (^{Krombein, 1967}). During the SA phase, traps made using Tatajuba wood with diameters of 8, 10, 12, 14 and 16 mm were set in groups of ten inside polyethylene terephthalate (PET) bottles, two of each diameter, each set considered a sampling unit (SU). At APABM, traps with diameters of 4, 6, 8, 10, 13 and 15 mm were used, two for each diameter, comprising 12 traps per SU placed in rectangular tubes of polyvinyl chloride (PVC). The tunnel of each trap was 11 cm deep and the SUs were placed hanging from three branches at 1.5 m off the ground. The number of SUs in field and the period of study in each area are summarized in Table 1.

Table 1
Arrangement of sampling units (SU) in both areas during the four study periods.

Sample periods	Sitio Aguahy				APABM			Total
	PF	SF	RE		VM	PE	PM
Period I (Mar/2007-Feb/2008)	20	30	-		-	-	-	500
Period II (Feb/2009-Jan/2010)	20	20	-		-	-	-	400
Period III (May/2011-Apr/2013)	15	15	12a		-	-	-	420
Period IV (Jul/2013-Jul/2014)	-	-	-		18	18	18	540

• PF, primary forest; SF, secondary forest; RE, restinga; VM, Vitória do Mearim; PE,Penalva; PM, Peri-Mirim.
• ^a
  Only 20 traps of 16 mm were used during this period.

Each SU was inspected monthly and traps occupied were replaced by empty traps and brought to laboratory to be monitored until adult emergence, when we registered date and order of emergence. Adults were sacrificed in jars with Ethyl Acetate, sexed, and pinned. Voucher specimens were deposited in the LESPP Collection of the Biology Department at “Universidade Federal do Maranhão”, in the Entomological Collection of the Taxonomic Collections at “Universidade Federal de Minas Gerais” and in the Padre Jesus Santiago Moure Collection at “Universidade Federal do Paraná”.

Nests where adults had emerged or died were used to describe architecture. Measurements of the nest and brood cells were taken. The process of nest analysis was recorded through photographs. This study is a data compilation of three studies carried out at different periods and with different traps as showed in Table 1. Collects occurred during periods I and II were meant to perform a survey of solitary Hymenoptera species which nest in pre-existent cavities. The aim of collects during Period III was to study of Megachile species biology, then development time of bees were recorded and data presented is considering this period and period IV. Collects during Period IV aimed to evaluated aspects of vegetation cover and nesting activity of solitary Hymenoptera (DBM, unpublished data). For building activity we followed the methodology described by ^{Cardoso and Silveira (2003)} to infer the span of minimum and maximum development time of bees. For the first span we considered the time of development as being from the day the nest was collected up to the emergence of the adult bee in the laboratory; for the second span of development time we considered that the first bee was originated from an egg laid the day after the last survey of nests until the emergence of the adult bee in laboratory. Regarding bee mortality, brood cells were opened a year after nests were collected, in the case of Periods III and IV. Nests of Period I and II had their brood cells analyzed during Period III when bees did not emerge.

Acetolysis process was used to analyze the residual pollen content and larval feces from brood cells following protocols by ^{Erdtman (1943)} modified by ^{Lieux (1980)}. Three slides were made from each nest and 400 pollen grains were counted whenever possible. We followed ^{Michener (2007)} classification in order to classify bees according to pollen contents. Statistical tests were run in the programming package R version 3.2.2 (^{R Development Core Team, 2015}). Data normality was assessed using the Shapiro-Wilk test. Correlation between rainfall and nesting was evaluated through Fisher's exact test; sex-ratio of emerged bees and the influence of diameter in trap-nest use was assessed using the Chi-square test. Occupancy rate (OR) of the trap-nests was calculated by dividing the total of nests occupied during the study in each sampled period by the total of nests offered in the field (^{Table I, supplementary material}). Confidence intervals used in tests were p < 0.05.

Results

Nest abundance and nesting activity

We collected 78 nests of Megachile in the studied areas: 24 nests of M. (Melanosarus) brasiliensis Dalla Torre, 26 of M. (Austromegachile) sejuncta Cockerell, and 28 of M. (Pseudocentron) stilbonotaspis Moure. Occupancy rate of traps at SA was 3.75% during period I, 1.25% in period II and 4.04% in period III; at APABM, occupancy rate was 7.77% considering all the areas together in APABM (but see ^{Table I} of the ^{supplementary material}, for detailed occupancy rates). The most abundant species were M. brasiliensis at SA and M. stilbonotaspis at APABM. Detailed number of nests by species and area, as well as emerged specimens are shown in Table 2.

Table 2
Data of emerged adults (excluding parasites) from Megachile at Sítio Aguahy and APABM (Penalva, Peri-Mirim and Vitória do Mearim) in Maranhão state, Brazil.

Species	Sítio Aguahy					Penalva					Peri-Mirim					Vitória do Mearim
	Nests				ES	Nests				ES	Nests				ES	Nests				ES
	N	Ø	(n)	N	%	N	Ø	(n)	N	%	N	Ø	(n)	n	%	N	Ø	(n)	N	%
M. brasiliensis	19	8	(8)	55	59.1	-	-		-	-	4	8	(2)	16	59.2	1	10	(1)	1	1.7
		10	(2)									6	(2)
		12	(4)
		14	(3)
		16	(2)
M. sejuncta	13	8	(2)	30	32.4	2	6	(1)	3	10.7	1	10	(1)	1	3.7	10	8	(4)	34	56.6
		10	(9)				10	(1)									10	(4)
		12	(1)														13	(1)
		14	(1)														15	(1)
M. stilbonotaspis	5	10	(2)	7	7.5	9	6	(4)	25	89.3	4	6	(2)	10	37.1	11	8	(6)	25	41.7
		12	(2)				8	(3)				10	(2)				10	(2)
		14	(1)				10	(2)									13	(3)
Total	37			92		11			28		9			27		220			60

• Ø, trap diameter; (n), absolute number of traps used or emerged specimens; ES,emerged specimens.

Megachile brasiliensis and M. sejuncta nested mainly between July and January in both areas (Fig. 2). Due to few M. stilbonotaspis nests collected at area A we cannot infer any nesting pattern, as this species founded nests sparsely during the study period. Only M. stilbonotaspis nesting activity was correlated with rainfall at area B (Fisher's exact, p = 0.003).

Fig. 2
Recorded rainfall and number of sampled nests during the study in SA and APABM: (A-C) nests collected and rainfall recorded in SA; (D) nests collected and rainfall recorded in APABM.

