Nesting biology of Centris (Paracentris) burgdorfi (Apidae: Centridini)

Item Type Article

Authors Sabino, William O.; Alves-dos-Santos, Isabel; Queiroz, Elisa Pereira; de Faria, Letícia Biral; Papaj, Daniel R.; Buchmann, Stephen L.; da Silva, Cláudia Inês

Citation William O. Sabino, Isabel Alves-dos-Santos, Elisa Pereira Queiroz, Letícia Biral de Faria, Daniel R. Papaj, Stephen L. Buchmann & Cláudia Inês da Silva (2020) Nesting biology of Centris (Paracentris) burgdorfi (Apidae: Centridini), Journal of Apicultural Research, DOI: 10.1080/00218839.2020.1717760

DOI 10.1080/00218839.2020.1717760

Publisher TAYLOR & FRANCIS LTD

Journal JOURNAL OF APICULTURAL RESEARCH

Rights © 2020 International Bee Research Association.

Download date 03/10/2021 23:39:10

Item License http://rightsstatements.org/vocab/InC/1.0/

Version Final accepted manuscript

Link to Item http://hdl.handle.net/10150/637738 0DQXVFULSWZLWKDXWKRUGHWDLOV

1 Nesting biology of Centris (Paracentris ) burgdorfi (Apidae: Centridini) 1 2 2 3 4 5 3 William O. Sabino 1,2*, Isabel Alves-dos-Santos 1, Elisa Pereira Queiroz 1, Letícia 6 7 4 Biral de Faria 1, Daniel R. Papaj 3, Stephen L. Buchmann 3,4 and Cláudia Inês da Silva¹ 8 9 10 5 11 12 6 1. Departamento de Ecologia, Instituto de Biociências, Universidade de São 13 14 15 7 Paulo, São Paulo, SP, 05508-900, Brazil; 2. Coordenação de Zoologia, Museu Paraense 16 17 8 Emílio Goeldi, Belém, PA, 66077-530, Brazil; 3. Department of Ecology and 18 19 20 9 Evolutionary Biology, University of Arizona, Tucson, AZ, 85721, USA; 4. Department 21 22 10 of Entomology, University of Arizona, Tucson, AZ, 85721, USA. 23 24 11 25 26 27 12 * Corresponding author: [email protected] ; ORCID iD: 0000-0003-2151- 28 29 13 0723 30 31 32 14 33 34 15 Running title: Nesting biology of Centris burgdorfi 35 36 37 16 38 39 17 40 41 18 42 43 44 19 45 46 20 47 48 49 21 50 51 22 52 53 54 23 55 56 24 57 58 59 25 60 1 61 62 63 64 65 26 Abstract: 1 2 27 3 28 We describe the nesting biology of Centris (Paracentris ) burgdorfi , a solitary bee that nests in sandstone in northeastern 4 5 29 Brazil. The nest consists of a shallow tunnel with access to the brood cells. Females of C. burgdorfi made 1 to 7 brood 6 7 30 cells per nest with each cell requiring 2.58 ± 0.40 (X ± SD) days to construct. The average cell-building construction 8 9 31 time was longer when compared to other Centris species. Females were larger than males, and this difference was 10 11 32 reflected in the size of their respective emergence cells. The temperature within C. burgdorfi nests was lower when 12 13 33 compared to ambient temperature. Our study is the first to report the nesting biology of C. burgdorfi . The detailed 14 34 behavior of the female inside the nest was also described, which is unusual in the study of solitary bee nesting biology. 15 16 35 17 18 36 Keywords: foraging behavior; nesting behavior; oil-collecting bees; sex ratios; solitary bees 19 20 37 21 22 23 38 Introduction 24 25 39 Many bees within the tribe Centridini are known as oil-collecting bees, because females 26 27 gather and use floral oils as resources in brood cell construction (hardened and 28 40 29 30 41 hydrophobic cell linings) and/or high caloric and nutritious food for their larvae (Vogel, 31 32 42 1974; Buchmann, 1987). The genus Centris Fabricius comprises about 250 species 33 34 35 43 distributed from Argentina to Mexico and the southwestern United States. The subgenus 36 37 44 Centris (Paracentris ) Cameron is more common in semi-arid regions, including the 38 39 40 45 Sonoran Desert of northern Mexico and southwestern United States (Alcock, Jones, & 41 42 46 Buchmann, 1976; Rozen & Buchmann, 1990), Andean regions and the semiarid region 43 44 45 47 of northeastern Brazil (Zanella, 2002; Vivallo & Zanella, 2012). 46 47 48 All species of oil-collecting bees are solitary, although females of some species 48 49 49 form dense nesting aggregations. Centridine bees are important pollinators of wild and 50 51 52 50 cultivated species (Freitas & Paxton, 1998; Oliveira & Schlindwein, 2009) and some plants 53 54 51 (e.g. Malpighia , Byrsonima ) depend exclusively on the pollination services of Centris 55 56 57 52 bees because of their specialized oil-foraging behavior (Aguiar, 2003). 58 59 60 2 61 62 63 64 65 53 In Brazil, the best-studied Centris species are those that nest in vacant beetle 1 2 54 burrows in dead trees, pre-existing cavities, and especially in the drilled holes of wooden 3 4 5 55 trap nests, such as C. analis Fabricius and C. tarsata Smith (Garófalo, Camillo, & 6 7 56 Serrano, 1989). Studies of ground-nesting species are less common (Aguiar & 8 9 10 57 Gaglianone, 2003). In Paracentris there are studies of only 6 other species (out of 59): C. 11 12 58 nigerrima Spinola (Janvier, 1926) in Chile, C. autrani Vachal (probably C. neffi Moure) 13 14 15 59 in Bolivia (Janvier, 1955), C. rhodopus Cockerell, C. cockerelli Cockerell, C. 16 17 60 caesalpiniae Cockerell and C. pallida Fox (Rozen & Buchmann, 1990) conducted in 18 19 20 61 Arizona, USA. In Brazil we found a nesting aggregation of Centris (Paracentris ) 21 22 62 burgdorfi Friese, in northeastern Natal, in the state of Rio Grande do Norte (RN), Brazil. 23 24 63 This species is widely distributed, occurring in other states within the northeastern region 25 26 27 64 and central and south Brazil (Silveira, Melo, & Almeida, 2002; Zanella, 2002). 28 29 65 This study investigates the nesting biology of C. burgdorfi in a stabilized 30 31 32 66 sandstone dune (“fossilized dunes”) region of coastal Brazil. To our knowledge, there is 33 34 67 no previously published information about the nesting behavior of C. burgdorfi . The 35 36 37 68 mating biology and the plasticity of the trophic niche of this bee have been published 38 39 69 elsewhere (Sabino, Silva, & Alves-dos-Santos, 2017; Sabino, Alves-dos-Santos, & Silva, 40 41 70 2018). Here, we evaluate the activity periods and construction of nests of the species. We 42 43 44 71 sought to answer the following questions: 1) How many nests does a female provision in 45 46 72 her lifetime, and how many brood cells do these nests contain? 2) How much time do 47 48 49 73 females spend on nest and cell construction? 3) How long are foraging trips for pollen 50 51 74 and oil? 4) How long do females spend provisioning their nests? 5) What are the seasonal 52 53 54 75 emergence and activity periods of the adult bees? 6) Is C. burgdorfi protandric with males 55 56 76 emerging first? 7) What insect parasites are associated with this species? 