The estimates for minimum and maximum development time of bees are summarized in Fig. 3. At SA, M. brasiliensis was multivoltine; M. stilbonotaspis could not have its emergence pattern assessed because some individuals emerged before the next survey of nests. Megachile sejuncta had a population partially univoltine; some individuals (n = 4) from two nests collected in August/2011 and two nests collected in April/2012 (n = 4) had their development delayed and took more than one year in average to emerge (Fig. 3). At APABM, all species had a multivoltine pattern. Number of generations by species and local studied are presented in Fig. 4.

Fig. 3
Estimatives of minimum and maximum time development of Megachile bees in SA and APABM.

Fig. 4
Estimated number of generations of Megachile bees in SA and APABM. The number on the columns refer to the number of generations in each period (roman numerals) by each area sampled (A) “Sítio Aguahy”; (B)“Área de Proteção Ambiental da Baixada Maranhense”.

Pollen use

We performed acetolysis on pollen residuals and bee feces from ten M. brasiliensis nests, eight M. sejuncta nests and four M. stilbonotaspis nests from SA; from BM, we analyzed pollen residuals and bee feces of three M. brasiliensis nests, six M. sejuncta nests and thirteen M. stilbonotaspis nests (Tables 3 and 4).

Table 3
Pollen types used by Megachile species at Sítio Aguahy, Maranhão Island, Maranhão State, Brazil.

Family	Pollen type	M. brasiliensis	M. sejuncta	M. stilbonotaspis
Arecaceae	Attalea sp.	89.23%	59.09%	34.05%
Asteraceae	Tilesia sp.	0.34%	-	34.05%
	Lepidaploa sp.	-	-	31.90%
Euphorbiaceae	Croton sp.	0.59%	0.52%	-
	Phyllanthus sp.	1.79%	0.11%	-
Gentianaceae	Coutoubea sp.	0.57%	-	-
Lamiaceae	Hyptis sp.	-	0.18%	-
Myrtaceae	Myrcia sp.	-	28.88%	-
Rubiaceae	Spermacoce sp.	4.28%	-	-
Onagraceae	Ludwigia sp.	3.20%	0.07%	-
-	Type 1 sp.	-	11.15%	-

Table 4
Pollen types used by Megachile species at APABM, Maranhão State, Brazil.

Family	Pollen type	M. brasiliensis	M. sejuncta	M. stilbonotaspis
Arecaceae	Attalea sp.	99.97%	96.48%	54.43%
Asteraceae	Lepidaploa sp.	-	-	8.53%
Asteraceae	Tilesia sp.	-	-	16.26%
Asteraceae	Vernanothura sp.	-	-	0.90%
Convovulaceae	Ipomea sp.	-	-	0.97%
Euphorbiaceae	Croton sp.	-	-	0.89%
Fabaceae	Cesalpinoidea sp.1	0.03%	-	-
Fabaceae	Type 1	-	0.07%	-
Fabaceae	Bahunia sp.1	-	0.01%	-
Lamiaceae	Hyptis sp.	-	-	0.64%
Myrtaceae	Myrcia sp.	-	2.85%
Onagraceae	Ludwigia sp.	-	-	5.92%
Passifloraceae	Type 1	-	0.59%	-
Phyllanthaceae	Phyllanthus sp.	-	-	2.52%
Pontederiaceae	Aff. Eichornia sp.	-	-	3.37%
Solanaceae	Type 1	-	-	0.78%
Unidentified	Type 2	-	-	3.08%
Unidentified	Type 3	-	-	1.72%

M. brasiliensis was classified as oligolectic, using mainly pollen of type Attalea sp. (Arecaceae) (Fig. 5A). However, we observed one nest at SA whose pollen content varied greatly, with pollen types of Spermacoce sp., Ludwigia sp., Phyllantus sp., Cotobea sp. and Tilesia sp. At APABM, almost 100% of pollen grains counted belonged to Attalea sp.

Fig. 5
Main pollen types identified on Megachile slides from larvae feces: (A) type Attalea sp.; (B) type Tilesia sp.; (C) type Lepidaploa sp.; (D) type Myrcia sp. Scale bar of figures A - C 10 µm; figure D: 20 µm.

Pollen types of Attalea sp., Tilesia sp. (Fig. 5B) and Lepidaploa sp. (Fig. 5C) were counted at all areas studied for M. stilbonotaspis, classified as narrowly polylectic. At PE and VM, the pollen type Attalea sp. was not present in August/2013 and May/2014 and pollen types of Tilesia sp. and Lepidaploa sp. were abundant on the slides analyzed. M. sejuncta was classified as narrowly polylectic at SA.

Nests of M. sejuncta presented pollen types of Attalea sp. and Myrcia sp. in the nest slides of February and Attalea sp. as predominant in April and July. An unidentified pollen type was present in nests collected in August/2011 (n = 1) and January/2012 (n = 3). Type Myrcia sp. (Fig. 5D) was predominant in samples of nests collected in February/2008 (n = 1) and December/2011 (n = 1). At area B, M. sejuncta used mainly type Attalea sp. (Table 4).

Sex ratio and diameter cavity occupancy

Sex ratio did not differ from 1:1 in M. brasiliensis and M. stilbonotaspis. For M. sejuncta, sex ratio was significantly biased toward males at SA, and toward females at APABM, although the latter was only marginally statistically significant (Table 5).

Table 5
Sex ratio and χ² values of Megachile emerged from nests in SA and in APABM.

Species	SA					APABM
	(♂:♀)	SR	χ2	p-Value		(♂:♀)	SR	χ2	p-Value
M. brasiliensis	26:25	1.08:1	0.0196	0.886		07:10	0.7:1	0.52941	0.4669
M. stilbonotaspis	-	-	-	-		30:29	1.03:1	0.0169	0.8964
M. sejuncta	19:05	3.03:1	5.761	0.016		14:26	0.53:1	3.6	0.0577

• ♂, males; ♀, females; sex ratio, SR.

Nests of 8 and 10 mm diameter were used the most by M. sejuncta in SA (χ² = 20.462; p < 0.001) and in APABM (χ² = 12.385; p < 0.05) (Table 2). The same trend was observed for M. stilbonotaspis in APABM where 6, 8 and 10 mm were used the most (χ² = 24.5; p < 0.001) (Table 2).

Nest architecture: general aspects

Nests of all three species were characterized by having brood cells lined and covered with layers of interwoven leaf fragments; nests had a cylindrical shape (Fig. 6). No vestibular cell was accounted for; neither were any intercalary cells or any gaps between the first and last cells built (Table 6).

Fig. 6
Photographs of nests from the three Megachile species: (A)-(C) M. brasiliensis; (D)-(F) M. sejuncta; (G)-(I) M. stilbonotaspis. Nests A, D and G were collected at SA, whereas nests B, E and H were collected at APABM. Scale bar = 1 cm.

Table 6
Nest architecture data taken from Megachile nests at area A and B. Measures were taken in centimeters.

Species	Area	Nests (n)	ML ±SD	BC(n)	BC/N	ML ±SD
M. brasiliensis	AS	12	6.6 ± 1.60	41	1-5	0.97 ± 0.20
	APABM	3	7.65 ± 2.69	17	3-7	1.07 ± 0.91
M. sejuncta	SA	11	5.17 ± 1.85	23	2-5	1.02 ± 0.16
	APABM	10	7.69 ± 1.85	37	2-6	1.20 ± 0.18
M. stilbonotaspis	SA	4	5.04 ± 0.39	12	3	1.20 ± 0.12
	APABM	11	6.80 ± 1.17	51	3-6	1.16 ± 1.99

• ML, mean length; SD, standard deviation; BC, Brood cells; BC/n, amplitude of broodcells by nests.

Each brood cell in a line series was snugly fitted, so that the apex of one brood cell faced the bottom of the next one (except for the innermost brood cell, which faced the bottom of the trap, and outermost brood cells, which faced the nest's opercula). The brood cells were covered with a batch of leaf fragments, making it difficult to observe how many cells were in the whole nest; but most nests were easily disassembled, and brood cells were separated without cell damage (Fig. 6, but see M. sejuncta description below). In all nests analyzed, feces of the immatures were adhered on the brood cell walls. Leaf fragments used to build the nests had indentations around the edges, giving them a crenulated outline. The leaf fragments used to build the brood cell walls were cut through the mid vein of the leaves and pressed together by the adult females, using their mandibles.

The nest of Megachile brasiliensis

Nests of this species did not differ between areas (Fig. 6A and B) and were typically covered by elongated subtrapezoidal pieces of leaves, with one side of the fragment much more rounded than the other (Fig. 6C). The sharp apex of the pieces points upwards, giving a scaly aspect to the nest's surface. The leaf fragments forming the walls of the brood cells were tightly fitted on the sides, forming a strongly fastened cup (although no trace of any gluing substance was found); those leaf fragments above the brood cell walls were partially overlapped, forming a projection that fits to the bottom of the next cell. The bottom of the cup was formed by the posterior side of bent leaf fragments (Fig. 7C). The cap of the brood cell was formed by several rounded pieces of leaves, which were bigger in diameter than the cup, so that when the brood cell was closed, these leaf fragments acquired a convex shape. After the last cell was built, several rounded pieces of leaves were stuck to the cavity floor, forming the nest opercula.

Fig. 7
Brood cells of Megachile species: (A) nest supersedure of Tetrapedia rugulosa over a Megachile brasiliensis nest; (B) cavity sharing by M. brasiliensis and a unidentified species of Centris; (C) unattached cells of M. brasiliensis; (D) normal and big-sized cell of M. brasiliensis; (E) M. sejuncta brood cells; (F) M. stilbonotaspis brood cells. Scale bar = 1 cm.

Two nests in a cavity of 16 mm had each two series of lined brood cells in the same tunnel (Fig. 6A). We also found another nest with two cells, from which three bees emerged, and with an abnormally big outermost cell (Fig. 7D), from which two males emerged. A nest of M. brasiliensis built in a 14 mm cavity had another nest of Tetrapedia rugulosa Friese (Hymenoptera: Apidae) built in front of it (Fig. 7A), evidencing nest supersedure. Another cavity was shared by M. brasiliensis with another Hymenoptera species, which we were unable to identify due to the fact that the larvae did not survive until adulthood. However, we suspect it may be Centris sp. because of the nest architecture (Fig. 7B).

The nest of Megachile sejuncta

Nests of this species varied between the two sites regarding types of leaf fragments used to build brood cells (Fig. 6D and E; Fig. 7E). The leaves used in area A were more coriaceous, while in area B leaves were coriaceous (apparently the same species used in area A) and others were hirsute. Leaf fragments were overlapped and not tightly fastened, so that when removing the nest from the trap, some brood cells got detached. The nest opercula had several leaf fragments arranged from the last cell built up to the entrance of the boring.

Leaf fragments used to construct the brood cell walls and the upper layers of the nest walls were elongated, the longer sides cut in a straight line, while the shorter sides had a more rounded cut. The brood cells were formed by leaf fragments fastened by their sides, forming the cell cup; the bottom of the nest is formed by leaf fragments bent to adjust the bottom nest structure to the substrate. The cap of the cell is formed by several rounded leaf fragments and has a bigger diameter than the nest structure, sealing it with a convex shape.

The nest of Megachile stilbonotaspis

This species presented noticeable variation in its nest structure, with two kinds of leaf material used to construct nests: leaf fragments and leaflets (Fig. 6G and H). The leaflets were hirsute and cut out at the petiole, so that the whole structure was used to build the cup of the brood cells. Leaf fragments varied regarding cut patterns ranging from almost rectangular shaped, to oblong (Fig. 6I). The bottom of each cell was built by bending the tip of the leaflet or the leaf fragment. The cap of the brood cells was built with rounded leaf fragments and seemed to be made from leaflets (Fig. 7F). The diameter of such rounded fragments was almost the same as the nest diameter, fitting snugly with the entrance of the cell. The nest opercula had several leaflets and leaf fragments arranged from the last built cell and the entrance of the boring. The pattern of leaves used did not vary between areas; however, the leaflets seemed to be from two different species of Fabaceae. Some of the leaflets had cut marks around the edges, probably made by the bee to better fit them into the nest.

Mortality and associated species

Data on general and specific mortality rates are summarized in Table 7, and in ^{Table II of the supplementary material} for detailed number of brood cells. The highest mortality rate recorded, considering all periods sampled, was for M. brasiliensis at SA by cleptoparasitism. Brachymeria sp. 1 and Coelioxys tolteca Cresson parasitized nests of M. brasiliensis in periods I and II. In period III, Coelioxys otomita and Pyralidae (Lepidoptera) were recorded. At APABM, C. otomita Cresson was the only parasite of M. brasiliensis and one bee died of unknown causes as an adult, but only fragments of the body was found.

Table 7
Mortality rate, parasites and pathogens recorded for the Megachile nests at SA and three municipalities of APABM.