8) Do the cells 57 58 59 77 in which males versus females develop differ in size and shape? 60 3 61 62 63 64 65 78 1 2 79 Materials and methods 3 4 5 6 7 80 Centris burgdorfi nesting behavior 8 9 81 We located a nesting aggregation of C. burgdorfi within a coastal dunes region 10 11 82 close to the city of Natal, RN, in the northeastern of Brazil (05°36.310’S, 35°14.435’W; 12 13 14 83 elevation: 41m). The average annual temperature is 26.7 °C and the rainy season extends 15 16 84 from March to August, with an average annual rainfall of 1643.48 mm (INMET / UFRN, 17 18 19 85 2002). 20 21 86 The nesting season of C. burgdorfi is between April and July. The behavior of 22 23 24 87 foraging and nesting females was observed for two successive breeding seasons, during 25 26 88 19 consecutive days of observation in 2014 (April 21 th to May 9 th - 247 hours of 27 28 89 observations) and 17 consecutive days of observation in 2015 (May 25 th to June 10 th - 29 30 31 90 208 hours of observations). The fieldwork was conducted daily from 04:30 h to 17:30 h. 32 33 91 In this location at this time of year, the sun rises at 05:15 h and sets at 17:30 h. 34 35 36 92 Centris burgdorfi females always built nests in a formation of sandstone dunes. 37 38 93 We did not find them nesting anywhere else. The dunes exhibit sharp discontinuities in 39 40 41 94 the ground surface, forming slopes reaching 50 meters high (Neto, Costa, Severo, Júnior, 42 43 95 & Scudelari, 2005) dating to between the upper Pleistocene and the Holocene (Nogueira, 44 45 96 1981). 46 47 48 49 50 97 Nest architecture and adult emergence 51 52 98 Measurements were made of nest entrances and lateral tunnels leading to brood 53 54 99 cells. Some nests (n = 20) were excavated completely, and the cells were taken to the 55 56 57 100 laboratory in April 2011, February 2012 and April 2014 for analyses. Brood cells were 58 59 101 separated into individual vials and maintained in an incubator (FANEM-347 CDG) at the 60 4 61 62 63 64 65 102 same internal temperature (about 28 °C) as found within outdoor nests. After all bees had 1 2 103 emerged, their sex ratios and parasitism rates were determined. To test for any correlation 3 4 5 104 between the sexes and the sizes of their brood cells, intertegular distance for 25 males and 6 7 105 25 females was measured (Cane, 1987). We measured the height of the cells and the cell 8 9 10 106 diameters at their widest point (base) and opening for 50 cells of each sex. All 11 12 107 measurements were performed with a digital caliper (Digimess- 100.174BL). 13 14 15 108 To assess the water retardant nature and resistance of the cell wall, we wrapped 16 17 109 10 C. burgdorfi cells completely in well-moistened tissue paper inside a vial for 24h. We 18 19 20 110 then unwrapped the cell and placed a drop of water on the wall inside the cells (n=5) and 21 22 111 used a dissection microscope to determine if the droplet was absorbed or not. 23 24 25 26 112 Nesting activity of females 27 28 113 The nesting behavior of 20 females that initiated nest excavations was monitored 29 30 31 114 throughout the day (12 females in 2014 and 8 females in 2015). To identify individuals, 32 33 115 females were captured and marked on their thoracic dorsa with different color 34 35 36 116 combinations using non-toxic paint pens (Posca®; Tokyo, Japan). We quantified the 37 38 117 number of cells each female constructed, as well as the foraging time and the time 39 40 41 118 required to complete the construction of a cell and of a complete nest. We distinguished 42 43 119 between floral oil and pollen-collecting trips made by females C. burgdorfi through 44 45 120 observations of the material on the female scopae. We did not measure the nectar- 46 47 48 121 collecting trips because this behavior is too difficult to visualize since the female can 49 50 122 transport nectar together with other materials. Thus, it was possible to quantify the 51 52 53 123 average time spent during oil-foraging trips, lining their cells with oil, the average 54 55 124 duration of pollen-collecting trips, and the time females required to close/seal their cells. 56 57 58 59 60 5 61 62 63 64 65 125 Based on these data, the total duration of cell construction times was measured, allowing 1 2 126 comparisons of the time spent on each activity (Martins, Peixoto, & Aguiar, 2014). 3 4 5 127 To monitor female activity, we made a small hole in the lateral part of each nest 6 7 128 (n = 20). The orifice was small enough (about 1.5 cm) to avoid interference in the 8 9 10 129 behavior of the females, and the observation involved a quick inspection, after which the 11 12 130 hole was re-sealed with the same sand substrate. 13 14 15 131 Another five nests were selected to record in detail the construction activity. We 16 17 132 filmed the behavior of the nesting females and later analyzed and timed these behaviors 18 19 20 133 in greater detail. These nests were not used for daily monitoring because, for video 21 22 134 recording, it was necessary to make a larger hole at the top of the nest (about 3 cm in 23 24 135 diameter) and, inevitably, the female abandoned her nest after this larger disturbance was 25 26 27 136 made. 28 29 137 To evaluate the geographical dispersion of nests across the sandstone, we divided 30 31 2 32 138 the total area into 1 m quadrants (following Martins et al., 2014) and all the active nests 33 34 139 within each quadrat were counted. 35 36 37 140 Temperature data both inside and outside the nests and the humidity of the 38 39 141 environment were collected using a standardized digital thermo-hygrometer (Instrutemp- 40 41 142 ITHT 2250). The external and internal nest temperatures at an average cell depth were 42 43 44 143 measured every 30 minutes (from 05:30 h to 17:30 h) for 12 days. Nests for temperature 45 46 144 measurements were selected randomly every day, at different heights in the sandstone 47 48 49 145 dunes (lower positions near the ground, and the average middle height and the top of 50 51 146 aggregation) to avoid a bias in the analyses. Outside ambient temperature was always 52 53 54 147 measured one meter in front of the same nest that internal temperature was measured. 55 56 57 58 59 60 6 61 62 63 64 65 148 We also observed the presence of natural enemies, including cleptoparasitic bees, 1 2 149 a parasitic wasp and fly species and their behaviors in the Centris burgdorfi nesting 3 4 5 150 aggregation. 6 7 8 9 151 Data analysis 10 11 152 Continuous data are generally reported here as means + standard errors. A student 12 13 14 153 t-test was used to evaluate the sizes of male versus female bees as well as possible 15 16 154 morphological differences between male and female-destined cells. 17 18 19 155 A Chi-square test was used to assess if the proportion of males and females that 20 21 156 emerged in the laboratory differed from a standard 1:1 Fisherian sex ratio. 22 23 24 157 A Kruskal-Wallis test with Dunn post hoc test was used to evaluate the three main 25 26 158 activities of females during the construction of a cell (i.e. adding floral oil to form the cell 27 28 159 lining, mixing pollen and sometimes oil, to form larval food (provision) masses, and oil 29 30 31 160 for “operculation ”) and the duration of these behaviors inside the nest. A Mann-Whitney 32 33 161 U test was used to compare the duration of pollen-collection trips with that of oil-foraging 34 35 36 162 trips used for both larval food provisioning and forming the cell wall linings and cell 37 38 163 closures (operculation). We used an ANOVA with Tukey post hoc test to evaluate the 39 40 41 164 mean number of trips for oil collection for cell lining, for pollen collection, and for oil 42 43 165 collection. 44 45 166 A repeated-measures ANOVA, with day as a factor was used, to evaluate 46 47 48 167 differences in the average temperatures of the external environment with temperatures 49 50 168 inside the nests. 51 52 53 169 Statistical analyses were performed using R (R Development Core Team, 2018) 54 55 170 and Statistica 7.0 software, based on Zar (1996) and Sokal & Rohlf (1995). 56 57 58 171 59 60 7 61 62 63 64 65 172 Results 1 2 173 Architecture of the nests and emergence of individuals 3 4 5 174 Female bees built nests exclusively in the sandstone portion of the dunes, only in 6 7 175 the vertical sides; no nests were observed in the loose sand (mobile or modern dunes). 8 9 10 176 Females excavated horizontal tunnels with small circular entrance holes 1.17 cm ± 1.05 11 12 177 (n = 134) diameter leading to a relatively straight horizontal tunnel of 5.60 ± 0.96 cm in 13 14 15 178 length. Immediately after digging the nest burrow the female constructed linearly aligned 16 17 179 cells adjacent to each other (Fig. 1a-b). 18 19 20 180 The emergence period extended from the end of March to the end of July 2011, 21 22 181 and from the end of March to the beginning of August in 2012. Adults emerged from just 23 24 182 62% of all brood cells collected (n = 125), indicating a relatively high level of mortality 25 26 27 183 in the immature stages. The observed male:female sex ratio of 0.825 did not differ 28 29 184 statistically from a 1:1 sex ratio (χ2 = 0.67, p = 0.48; 33 males and 40 females). Centris 30 31 32 185 burgdorfi is highly protandric, since in all completed nests, males in a given nest always 33 34 186 emerged earlier than females in that nest. Protandry at the level of the population was 35 36 37 187 observed in the field, with only males flying in the nesting site at the beginning of the two 38 39 188 breeding seasons. 40 41 189 Females were larger than conspecific males (t= 2.04, p <0.05, d.f.= 48; Fig. 2a), 42 43 44 190 the size difference being reflected in the size of their respective brood cells. Male-destined 45 46 191 cells were 2.12 cm ± 0.13 in height (n = 50) while female-destined cells were 2.19 cm ± 47 48 49 192 0.14 in height (n = 50). This difference was significant (t= 2.79; p <0.05, d.f.= 98; Fig. 50 51 193 2b). The same sexual dimorphism was observed in the size of the brood cell opening (with 52 53 54 194 the inner part of the cell closure with spiral pattern), with male brood cells averaging 0.78 55 56 195 cm ± 0.1 in diameter (n = 50) and female brood cells averaging 0.83 ± 0.09 cm (t= 2.69, 57 58 59 196 p <0.05, d.f.= 98; Fig. 2c). However, there were no differences between the sexes in width 60 8 61 62 63 64 65 197 of the brood cell at its widest point (typically, the cell base) (males = 1.51 ± 0.1 cm; 1 2 198 Females = 1.52 ± 0.14 cm; t= 0.33, p = 0.77, d.f.= 98; Fig. 2d). 3 4 5 6 199 Nesting activity of females 7 8 200 The period of nesting activity in Natal, northeastern Brazil, coincided with the 9 10 11 201 rainy season at this location. Females initiated nest-building immediately following 12 13 202 mating, digging their nest. Females were observed reusing tunnels of old nests, by 14 15 16 203 excavating a short lateral tunnel to the new brood cells. We commonly observed nests 17 18 204 from other years, which had been exposed to the weather and erosion after the immatures 19 20 21 205 had developed and emerged as new adults. This observation indicates that the same 22 23 206 nesting location is being used in multiple years by this species. 24 25 207 The first females were observed at 05:20 h, and the last females at 16:40 h. Despite 26 27 28 208 being active until almost sunset, most construction and provisioning of the nests occurred 29 30 209 in the morning hours, between 7 and 10 hours (Fig. 3). 31 32 33 210 Nests were built close together in aggregations. The density of nests in these 34 35 211 aggregations was estimated to be 9.64 nests/m 2. 36 37 38 212 The excavation of the tunnel entrance and the excavation and shaping of the first 39 40 213 brood cell within a nest required an average of 327.8 ± 79.9 min (n = 20). Excavation was 41 42 43 214 characterized by multiple pauses, during which the female left the nest presumably to 44 45 215 forage for nectar (and possibly water). It was common to see females initiating an 46 47 216 excavation in one place, then soon after, leaving that location to begin another nest 48 49 50 217 somewhere else (n = 17). Two marked females took about two days to find an ideal place 51 52 218 to start their nest excavation. 53 54 55 56 57 58 59 60 9 61 62 63 64 65 219 After digging the main burrow, females began to excavate and shape their brood 1 2 220 cells. The cells were barrel-shaped with the bottoms slightly larger than the upper portions 3 4 5 221 (Fig. 1b). 