Hosts	Mortality cause		Sítio Aguahy			Penalva	Peri-Mirim	Vitória do Mearim
		PI	PII		PIII	PIV	PIV	PIV
M. brasiliensis	Brachymeria sp.1	6	-		-	-	-	-
	C. tolteca	1	-		-	-	-	-
	C. otomita	-	-		3	-	1	-
	Pyralidae	-	-		2	-	-	-
	Fungi	-	-		6	-	-	-
	Unknown causes	-	-		3	-	1	-
	% BC mortality	15.9	0		58.3	0	11.76	0
M. sejuncta	Brachymeria sp.2	1	-		-	-	-	-
	Pyralidae	-	-		2	-	-	-
	Unknown causes	6	-		-	-	-	-
	% BC mortality	58.3	0		14.28	0	0	0
M. stilbonotaspis	C. tolteca	-	1		-	-	-	-
	C. simillima	-	-		-	1	-	-
	C. otomita	-	-		-	-	1	-
	Meloidae	-	-		-	-	-	1
	Unknown causes	-	-		-	3	-	-
	% BC mortality	0	20		0	17.39	9	5

• Brood cells, BC.

We recorded cleptoparasitism in M. sejuncta only in period III at SA by Bachymeria sp.2 and Pyralidae (Lepidoptera). At APABM, only deaths by unknown causes were reported, all in larval stage.

Regarding M. stilbonotaspis, C. tolteca was recorded at a secondary forest in SA; at APABM, we recorded Meloidae (Coleoptera), Ceolioxys simillima Smith, C. tolteca and C. otomita, besides unknown mortality causes. We also observed the occurrence of mites on the bodies of M. stilbonotaspis specimens emerged from Restinga nests of SA. They were found at the forewings, propodeum and first tergite of the bees.

Discussion

We have here provided data on nests of three species of Megachile for the first time. Previously, two species of bees from the subgenus Melanosarus had their biology described: M. (Melanosarus) xylocopoides Smith (^{Krombein, 1967}) and M. (Melanosarus) nigripennis Spinola (^{Marques and Gaglianone, 2013}). For the subgenus Austromegachile, there is data provided by ^{Morato (2003)} and notes on a wild nest found in an Araucaria forest by ^{Laroca et al. (1987)} in Brazil. Regarding the Pseudocentron subgenus, there is data from ^{Torretta et al. (2012)} for M. (Pseudocentron) gomphrenoides Vachal, ^{Laroca (1991)} for M. (Pseudocentron) apicipennis Schrottky.

In this study, occupancy rate differed between sampled areas, where SA showed low nest abundance compared with APABM, even though SA was sampled for a considerably longer period of time. This may be explained by the fact that SA is under more anthropic pressure and thus can only harbor smaller populations of the studied species. Occupancy rate of nests in APABM was higher than previously recorded in a similar study (^{Morato, 2003}), but among APABM areas, VM was the area where we found the highest number of nests. This area is considered the best preserved area we sampled at APABM and the higher number of nests may be linked to the availability of resources for nest building for Megachile species, such as suitable substrates and material to build nests, besides pollen and nectar for feeding larvae (^{Michener, 1953}; ^{Stefan-Dewenter and Schielle, 2008}; ^{Sheffield et al., 2011}). Other researchers have previously found a positive correlation with the presence of nesting bees and the vegetation cover (DBM, unpublished data), which might be linked with presence of resources bees use for nesting.

Regarding the places chosen for nesting, it is important to point out that M. sejuncta was collected only in the more preserved areas in this study (primary forest in SA and the municipality of Vitória do Mearim in APABM), suggesting that this species probably has little tolerance when it comes to disturbed areas, which was also observed for some Austromegachile species (^{Laroca et al., 1987}; ^{Morato, 2003}). For M. brasiliensis and M. stilbonotaspis, we could not find any pattern in the nesting places used. Few studies address factors that influence nesting activities of solitary bees in the Amazon forest biome. ^{Morato (2003)} and ^{Morato et al. (1999)} observed no correlation of rainfall with nesting activities of Megachile and Centris Fabricius, respectively, in the Amazonian region. In this study, only the nesting activity of M. stilbonotaspis had its peak in months with low rainfall records, and although not detected by the Fisher's test, M. brasiliensis showed the same trend. The lack of patterns recognizable in SA may be due to low abundance of nests, despite intense sampling efforts. Solitary bees' activity pattern seems to vary among and within genera throughout the tropics. ^{Becker et al. (1991)} found a positive correlation between rainfall and flowering seasons and activity peaks for Euglossa Latreille species, while studies focused on other taxa failed to show this association (^{Morato, 2003}; ^{Sabino and Antonini, 2017}).

Another factor that may determine nesting activity is pollen availability, as observed by ^{Cardoso and Silveira (2012)}. For M. brasiliensis and M. sejuncta, the pattern of nesting activity may be correlated with the peak of Attalea sp. bloom, since this was the predominant floral resource used by these two Megachile species. ^{Bernini et al. (2007)} noted peaks of A. speciosa bloom during dry and wet season in Roraima (north of Brazil, deep into the Amazonian region) and this peak variation may be related to pollinator visits during the year and probably the quantity of fruits produced. However, there was some variation in pollen use in SA by M. sejuncta, which also used Myrtaceae and an unidentified pollen type, what may be related to nesting activity during low blooming of Attalea sp. or competition for food with M. brasiliensis. This flexibility of pollen use was observed in other Megachilidae, such as Megachile brevis Say (^{Michener, 1953}), Osmia Panzer (^{Williams, 2003}) and species of the genus Chelostoma Latreille (^{Sedivy et al., 2008}). Nesting activity of the species we studied matched with the flowering season of Attalea in Maranhão and Mato Grosso states (Anderson et al., 1988; ^{Fava et al., 2011}; ^{Pinheiro, 2011}). On the other hand, M. stilbonotaspis was found predominantly using Asteraceae pollen in SA; in APABM, the resource used was mainly Attalea sp., what may reflect pollen use according with one or other available source, and thus explain the bigger number of nests of this species in APABM, since Attalea stands are abundant in the area.

It is important to point out that previous studies on Attalea (Babassu palm trees) pollinators have never recorded any Megachile species, only social bees such as Trigona spinipes Fabricius and Coleoptera species (^{Anderson et al., 1988}; ^{Fava et al., 2011}; ^{Pinheiro, 2011}). Furthermore, only one Arecaceae species, Sabal palmetto (Walt.) Lodd, was recorded for Megachile (^{Raw, 2007}). There are very few studies focused on pollen use by solitary bees in the tropics (^{Cardoso and Silveira, 2012}), and very little is known about the pattern of pollen use in brood cells for those species. Attalea species are widespread in several parts of the Neotropics (^{May et al., 1985}) and their distribution may play a role in explaining some Megachile species' distribution, since Arecaceae seems to be an important food resource.