6 7 222 Building a brood cell proceeded in the following manner: after digging the 8 9 10 223 entrance tunnel, the female shaped the cell to be used for the floral resource provisions 11 12 224 (pollen, nectar and oil). Females made 9.62 ± 1.81 trips (n = 91) to obtain floral oils from 13 14 15 225 Krameria tomentosa St. Hill (Krameriaceae) to line the cell walls. Upon returning from 16 17 226 a foraging trip, the female always inspected the interior of the cell, then turned around 18 19 20 227 and entered the cell from the front, using her hind legs to attach the sand grains to the top 21 22 228 of the cell along with the collected floral oil. Females then went back into the cell and 23 24 229 initiated a rotating movement to deposit the mixture of oil + sand (with brushes) evenly 25 26 27 230 over the cell wall. 28 29 231 After lining the brood cell, the female began depositing her scopal pollen load, 30 31 32 232 making an average of 4.91 ± 1.33 (n = 91) pollen-collection trips to provision one cell. 33 34 233 Pollen deposition involved the female re-entering the cell and making rotating 35 36 37 234 movements while discharging her scopal loads (Fig. 4a-d). We observed that the cell 38 39 235 linings hardened and became waxlike over a period of several days and also darkened in 40 41 236 color with age. In our water droplet test (see Methods), we observed that a droplet of 42 43 44 237 water placed on the inside surface of a cell was not absorbed into the cell lining, indicating 45 46 238 the formation of an impervious water barrier likely created by the added floral oils. 47 48 49 239 The last stage of brood cell construction was the collection and addition of 50 51 240 additional Krameria floral oil used for larval feeding and construction of the operculum. 52 53 54 241 The female made an average of 2.63 ± 0.77 (n = 91) oil-collecting trips to complete the 55 56 242 cell. Females initially deposited and mixed a small amount of oil into the pollen mass. 57 58 59 243 The oil was thoroughly mixed with the pollen provisions, resulting in a uniform texture 60 10 61 62 63 64 65 244 and consistency (Fig. 1 c-d). After the provisioning of the brood cell, the female began 1 2 245 the process of oviposition. The average time to lay the egg was 1.92 ± 0.21 min (n = 5). 3 4 5 246 The female then started the operculum by supported herself on the sides of the cell and 6 7 247 with a swift rotary motion, deposited the remaining oil on top to close the cell completely. 8 9 10 248 The resulting concave surface is then flattened by the female who deposits sand above it 11 12 249 using her front and hind legs, leaving the operculum flat (that is, without an apical process 13 14 15 250 (or nipple) characteristic of brood cells opercula in some species of Centris ). The number 16 17 251 of foraging trips made to provision cells differed in relation to what was being collected 18 19 20 252 and for what purpose. The mean number of trips for oil collection for cell lining, for pollen 21 22 253 collection, and for oil collection for larval sustenance differed significantly from each 23 24 254 other (F 2,220 = 614.02, p<0.0001). 25 26 27 255 Females built and provisioned from 1 to 7 cells for each nest (4.55 ± 1.6; n = 20). 28 29 256 No marked females were observed to build more than one nest. Each cell required an 30 31 32 257 average of about two and a half days to construct (2.58 ± 0.40; n = 20) (Table 1). To 33 34 258 unload oil and line the cell, females spent about 10.5 min (median; n = 749), while the 35 36 37 259 time required to release her scopal pollen load into a cell was about 5 min (median; n = 38 39 260 394). The time necessary to use oil during operculation was also about 5 min (median; n 40 41 261 = 119). Differences in time spent in these different tasks were significant (H 2,1262 = 196.4; 42 43 44 262 p<0.001). These times include both manipulation of the resource itself, as well as the time 45 46 263 spent inside the nest in complementary activities. 47 48 49 264 The plants where females collected resources (pollen, nectar and oil) occurred in 50 51 265 close proximity to their nests (ca. 160 m). The duration of foraging trips to these plants 52 53 54 266 by females differed according to the resources collected (oil versus pollen). Females spent 55 56 267 more time for pollen-collection trips (73 min., median; n = 394) than for oil-collection 57 58 59 268 trips (54 min., median; n = 868) [Z (U) =7.77; p<0.001] (Fig. 5). 60 11 61 62 63 64 65 269 Natural enemies 1 2 270 Natural enemies, parasites, were observed around the aggregation during the 3 4 5 271 nesting period. It was common to see females of mutillid wasps (Hoplomutilla biplagiata 6 7 272 Mickel and Tallium aracati Couple) walking on the sandstone dunes searching for open 8 9 10 273 nests. During these movements, the wasps antennated the ground and periodically entered 11 12 274 open nest tunnels. When a C. burgdorfi female was in her nest, the bee fought with the 13 14 15 275 wasp, causing the wasp fall from the surface of the dune to the ground below the nest site. 16 17 276 Ants of the genus Pheidole were also seen entering bee burrows. At sunset, there 18 19 20 277 was a noticeable increase in the number of ants circulating within the nest aggregation. 21 22 278 When C. burgdorfi females reached the nest in the morning they sometimes encountered 23 24 279 tunnels full of ants. Often females did not reenter their burrows until the ants had left. 25 26 27 280 Mesoplia regalis Smith, a cleptoparasitic bee species, was also observed at the 28 29 281 aggregation sites. These bees have a characteristic flight pattern, always close to the 30 31 32 282 ground and flying very fast, making them difficult to see before landing. We observed 33 34 283 individuals of M. regalis flying over the entrances, but none were observed to enter nests. 35 36 37 284 Parasitic flies in the genus Anthrax were also collected at the nesting site. The 38 39 285 Anthrax females were always found at the base of the nesting aggregation, awaiting the 40 41 286 departure of a female, at which time they hovered over the tunnel entrance and flicked 42 43 44 287 eggs inside while still hovering about the entrances. 45 46 288 Of 125 brood cells maintained in the laboratory in only 8% did we observe a 47 48 49 289 parasite to emerge. Three cells were parasitized by mutillid wasps: H. biplagiata (n = 1) 50 51 290 and T. aracati (n = 2). These wasp larvae produce a fibrous cocoon within the host 52 53 54 291 cocoon, which makes it easy to identify the parasitism when cells are opened. In the rest 55 56 292 of the parasitized cells, we saw the emergence of the cleptoparasitic bee M. regalis (n = 57 58 59 293 1) and the parasitic fly Anthrax sp. (n = 6). 60 12 61 62 63 64 65 294 Temperature analysis 1 2 295 The temperatures inside the nest were similar to the outside ambient temperatures 3 4 5 296 in the early morning, when the lowest values were measured. The internal temperature 6 7 297 barely reached 28 °C even during the hottest periods of the day, while the outside 8 9 10 298 temperatures were above 30 °C between 09:30 h and 16:00 h (peaks above 36 °C at 12:00 11 12 299 h), the time period in which we also recorded the lowest humidity levels on site. The 13 14 15 300 difference in average temperatures between internal and external (ambient) from the nest 16 17 301 was significantly different (F 1, 425 = 350.29, p <0.001) (Fig. 6). 