Oligolecy was observed for M. brasiliensis. ^{Buschini et al. (2009)} cite Ludwigia as a pollen source for M. brasiliensis based on observations from ^{Schlindwein (2004)}, which contradicts our results from the analyses of mounted brood cell slides. To our knowledge, only one study of Ludwigia sericea reports M. brasiliensis visiting its flowers sporadically in the state of São Paulo (^{Sazima and Santos, 1982}). However, flower visitation does not guarantee its effective pollination, as only direct contact with reproductive parts of the plant will ensure pollen collection. Indirect evidence of pollination (^{Michener, 1953}; ^{Schlindwein, 2004}) is also possible by analyzing the content of brood cells, as provided in this study. M. sejuncta and M. stilbonotaspis may be classified as strictly polylectic (^{Michener, 2007}) due to usage of a several unrelated pollen types, considering data from both studied areas. No pollen data exists for any species of the Austromegachile subgenera; regarding for Pseudocentron subgenus, ^{Torretta et al. (2012)} report the Asteraceae along with other families as pollen hosts for another co-subgeneric species, Megachile gomphrenoides Vachal in Argentina. The difference of nest abundance between the two areas may be explained by the pollen resources available, as M. stilbonotaspis used a wider range of pollen types and was highly abundant in APABM.

The voltinism pattern found for Megachile species, considering nesting period and developmental time span, was of at least two yearly generations. However, for M. sejuncta in SA we observed a two delayed emergence pattern in part of the population. From the data we collected, those specimens emerged around the same time as their nests were built in the following season. However, pollen provisions were different for each observed period (^{Supplementary material}). M. sejuncta population in SA had a strongly biased sex ratio toward males and three patterns of emergence: direct development, delayed emergence with individuals emerging in April/May from the past April/May and individuals emerging in August from the past August. The population from APABM was multivoltine, had a marginally statistically significant, biased sex ratio toward females and the pollen used remained constant all season. This pollen use stability may be linked with the nesting activity rate and biased sex ratio toward females.

Although the data gathered do not allow us to make general assumptions in SA, we believe two factors may be influencing the emergence patterns observed in M. sejuncta: females may be spawning more males to increase their fitness (sons do not require a father and are cheaper than daughters, so the genetic load of the offspring comes only from the mother; this theory is explored by ^{Buschini and Bergamaschi, 2010}), since those males will emerge only in the subsequent season, and mate with the females emerging by that time. However, we did not find females in the same proportion as males, and we could speculate females are being produced elsewhere, an opposite situation to the one observed by ^{Buschini and Bergamaschi (2010)} for Trypoxylom opacum Brèthes. The offspring from nests collected in April, and emerged in the subsequent period of April/May, benefit from pollen of Type I sp. and Myrcia when Babassu palms were in low bloom, which could be explained by the difference in pollen load observed in the slides. The pollen grains used to provision those nests may not be able to sustain female development due to its quality, which may explain the male-biased sex ratio, a pattern observed by ^{Brockmann and Grafen (1992)} for Trypoxylon politum Drury.

However, in light of our data for the Megachile species studied here, the topic remains speculative, because different patterns of emergence are known for other Hymenoptera groups, such as Trypoxylon (^{Brockmann, 2004}; ^{Brockmann and Grafen, 1992}; ^{Buschini and Bergamaschi, 2010}), and the authors of those studies enumerated several factors that might influence sex ratio and emergence time, such as latitudinal variation, and environmental cues such as prey availability (^{Brockmann, 2004}). Many emergence patterns, as well as varied sex ratios, were found for megachilid bees in different biomes of the Neotropical region (^{Cardoso and Silveira, 2003}; ^{Marques and Gaglianone, 2013}; ^{Rocha-Filho and Garófalo, 2016}; ^{Sabino and Antonini, 2011}; ^{Zilikens and Steiner, 2004}), in compliance with this study. The sex ratios and emergence patterns may vary under ecological factors such as nest width, parasites and competition for food (^{Rocha-Filho and Garófalo, 2016}; ^{Tepedino and Torchio, 1989}), but we did not record high parasitism rates and males were reared from wide cavity nests as well; it is worth noting that genetic factors may also play a role here (^{Buschini and Bergamaschi, 2010}).

The nest architecture of the studied species fits the Megachile pattern (^{Krombein, 1967}; ^{Michener, 2007}). All three species studied here used round pieces of leaves to build the nest and brood cells opercula; furthermore, the bottom of the nests was made of elongated pieces of leaves with a bent posterior side, forming the cup of the nest. These features were already pointed out by ^{Gonzalez (2008)} as being phylogenetically informative for Megachile lineages.

M. brasiliensis nests were similar to M. nigripennis' (^{Marques and Gaglianone, 2013}), including the behavior of building more than one cell series in wider traps, which shows the independence of confining walls that align their brood cells. They were also similar in shape and arrangement of leaf pieces to nests of M. xylocopoides, described by ^{Krombein (1967)}. The remarkable characteristic of these nests was the tip of the leaf fragments that faced the nest plug with a distinct upward tip, giving the nest a scaly aspect, and the nest opercula with rounded leaf fragments glued to the trap's floor (this study and ^{Marques and Gaglianone, 2013}), although according to the description and images provided by ^{Krombein (1967)}, we cannot state the same for M. xylocopoides.

We observed some variation in the structure of leaves used to build nests of M. stilbonotaspis in two sites in SA (secondary forest and restinga), i.e. the use of leaf fragments and entire leaflets. These two areas are 1.5 km apart from each other and the use of leaf fragments and leaflets may reflect the scarcity of resources used for nest building in the area (^{Cardoso and Silveira, 2012}) combined with a small foraging range (^{Gathmann and Tscharntke, 2002}). ^{Cardoso and Silveira (2012)} also observed similar variation in nest building, but with smaller distances (about 40 m). ^{Gathmann and Tscharntke (2002)} evaluated the foraging range of solitary bees and observed that on plant-species-poor sites the effort for finding suitable nest material seems to be greater than on plant-species-rich sites, although other factors may influence the foraging range, such as bees body size (^{Greenleaf et al., 2007}; ^{Zurbuchen et al., 2010}).

Females of M. sejuncta also showed variation between nests regarding the type of leaves used to build nests in the primary forest from SA, indicating that several plant species were used. However, we noted that each type of leaf fragments was put together in different parts of the nest, evidencing that the bees targeted and collected specific types of materials while those were available to them. This leaf collecting pattern was also observed for M. brevis by ^{Michener (1953)} who stated that such behavior enables a rapid gathering of plant material. We also observed nests of M. sejuncta with a nest building pattern similar to another co-subgeneric species, Megachile habilis Mitchell (^{Laroca et al., 1987}). These results agree with the observations from ^{Kambli et al. (2017)}.

Regarding nests of Megachile species in general, nests of M. brasiliensis and M. stilbonotaspis were more similar to each other than to nests of M. sejuncta. According to ^{Gonzalez (2008)}, species placed into Melanosarus and Pseudocentron subgenera belong to the Pseudocentron lineage, and data on nest architecture seem to corroborate his phylogenetic hypothesis. This means that the features of nests described here may be of interest for future phylogenetic research.

Mortality rates observed in this study were not as high as in other studies, where more than 50% of bees did not reach adulthood (^{Marques and Gaglianone, 2013}; ^{Rocha-Filho and Garófalo, 2016}; ^{Torretta and Durante, 2011}). Parasitism by Lepidoptera (Pyralidae) has been recorded in the literature only for M. rotundata Fabricius and Macrozamia concinna Smith (^{Raw, 2007}). Parasitism by Coelioxys is common (^{Cardoso and Silveira, 2012}; ^{Raw, 2007}) and only M. sejuncta was not a host for any Coelioxys species recorded in this study. Nesting in pre-existing cavities in wood may facilitate attacks from parasites such as Chalcidoidea (^{Grissell, 2007}), as recorded for M. sejuncta and M. brasiliensis in this study. Within Chalcidoidea, ^{Raw (2007)} describes Leucospidae and Eurytomidae as parasites of Megachile. All parasites reported here are new records for the studied species.

Besides parasites, other species were associated with Megachile. We observed cavity supersedure of T. rugulosa over M. brasiliensis, although it was unclear if it occurred due to competition for the cavity or if the T. rugulosa female took over the cavity to build its nest. It seems to be a common process among trap-nesting solitary Hymenoptera (^{Krombein, 1967}). Another unidentified Hymenoptera also shared a cavity with a M. brasiliensis whose immatures died before emergence, hindering identification, but as seen from its nest architecture it was probably a Centris species. Nest sharing also appears to be common and cavities may be shared by bees, whether they belong to the same species or not (^{Cardoso and Silveira, 2012}; ^{Krombein, 1967}). However, nest architecture of some species may be incompatible with another bee's nests, which is the case for Euglossini species (^{Zilikens and Steiner, 2004}) due to different material employed to build nests.

Another interspecific association observed was the occurrence of mites on some M. stilbonotaspis specimens that emerged from nests at the restinga in SA. Those mites were attached to the body of the adults and probably in their phoretic life stage. These mites probably did not feed on the provisions, since the bees completed their development, although observations during the bee life cycle must be made to examine the interaction between the immature bee and the mites in the brood cell. Associations like these occur in solitary (^{Cordeiro et al., 2010}; ^{Klimov and O'Connor, 2007}) and social bees (^{Sammataro et al., 2000}), and they may be commensal, mutualistic or predatory/parasitic.

Acknowledgements

We thank doctors Lea Carreira from “Museu Paraense Emílio Goeldi” (MPEG) and Gabriel Melo from “Universidade Federal do Paraná” (UFPR) for identifying pollen slides and bee specimens, respectively; the “Universidade Federal do Maranhão” for the PIBIC scholarship in the first year of project to the first author; the “Fundação de Amparo a Pesquisa e Desenvolvimento Científico do Maranhão”, FAPEMA, Brazil, (grant CBIOMA project 02986/12) and Conselho Nacional de Desenvolvimento Científico e Tecnológico CNPq, Brazil (grant 00878/15) for the financial support; “Quercegen S.A." and “Instituto Formação" for the permission to research in “Sitio Aguahy" and “Peri-Mirim" area., and for the scholarships provided for the first and second authors of this paper. We also thank Clarisse Figueiredo, Victor Nogueira and André Hoffmann for English revision and proofreading, respectively, and Cid Conde for one year of collecting at the “Sítio Aguahy”. Finally, we thank very much the two anonymous reviewers, who helped greatly at improving this paper.

Appendix A Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.rbe.2018.03.002.