18 19 20 302 21 22 303 Discussion 23 24 304 Our results provide detailed information about the nesting behavior of C. 25 26 27 305 burgdorfi . The activity of the females probably follows the phenology of the plant species 28 29 306 from which they forage . The nest substrate (sandstone dunes) provides some thermal 30 31 32 307 isolation, important in an environment with high temperatures alongside the day. 33 34 308 Following a pattern seen in many solitary bees, C. burgdorfi is a species that nest in 35 36 37 309 aggregations. But we observed an elevated investment of time in constructing the brood 38 39 310 cells. 40 41 311 Many species of Centris nest in aggregations [e.g. ranging from 328 nests/m 2 in 42 43 2 2 44 312 C. caesalpiniae , 3 nests/m or less in C. pallida (Rozen & Buchmann, 1990), 40 nests/m 45 46 313 in C. aenea Lepeletier (Aguiar & Gaglianone, 2003), 708 holes/m 2 in C. muralis 47 48 2 49 314 Burmeister (Cilla & Rolón, 2012) and 1.7 nests/m in C. flavifrons Fabricius (Martins et 50 51 315 al. 2014)]. Also, we found many old brood cells of C. burgdorfi at the nest site that had 52 53 54 316 been established in previous years. This observation suggests that nesting aggregations 55 56 317 are often located in the same location across years. We consider the possible reasons why 57 58 59 60 13 61 62 63 64 65 318 nests are found in aggregations and why these aggregations tend to occur in the same 1 2 319 location across years. 3 4 5 320 First, some locations for nests may be better than others with respect to favorable 6 7 321 edaphic conditions, close proximity to food plants, and/or reduced parasite loads. Desired 8 9 10 322 locations may be limited in availability, causing nests to be closely spaced to one another. 11 12 323 For C. burgdorfi , the occurrence of dense aggregations could be related to preferences for 13 14 15 324 the sandy substrate chosen for nest construction. Females have a clear preference for 16 17 325 nesting in the hard-sandy substrate (sandstone dunes), formations that are uncommon and 18 19 20 326 restricted in area. The distribution of Anthophora pueblo Orr, another bee species that 21 22 327 nests in similar substrates, is apparently linked to sandstone, despite an elevated cost 23 24 328 associated with nesting in such substrate (Orr, Griswold, Pitts, & Parker, 2016). Other 25 26 27 329 aggregations of C. burgdorfi nests were found in similar dune structures indicating such 28 29 330 a possible preference. 30 31 32 331 Alternatively, bees might actively choose to place their nests close to where other 33 34 332 nests are because there are benefits of aggregations that outweigh the costs. Factors that 35 36 37 333 favor aggregation behavior, such as facilitating encounters by partners for mating might 38 39 334 offset factors that disfavor such behavior, such as increased attraction of parasites or 40 41 335 predators to these food-rich island-like targets (Michener, Lange, Bigarella, Salamuni, 42 43 44 336 1958). The two possible determinants of nest aggregations are not mutually exclusive. 45 46 337 Both could account for the aggregations of nests observed in these bees. 47 48 49 338 At this time, we do not know if nesting aggregations result in increased levels of 50 51 339 predation and parasitism. Among the natural enemies of Centris are several species of 52 53 54 340 cleptoparasitic bees in the Ericrocidini including Mesocheira , Mesoplia and 55 56 341 Mesonychium (Snelling & Brooks, 1985; Alves-dos-Santos, 2009). During studies of the 57 58 59 342 nesting biology of C. (Paracentris ) caesalpiniae in Arizona, Rozen & Buchmann (1990) 60 14 61 62 63 64 65 343 observed a species of Ericrocis (E. lata Cresson) parasitizing some nests but at a very 1 2 344 low level. However, since this genus does not occur in Brazil, other species of Ericrocidini 3 4 5 345 (e.g. Mesoplia regalis ) would be expected to parasitize C. burgdorfi nests. It was noted 6 7 346 in previous studies that C. burgdorfi females do not spend the night inside their nests, 8 9 10 347 unlike most species of Centris (Sabino et al., 2017) and most other bees. Not spending 11 12 348 the night in the nest is a highly unusual behavior in all Centris , and most solitary bees, 13 14 15 349 that could conceivably lead to more cleptoparasitic bee attacks and more predation by 16 17 350 other insects (since females would not be present to defend their nests). The same 18 19 20 351 behavior was observed recently in Epicharis nigrita Friese (Martins, Neto, Cruz, 2019). 21 22 352 Nevertheless, studies show that cleptoparasitic bees are active mainly during the day 23 24 25 353 (Wcislo et al., 2004) which would not necessarily mean a greater attack rate even within 26 27 354 unprotected nests in C. burgdorfi . 28 29 355 The choice of the nesting site probably also influences the development of the 30 31 32 356 brood. The difference between internal and external temperatures inside C. burgdorfi 33 34 357 nests demonstrates that the substrate acts as an excellent thermal insulator in protecting 35 36 37 358 the immatures. The development of immature bees, in general, is influenced by high 38 39 359 temperatures, which can cause high mortality, particularly in the early larval stages 40 41 42 360 (Frankie, Newstrom, Vinson, Barthell, & 1993). Temperatures between 38 °C and 40 °C 43 44 361 would be lethal to larvae of C. analis (Frankie, Vinson, Newstrom, & Barthell, 1988). 45 46 47 362 Nesting females of Nomia melanderi Cockerell, for example, respond to hot ambient 48 49 363 temperatures by excavating deeper nest burrows in the soil (Stephen, 1965). We do not 50 51 364 know the critical thermal maxima for bee larvae in the genus Centris nor the actual 52 53 54 365 temperatures inside the cells, and we do not have such data for the species. This would 55 56 366 be an interesting avenue for future research. 57 58 59 60 15 61 62 63 64 65 367 The shape of the brood cells of C. burgdorfi resembles that of other Centris species 1 2 368 (e.g. Coville, Frankie, Vinson, 1983; Silva, Viana, & Neves, 2001; Aguiar & Gaglianone, 3 4 5 369 2003; Cilla & Rolón, 2012; Martins et al., 2014). However, in C. burgdorfi , the top of the 6 7 370 operculum is concave, which differs from many other Centris species that add a central 8 9 10 371 process or “nipple” in the operculum (cell cap), such as C. (Centris ) aenea (Aguiar & 11 12 372 Gaglianone, 2003), C. (Centris ) varia Erichson (cited by Coville et al., 1983 as C. 13 14 15 373 segregate Crawford), C. (Centris ) flavifrons (Martins et al., 2014), and C. (Hemisiella ) 16 17 374 transversa Pérez (Batra & Schuster, 1977). A central process is present in other Centris 18 19 20 375 (Paracentris ) species, including C. pallida , C. rhodopus , C. cockerelli (Alcock et al., 21 22 376 1976) and C. caesalpiniae (Rozen & Buchmann, 1990). According to these authors, the 23 24 377 central process likely allows gas exchange between the external and the intranidal 25 26 27 378 environments through a pore, because gases do not diffuse as easily through the cell wall, 28 29 379 and the tough cells are also protected against flooding and from all natural enemies except 30 31 32 380 for small vertebrates (e.g. skunks or squirrels) that occasionally find, dig and eat the 33 34 381 Centris pallida cells (S. Buchmann, unpubl. data). We do not know exactly if differences 35 36 37 382 in the operculum shape in C. burgdorfi means differences in gas exchange and protection 38 39 383 against enemies. 40 41 384 The brood cells were also influenced by the sex of the larva. Females of C. 42 43 44 385 burgdorfi are larger than conspecific males and this resulted in larger cells built for 45 46 386 female-destined larvae, and more provisioned pollen and oil per cell. Although we did 47 48 49 387 not weight the pollen balls (provision masses), or the weight of fecal material, some 50 51 388 authors have reported that adult body size is associated with the amount of food consumed 52 53 54 389 by the larvae (e.g. Krombein, 1967; Klostermeyer et al., 1973; Alcock, 1979; Kim, 1997). 55 56 390 In C. analis , Jesus & Garofalo (2000) observed behavioral differences in foraging in 57 58 59 391 females according to whether they were provisioning a cell destined to be a female versus 60 16 61 62 63 64 65 392 a male bee. A larger number of pollen-collecting trips were made when the cell produced 1 2 393 a female offspring (Jesus & Garofalo, 2000). These results indicate that the production of 3 4 5 394 females costs more than the production of males (Jesus & Garofalo, 2000; Peterson & 6 7 395 Roitberg, 2006), which could affect local population sex ratios in the event of food 8 9 10 396 shortages during periods of drought. Previous studies in solitary bees have shown that the 11 12 397 ratio of males increases when food availability declines (Torchio & Tepedino, 1980; Kim, 13 14 15 398 1999). 16 17 399 Most activity occurs in the cooler morning hours. In the nesting location, the 18 19 20 400 climate is typical of that in the equatorial belt, with little annual variation in temperature, 21 22 401 but a pronounced division between the wet season (period where the plants are blooming 23 24 402 in the region) and dry season. Decreases in female flight activity in the afternoon are 25 26 27 403 likely related to the high ambient temperatures and possibly a decrease in the availability 28 29 404 of floral resources (standing crop of nectar and pollen) on-site by noon. Freitas & Pereira 30 31 32 405 (2004) reported a large drop in the flight activity of C. tarsata , and C. bicolor Lepeletier, 33 34 406 around 12:00 h due to the lower availability of pollen and oil after that period. 35 36 37 407 The end of the activity period is only one of the factors that can interrupt the 38 39 408 building of a brood cell and increase the overall time required to build it. The average 40 41 409 time required by C. burgdorfi (2.59 days) females to build a cell might be considered high 42 43 44 410 when compared with records for other Centris species. For example, C. flavifrons spends 45 46 411 an average of 1.61 days to construct each cell (Martins et al., 2014). But C. flavifrons and 47 48 49 412 C. burgdorfi species differ in the number of oil-collection trips required for lining their 50 51 413 cells. C. burgdorfi made 9.62 oil-collecting trips while C. flavifrons was observed taking 52 53 54 414 only three trips to collect floral oils. Similarly, Centris aenea was reported to make fewer 55 56 415 oil-collection trips to line its cells (two to three trips) and females spent less time in pollen 57 58 59 416 collection (33.6 min in C. aenea versus 68 min in C. burgdorfi ) (Aguiar & Gaglianone, 60 17 61 62 63 64 65 417 2003). The time that Centris females spend gathering each type of floral resource is highly 1 2 418 variable and likely related to the distance between the nest and the floral species that they 3 4 5 419 utilize. Distance to the floral resources does not appear to be a problem for C. burgdorfi 6 7 420 females since the plants used to occur in the vicinity (c.a. 200 m) to the nesting site. 8 9 10 421 The present study contributes to our overall understanding of the nesting biology 11 12 422 of Centris bees. We successfully followed the construction of brood cells in nature step 13 14 15 423 by step, which is rare in the study of bee nesting biology. Since C. burgdorfi is highly 16 17 424 unusual in several aspects of its nesting behavior (e.g. nesting in the sandstone and 18 19 20 425 females not spending the night in their burrows), it would be helpful to locate other more 21 22 426 geographically distant nesting localities for this bee, to confirm many of the observations 23 24 427 at the northeastern Brazil study sites. In fact, we are undertaking an expanded study of C. 25 26 27 428 burgdorfi by searching for other nesting sites of this remarkable centridine bee. 28 29 429 30 31 32 430 Acknowledgments 33 34 431 The authors are grateful to Paulo Roberto de Castro for the help and all support in 35 36 37 432 the fieldwork and to Astrid Kleinert, Celso Martins, Clemens Schlindwein, Felipe 38 39 433 Vivallo, Maria Cristina Gaglianone and the two anonymous reviewers for comments and 40 41 434 suggestions in the manuscript. The research was supported by Fundação de Amparo à 42 43 44 435 Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Desenvolvimento 45 46 436 Científico e Tecnológico (CNPq). 47 48 49 437 50 51 438 References 52 53 54 439 Aguiar, C. M. L. (2003). Flower visits of Centris bees (Hymenoptera: Apidae) in an area 55 56 440 of caatinga (Bahia, Brazil). Studies on Neotropical Fauna and Environment , 38 57 58 59 441 (1), 41-45. 60 18 61 62 63 64 65 442 Aguiar, C. M. L., & Gaglianone, M. C. (2003). Nesting biology of Centris (Centris ) aenea 1 2 443 Lepeletier (Hymenoptera, Apidae, Centridini). Revista Brasileira de Zoologia , 20 3 4 5 444 (4), 601-606. 