References

• ^{Anderson et al., 1988}
  Anderson, A.B., Overal, W.L., Henderson, A., 1988. Pollination ecology of a forest-dominant palm (Orbignya phalerata Mart.) in Northern Brazil. Biotropica 20, 192-205.
• ^{Becker et al., 1991}
  Becker, P., Moure, J.S., Peralta, F.J.A., 1991. More about Euglossine bees in Amazonian forest fragments. Biotropica 23, 586-591.
• ^{Bernini et al., 2007}
  Bernini, H., Moret, A., Almeida, R., Bernardi, J.V.E., 2007. Levantamento fitossociológico da (Orbignya speciosa) a partir de técnicas multivariadas para a produção sustentável em RESEX. In: VIII Congresso de Ecologia do Brasil, Caxambu, MG, Brasil, pp. 1-2.
• ^{Brockmann, 2004}
  Brockmann, H.J., 2004. Variable life-history and emergence patterns of the pipe-organ mud-daubing wasp, Trypoxylon politum (Hymenoptera: Sphecidae). J. Kans. Entomol. Soc. 77, 503-527.
• ^{Brockmann and Grafen, 1992}
  Brockmann, H.J., Grafen, A., 1992. Sex ratios and life-history patterns of a solitary wasp, Trypoxylon (Trypargilum) politum (Hymenoptera: Sphecidae). Behav. Ecol. Sociobiol. 30, 7-27.
• ^{Buschini and Bergamaschi, 2010}
  Buschini, M.L.T., Bergamaschi, A.C.B., 2010. Strongly female-biased sex allocation in a trivoltine population of Trypoxylon (Trypargilum) opacum Brèthes (Hymenoptera, Crabronidae). Acta Zool. (Stockh.) 91, 433-439.
• ^{Buschini et al., 2009}
  Buschini, M.L.T., Rigon, J., Cordeiro, J., 2009. Plants used by Megachile (Moureapis) sp. (Hymenoptera: Megachilidae) in the provisioning of their nests. Braz. J. Biol. 69, 1187-1194.
• ^{Cardoso and Silveira, 2003}
  Cardoso, F.C., Silveira, F.A., 2003. Ecologia de nidificação de espécies de Megachile (Hymenoptera: Apoidea) em ninhos armadilha na Estação Ambiental de Peti (MG). In: VI Congresso de Ecologia do Brasil, Fortaleza, CE, Brasil, pp. 69-71.
• ^{Cardoso and Silveira, 2012}
  Cardoso, C.F., Silveira, F.A., 2012. Nesting biology of two species of Megachile (Moureapis) (Hymenoptera: Megachilidae) in a semideciduous forest reserve in southeastern Brazil. Apidologie 43, 71-81.
• ^{Cordeiro et al., 2010}
  Cordeiro, G.D., Taniguchi, M., Flechtmann, C.H.W., Alves-dos-Santos, I., 2010. Phoretic mites (Acari: Chaetodactylidae) associated with the solitary bee Tetrapedia diversipes (Apidae: Tetrapediini). Apidologie 42, 128-139.
• ^{Eickwort et al., 1981}
  Eickwort, G.C., Matthews, R.W., Carpenter, J., 1981. Observation on the nesting behaviour of Megachile rubi and M texana with a discussion of the significance of soil nesting in the evolution of megachilid bees (Hymenoptera: Megachilidae). J. Kans. Entomol. Soc. 54, 557-570.
• ^{Erdtman, 1943}
  Erdtman, G., 1943. An Introduction to Pollen Analysis. An Introduction to Pollen Analysis. Waltham, MA, USA.
• ^{Fava et al., 2011}
  Fava, W.V., Covre, W.S., Sigrist, M.R., 2011. Attalea phalerata and Bractis glaucescens (Arecaceae, Arecoideae): phenology and pollination ecology in the Pantanal, Brazil. Flora 206, 575-584.
• ^{Garófalo et al., 2004}
  Garófalo, C.A., Martins, C.F., Alves-dos-Santos, I., 2004. The Brazilian solitary bee species caught in trap nests. In: Freitas, B.M., Pereira, J.O.P. (Eds.), Solitary Bees: Conservation, Rearing and Management for Pollination. Imprensa Universitária, Ceará, pp. 77-84.
• ^{Gathmann and Tscharntke, 2002}
  Gathmann, A., Tscharntke, T., 2002. Foraging range of solitary bees. J. Anim. Ecol. 71, 757-764.
• ^{Gonzalez, 2008}
  Gonzalez, V.H., 2008. Phylogeny and Classification of the Bee tribe Megachilini (Hymenoptera: Apoidea: Megachilidae), with Emphasis on the Genus Megachile (Ph.D. thesis). Department of Ecology and Evolutionary Biology and the College of Liberal Arts and Science of the University of Kansas, 274 pp.
• ^{Greenleaf et al., 2007}
  Greenleaf, S.S., Williams, N.M., Winfree, R., Kremen, C., 2007. Bee foraging ranges and their relationship to body size. Oecologia 153, 589-596.
• ^{Grissell, 2007}
  Grissell, E.E., 2007. Torymidae (Hymenoptera: Chalcidoidea) associated with bees (Apoidea), with a list of Chalcidoid bee parasitoids. J. Hymenopt. Res. 16, 234-265.
• ^{Kambli et al., 2017}
  Kambli, S.S., Aiswarya, M.S., Manoj, K., Varma, S., Asha, G., Rajesh, T.P., Sinu, P.A., 2017. Leaf foraging sources of leafcutter bees in a tropical environment: implications for conservation. Apidologie , 1-10.
• ^{Klimov and O'Connor, 2007}
  Klimov, P.B., O'Connor, B.M., 2007. Ancestral area analysis of chaetodactylid mites (Acari: Chaetodactylidae), with description of new early derivative genus and six new species from the Neotropics. Ann. Entomol. Soc. Am. 100, 810-829.
• ^{Krombein, 1967}
  Krombein, K.V., 1967. Trap-Nesting Wasps and Bees Life Histories, Nest and Associates. Smithsonian Press, Washington.
• ^{Laroca, 1991}
  Laroca, S., 1991. Euglossa stellfeldi: Arquitetura do ninho e coexistência com Pseudocentron apicipennis em uma mesma cavidade (Hymenoptera, Apoidea). Acta Biol. Parana. 20, 103-108.
• ^{Laroca et al., 1992}
  Laroca, S., Corbella, E., Varela, G., 1992. Biologia de Dactylomegachile affabilis (Hymenoptera, Apoidea): 1. Descrição do ninho. Acta Biol. Parana. 21, 23-29.
• ^{Laroca et al., 1987}
  Laroca, S., Schwartz-Filho, D.L., Zanella, F.C., 1987. Ninho de Austromegachile habilis e notas sobre a diversidade de Megachile (Apoidea, Megachilidae) em biótopos neotropicais. Acta Biol. Parana. 16, 93-105.
• ^{Lieux, 1980}
  Lieux, M.H., 1980. Acetolysis applied to microscopical honey analysis. Grana 19, 57-61.
• ^{Marques and Gaglianone, 2013}
  Marques, M.F., Gaglianone, M.C., 2013. Biologia de nidificação e variação altitudinal na abundância de Megachile (Melanosarus) nigripennis Spinola (Hymenoptera, Megachilidae) em um inselberg na Mata Atlântica, Rio de Janeiro. Biosci. J. 29, 198-208.
• ^{May et al., 1985}
  May, P.H., Anderson, A.B., Frazão, J.M.F., Balick, M.J., 1985. Babassu palm in the agroforestry systems in Brazil's Mid-North region. Agrofor. Syst. 3, 275-295.
• ^{Michener, 1953}
  Michener, C.D., 1953. The biology of a leafcutter bee (Megachile brevis) and its associates. Univ. Kans. Sci. Bull. 35, 1659-1748.
• ^{Michener, 2007}
  Michener, C.D., 2007. The Bees of the World, 2nd ed. John Hopkins University Press, Baltimore.
• ^{Morato et al., 1999}
  Morato, E.F., Garcia, M.V.B., Campos, L.A.O., 1999. Biologia de Centris Fabricius (Hymenoptera, Anthophoridae, Centridini) em matas contínuas e fragmentadas na Amazônia Central. Rev. Bras. Zool. 16, 1213-1222.
• ^{Morato, 2003}
  Morato, E.F., 2003. Biologia de Megachile (Austromegachile) orbiculata Mitchell (Hymenoptera, Megachilidae) em matas contínuas e fragmentos na Amazônia Central. In: Melo, A.R., Alves-dos-Santos, I. (Eds.), Apoidea neotropica: homenagem aos 90 anos de Jesus Santiago Moure. UNESC, Criciúma, pp. 157-162.
• ^{Orr et al., 2015}
  Orr, M.C., Portman, Z.M., Griswold, T., 2015. Megachile (Megachile) montivaga (Hymenoptera: Megachilidae) nesting in live thistle (Asteraceae: Cirsium). J. Melittol. 48, 1-6.
• ^{Peel et al., 2007}
  Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen-Geiger climate classification. Hydrol. Earth Syst. Sci. Discuss. 4, 439-473.
• ^{Pinheiro et al., 2010}
  Pinheiro, C.U.B., Araujo, N.A., Arouche, G.C., 2010. Plantas úteis do Maranhão: Região da Baixada Maranhense. Gráfica e Editora Aquarela, São Luís.
• ^{Pinheiro, 2011}
  Pinheiro, C.U.B., 2011. Palmeiras do Maranhão: onde canta o sabiá. Gráfica e Editora Aquarela, São Luís.
• ^{Pinheiro-Machado et al., 2002}
  Pinheiro-Machado, C., Alves-dos-Santos, I., Imperatriz-Fonseca, V.L., Kleinert, A.D.M.P., Silveira, F.A., 2002. Brazilian bee surveys: state of knowledge, conservation and sustainable use. In: Kevan, P., Imperatriz Fonseca, V.L. (Eds.), Pollinating Bees - The Conservation Link Between Agriculture and Nature. Ministério do Meio Ambiente, Brasília, pp. 115-129.
• ^{Pitts-Singer and Cane, 2011}
  Pitts-Singer, T.L., Cane, J.H., 2011. The Alfalfa Leafcutting Bee, Megachile rotundata: The World's Most Intensively Managed Solitary Bee. Annu. Rev. Entomol. 56, 221-237.
• ^{R Development Core Team, 2015}
  R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria https://www.R-project.org
  » https://www.R-project.org
• ^{Raw, 2007}
  Raw, A., 2007. An annotated catalogue of the leafcutter and mason bees (genus Megachile) of the Neotropics. Zootaxa 1601, 1-127.
• ^{Rocha-Filho and Garófalo, 2016}
  Rocha-Filho, L., Garófalo, C.A., 2016. Nesting biology of Megachile (Chrysosarus) guaranitica and high mortality caused by its cleptoparasite Coelioxys bertonii (Hymenoptera: Megachilidae) in Brazil. Aust. Entomol. 55, 25-31.
• ^{Sabino and Antonini, 2011}
  Sabino, W.O., Antonini, Y., 2011. Use of trap-nests with a Neotropical leaf-cutter bee Megachile (Moureapis) anthidioides (Hymenoptera: Megachilidae). J. Kans. Entomol. Soc. 84, 78-79.
• ^{Sabino and Antonini, 2017}
  Sabino, W.O., Antonini, Y., 2017. Nest architecture, life cycle, and natural enemies of the Neotropical leafcutting bee Megachile (Moureapis) maculata (Hymenoptera: Megachilidae) in a montane forest. Apidologie 48, 450-460.
• ^{Sammataro et al., 2000}
  Sammataro, D., Gerson, U., Needham, G., 2000. Parasitic mites of honey bees: life history, implications, and impact. Annu. Rev. Entomol. 45, 519-548.
• ^{Sazima and Santos, 1982}
  Sazima, M., Santos, J.U.M., 1982. Biologia floral e insetos visitantes de Ludwigia sericea (Onagraceae). Boletim do Museu Paraense Emílio Goeldi 54, 1-10.
• ^{Schlindwein, 2004}
  Schlindwein, C., 2004. Are oligolectic bees always the most effective pollinators? In: Freitas, B.M., Pereira, J.O.P. (Eds.), Solitary Bees: Conservation, Rearing and Management for Pollination. Imprensa Universitária, Ceará, pp. 231-240.
• ^{Sedivy et al., 2008}
  Sedivy, C., Praz, C., Dorn, S., 2008. Patterns of host-plant choice in bees of the genus Chelostoma: the constraint hypothesis of host-range evolution in bees. Evolution 62, 2487-2507.
• ^{Serra et al., 2016}
  Serra, F.C.V., Lima, P.B., Almeida, E.B., 2016. Species richness in restinga vegetation on the eastern Maranhão State, Northeastern Brazil. Acta Amaz., .
  » https://doi.org/10.1590/1809-4392201504704
• ^{Sheffield, 2017}
  Sheffield, C.S., 2017. Unusual nesting behavior in Megachile (Eutricharaea) rotundata (Hymenoptera: Megachilidae). J. Melittol. 69, 1-6.
• ^{Sheffield et al., 2011}
  Sheffield, C.S., Ratti, C., Packer, L., Griswold, T., 2011. Leafcutter and Mason Bees of the genus Megachile Latreille (Hymenoptera: Megachilidae) in Canada and Alaska. Can. J. Arthropod Identif. 18, 1-107.
• ^{Stefan-Dewenter and Schielle, 2008}
  Stefan-Dewenter, I., Schielle, S., 2008. Do resources or natural enemies drive bee population dynamics in fragmented habitats?. Ecology 89, 1375-1387.
• ^{Strickler et al., 1996}
  Strickler, K., Scott, V.L., Fischer, R.L., 1996. Comparative nesting ecology of two sympatric leafcutting bees that differ in body size (Hymenoptera: Megachilidae). J. Kans. Entomol. Soc. 69, 26-44.
• ^{Tepedino and Torchio, 1989}
  Tepedino, V.J., Torchio, P.F., 1989. Influence of nest hole selection on sex ratio and progeny size in Osmia lignaria propinqua (Hymenoptera: Megachilidae). Ann. Entomol. Soc. Am. 82, 355-360.
• ^{Torretta and Durante, 2011}
  Torretta, J.P., Durante, S.P., 2011. Nesting ecology of Megachile (Sayapis) mendozana Cockerell and its synonymy with Megachile (Sayapis) santiaguensis Durante (Hymenoptera: Megachilidae). Zootaxa 3008, 63-68.
• ^{Torretta et al., 2012}
  Torretta, J.P., Durante, S.P., Colombo, M.G., Basilio, A.M., 2012. Nesting biology of the leafcutter bee Megachile (Pseudocentron) gomphrenoides (Hymenoptera: Megachilidae) in an agro-ecosystem. Apidologie 43, 624-633.
• ^{Torretta et al., 2014}
  Torretta, J.P., Durante, S.P., Basilio, A.M., 2014. Nesting ecology of Megachile (Chrysosarus) catamarcensis Schrottky (Hymenoptera: Megachilidae), a Prosopis-specialist bee. J. Apic. Res. 53, 590-598.
• ^{Williams, 2003}
  Williams, N.M., 2003. Use of novel pollen species by specialist and generalist solitary bee (Hymenoptera: Megachilidae). Oecologia 134, 228-237.
• ^{Zilikens and Steiner, 2004}
  Zilikens, A., Steiner, J., 2004. Nest architecture, life cycle and cleptoparasite of the Neotropical leaf-cutting bee Megachile (Chrysosarus) pseudanthidioides Moure (Hymenoptera: Megachilidae). J. Kans. Entomol. Soc. 77, 193-202.
• ^{Zurbuchen et al., 2010}
  Zurbuchen, A., Landert, L., Klaiber, J., Müller, A., Hein, S., Dorn, S., 2010. Maximum foraging ranges in solitary bees: only few individuals have the capability to cover long foraging distances. Biol. Conserv. 143, 669-676.

Publication Dates

• Publication in this collection
  Apr-Jun 2018

History

• Received
  18 Nov 2017
• Accepted
  8 Mar 2018