6 7 445 Alcock, J. (1979). The relation between female body size and provisioning behavior in 8 9 10 446 the bee Centris pallida Fox (Hymenoptera: Anthophoridae). Journal of the 11 12 447 Kansas Entomological Society , 52 (3), 623-632. 13 14 15 448 Alcock, J., Jones, C. E., & Buchmann, S. L. (1976). The nesting behavior of three species 16 17 449 of Centris bees (Hymenoptera: Anthophoridae). Journal of the Kansas 18 19 20 450 Entomological Society , 49 (4), 469-474. 21 22 451 Alves-dos-Santos, I. (2009). Cleptoparasite bees, with emphasis on the oil bees hosts. 23 24 452 Acta Biológica Colombiana , 14, 107-114. 25 26 27 453 Batra, S. W. T., & Schuster, J. C. (1977). Nests of Centris , Melissodes , and Colletes in 28 29 454 Guatemala (Hymenoptera: Apoidea). Biotropica , 9 (2), 135-138. 30 31 32 455 Buchmann, S. L. (1987). The ecology of oil flowers and their bees. Annual Review of 33 34 456 Ecology and Systematics , 18, 343-369. 35 36 37 457 Cane, J. H. (1987). Estimation of bee size using intertegular span (Apoidea). Journal of 38 39 458 the Kansas Entomological Society , 60 (1), 145-147. 40 41 459 Cilla, G., & Rolón, G. (2012). Macroscopic and microscopic studies of the nests and the 42 43 44 460 stages involved in the nesting process of Centris muralis Burmeister 45 46 461 (Hymenoptera: Apidae: Centridini) bee in the adobe walls, in La Rioja, Argentina. 47 48 49 462 Biologia , 67 (3), 573-583. 50 51 463 Coville, R. E., Frankie, G. W., & Vinson, S. B. (1983). Nests of Centris segregata 52 53 54 464 (Hymenoptera: Anthophoridae) with a review of the nesting habits of the genus. 55 56 465 Journal of the Kansas Entomological Society , 56 (2), 109-122. 57 58 59 60 19 61 62 63 64 65 466 Frankie, G. W., Vinson, S. B., Newstrom, L., & Barthell, J. F. (1988). Nest site habitat 1 2 467 preferences of Centris bees in the Costa Rican dry forest. Biotropica , 20 (4), 301- 3 4 5 468 310. 6 7 469 Frankie, G. W., Newstrom, L., Vinson, S. B., & Barthell, J. F. (1993). Nesting-habitat 8 9 10 470 preferences of selected Centris bee species in Costa Rican dry forest. Biotropica , 11 12 471 25 (3), 322-333. 13 14 15 472 Freitas, B. M., & Paxton, R. J. (1998). A comparison of two pollinators: the introduced 16 17 473 honey bee Apis mellifera and an indigenous bee Centris tarsata on cashew 18 19 20 474 Anacardium occidentale in its native range of NE Brazil. Journal of Applied 21 22 475 Ecology , 35, 109–121. 23 24 476 Freitas, B. M., & Pereira, J. O. P. (2004). Crop consortium to improve pollination: can 25 26 27 477 West Indian Cherry ( Malpighia emarginata ) attract Centris bees to pollinate 28 29 478 cashew ( Anacardium occidentals ). In Freitas, B. M., & Pereira, J. O. P (Eds.). 30 31 32 479 2004. Solitary bees: conservation, rearing and management for pollination (pp. 33 34 480 193-201). Fortaleza: Imprensa Universitária. 35 36 37 481 Garófalo, C. A., Camillo, E., & Serrano, J. C. (1989). Espécies de abelhas do gênero 38 39 482 Centris (Hymynoptera, Anthophoridae) nidificando em ninhos-armadilha. [Bee 40 41 483 species of Centris genus (Hymenoptera, Anthophoridae) nesting in trap nests] 42 43 44 484 Ciencia e Cultura, 41, 799. 45 46 485 INMET/UFRN. (2002). Instituto Nacional de Meteorologia/UFRN. Boletins 47 48 49 486 Climatológicos: anos 1984 a Maio/2002. [Climatological Bulletins: years 1984 to 50 51 487 May / 2002] Estação Climatológica Principal. Natal. 52 53 54 488 Janvier, H. (1926). Recherches biologiques sur les Hymenopteres du Chili. [Biological 55 56 489 research on Hymenopteres of Chile]. Annales des Sciences Naturelles, 10, 133- 57 58 59 490 268. 60 20 61 62 63 64 65 491 Janvier, H. (1955) Le nid et la nidification chez quelques abeilles des Andes tropicales. 1 2 492 [Nest and nesting of some tropical Andes bees]. Annales des Sciences Naturelles , 3 4 5 493 17 (11), 311-49. 6 7 494 Jesus, B. M. V., & Garófalo, C. A. (2000). Nesting behavior of Centris (Heterocentris ) 8 9 10 495 analis (Fabricius) in southeastern Brazil (Hymenoptera, Apidae, Centridini). 11 12 496 Apidologie , 31, 503-515. 13 14 15 497 Kim, J. Y. (1997). Female size and fitness in the leaf-cuter bee Megachile apicalis . 16 17 498 Ecological Entomology , 22, 275-282. 18 19 20 499 Kim, J. Y. (1999). Influence of resource level on maternal investment in a leaf-cutter bee 21 22 500 (Hymenoptera: Megachilidae). Behavioral Ecology , 10, 552–556. 23 24 501 Klostermeyer, E. C., Mech, S. J., & Rasmussen, W. B. (1973). Sex and weight of 25 26 27 502 Megachile rotundata (Hymenoptera: Megachilidae) progeny associated with 28 29 503 provision weights. Journal of the Kansas Entomological Society , 46, 536-548. 30 31 32 504 Krombein, K. V. (1967). Trap-nesting wasp and Bees: life histories nest and associates. 33 34 505 Washington, DC: Smithsonian Press. 35 36 37 506 Martins, C. F., Peixoto, M. P., & Aguiar, C. M. L. (2014). Plastic nesting behavior of 38 39 507 Centris (Centris ) flavifrons (Hymenoptera: Apidae: Centridini) in an urban area. 40 41 508 Apidologie , 45 (2), 156-171. 42 43 44 509 Martins, C. F., Neto, V. I. D. S., & Cruz, R. D. M. (2019). Nesting biology and mating 45 46 510 behavior of the solitary bee Epicharis nigrita (Apoidea: Centridini). Journal of 47 48 49 511 Apicultural Research , 1-10. 50 51 512 Michener, C. D., Lange, R. B., Bigarella, J. J., & Salamuni, R. (1958). Factors influencing 52 53 54 513 the distribution of bees' nests in earth banks. Ecology , 39 (2), 207-217. 55 56 514 Neto, O. F., Costa, F. A. A., Severo, R.N. F., Júnior, O. F. S., & Scudelari, A. C. (2005). 57 58 59 515 Estudo da dinâmica de Falésias do município de Tibau do Sul – RN. [Study of the 60 21 61 62 63 64 65 516 Dynamics of Cliffs of the city of Tibau do Sul - RN.] Anais do 11° Congr. Brasil. 1 2 517 Geologia de Engenharia e Ambiental. Florianópolis – SC. 3 4 5 518 Nogueira, A. M. B. (1981). O cenozoico continental da região de Natal. [The continental 6 7 519 Cenozoic of the Natal region] (Master’s thesis) . CT/UFPE, Recife. 8 9 10 520 Oliveira, R., & Schlindwein, C. (2009). Searching for a manageable pollinator for acerola 11 12 521 orchards: the solitary oil-collecting bee Centris analis (Hymenoptera: Apidae: 13 14 15 522 Centridini). Journal of Economic Entomology , 102 (1), 265-273. 16 17 523 Orr, M. C., Griswold, T., Pitts, J. P., & Parker, F. D. (2016). A new bee species that 18 19 20 524 excavates sandstone nests. Current Biology , 26 (17), R792-R793. 21 22 525 Perterson, J. H., & Roitberg, B. D. (2006). Impacts of flight distance on sex ratio and 23 24 526 resource allocation to offspring in the leafcutter bee, Megachile rotundata . 25 26 27 527 Behavioral Ecology and Sociobiology . 59, 589–596. 28 29 528 R Core Team. (2018). R: A language and environment for statistical computing. R 30 31 32 529 Foundation for Statistical Computing, Vienna, Austria. URL https://www.R- 33 34 530 project.org/. 35 36 37 531 Rozen, J. G., & Buchman, S. L. (1990). Nesting biology and immature stages of the bees 38 39 532 Centris caesalpiniae , C. pallida and the cleptoparasite Ericrocis lata 40 41 533 (Hymenoptera: Apiodea: Anthophoridae). American Museum Novitates , 2985, 1- 42 43 44 534 30. 45 46 535 Sabino, W. O., Silva, C. I., & Alves-dos-Santos, I. (2017). Mating system and sleeping 47 48 49 536 behaviour of the Male and Female Centris (Paracentris ) burgdorfi Friese (Apidae, 50 51 537 Centridini). Journal of Insect Behavior , 30 (1), 103-118. 52 53 54 538 Sabino, W. O., Alves-dos-Santos, I., & Silva, C. I. (2018). Versatility of the trophic niche 55 56 539 of Centris (Paracentris ) burgdorfi (Apidae, Centridini). Arthropod-Plant 57 58 59 540 Interactions , 1-11. 60 22 61 62 63 64 65 541 Silva, F. O., Viana, B. F., & Neves, E. L. (2001). Biologia e arquitetura de ninhos de 1 2 542 Centris (Hemisiella ) tarsata Smith (Hymenoptera, Apidae, Centridini). [Biology 3 4 5 543 and nest architecture of Centris (Hemisiella ) tarsata Smith (Hymenoptera, 6 7 544 Apidae, Centridini)]. Neotropical Entomology, 30 (4), 541-545. 8 9 10 545 Silveira, F. A., Melo, G. A. R., & Almeida, E. A. B. (2002). Abelhas brasileiras: 11 12 546 sistemática e identificação. [Brazilian bees: systematics and identification]. Belo 13 14 15 547 Horizonte: Fundação Araucária. 16 17 548 Snelling, R. R., & Brooks, R. W. (1985). A review of the genera of cleptoparasitic bees 18 19 20 549 of the tribe Ericrocini (Hymenoptera: Anthophoridae). Natural History Museum 21 22 550 of Los Angeles County , 369, 1-34. 23 24 551 Sokal, R., & Rohlf, F. J. (1995). Biometry – The principles and practice of Statistics in 25 26 27 552 biological research. New York, NY: W. H. Freeman and Co. 28 29 553 Stephen, W. P. (1965). Temperature effects on the development and multiple generations 30 31 32 554 in the alkali bee, Nomia melanderi Cockerell. Entomologia Experimentalis et 33 34 555 Applicata , 8(3), 228-240. 35 36 37 556 Torchio, P. F., & Tepedino, V. J. (1980). Sex ratio, body size and seasonality on a solitary 38 39 557 bee, Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae). Evolution , 40 41 558 34, 993-1003. 42 43 44 559 Vivallo, F., & Zanella, F. C. V. (2012). A new species of Centris ( Paracentris ) Cameron, 45 46 560 1903 from northeastern Brazil, with a key for the Centris species of the Caatinga 47 48 49 561 region (Hymenoptera: Apidae). Zootaxa , 3298, 1-16. 50 51 562 Vogel, S. (1974). Ölblumen und ölsammelnde Bienen. [Oil flowers and oil-collecting 52 53 54 563 bees]. Trop. Subtrop. Pflanzenwelt , 7, 285-547. 55 56 564 Wcislo, W. T., Arneson, L., Roesch, K., Gonzalez, V., Smith, A., & Fernández, H. (2004). 57 58 59 565 The evolution of nocturnal behaviour in sweat bees, Megalopta genalis and M. 60 23 61 62 63 64 65 566 ecuadoria (Hymenoptera: Halictidae): an escape from competitors and enemies? 1 2 567 Biological Journal of the Linnean Society , 83, 377–387. 3 4 5 568 Zanella, F. C. V. (2002). Sistemática, filogenia e distribuição geográfica das espécies sul 6 7 569 americanas de Centris (Paracentris ) Cameron, 1903 e de Centris (Penthemisia ) 8 9 10 570 Moure, 1950, incluindo uma análise filogenética do “grupo Centris” sensu Ayala, 11 12 571 1998 (Hymenoptera, Apoidea, Centridini). [Systematics, phylogeny and 13 14 15 572 geographical distribution of South American species of Centris (Paracentris ) 16 17 573 Cameron, 1903 and Centris (Penthemisia ) Moure, 1950, including a phylogenetic 18 19 20 574 analysis of the " Centris group" sensu Ayala, 1998 (Hymenoptera, Apoidea, 21 22 575 Centridini)]. Revista Brasileira de Entomologia, 46, 435-488. 23 24 576 Zar, J. H. (1996). Biostatistical Analysis. (3rd ed.). New Jersey, NJ: Prentice Hall. 25 26 27 577 28 578 29 30 579 31 32 580 33 34 581 35 582 36 37 583 38 39 584 40 41 585 42 43 586 44 45 587 46 588 47 48 589 49 50 590 51 52 591 53 54 592 55 593 56 57 594 58 59 595 60 24 61 62 63 64 65 596 Figure captions 1 2 597 3 4 5 598 Figure 1. A Centris burgdorfi nest in the sandstone substrate in northeastern, Brazil, 6 7 599 showing (a) the clumped arrangement of seven brood cells constructed by the female with 8 9 10 600 the access tunnel (arrow); (b) a cross-section of a cell under construction, showing the 11 12 601 cavity excavated by the female (arrow) that would subsequently be lined with oil; c) a 13 14 15 602 cross-section showing the egg deposited on top of the oil and pollen; d) removal of the 16 17 603 cell cap showing second instar of C. burgdorfi larva on top of a mass of pollen and oil. 18 19 20 604 21 22 605 Figure 2. Morphological analysis of Centris burgdorfi adult bees and 50 brood cells 23 24 606 of each sex. a) intertegular distance of random adult bees (25 males and 25 females); b) 25 26 27 607 height of brood cells; c) diameter of the cell opening; d) diameter of the cell base. The 28 29 608 main horizontal line shows the mean, boxes represent standard error, and whiskers depict 30 31 32 609 95% intervals. Asterisk symbol (*) represent significant differences and “ns” no 33 34 610 significant difference. 35 36 37 611 38 39 612 Figure 3. Number of resource collection trips made by Centris burgdorfi females in a 40 41 613 dune region in northeastern, Brazil, separated by time class. Data are pooled across 20 42 43 44 614 females. 45 46 615 47 48 49 616 Figure 4. Centris burgdorfi female during the construction of a cell: (a) female returns 50 51 617 to nest with pollen; (b) checks the cell before depositing material; (c) using rotational 52 53 54 618 movements, the female arranges the material uniformly using the hind legs; (d) exits 55 56 619 again. 57 58 59 620 60 25 61 62 63 64 65 621 Figure 5. Time spent by one Centris burgdorfi female (CB1 - Table 1) inside and 1 2 622 outside the nest during cell construction, provisioning, oviposition, and cell closure. 3 4 5 623 Black arrows indicate oil-collecting trips, and gray arrows indicate pollen-collecting trips. 6 7 624 Cell figure indicate the time in which one cell was completed, and the vertical dashed 8 9 10 625 lines separate each day. 11 12 626 13 14 15 627 Figure 6. Average temperatures measured simultaneously inside (continuous black 16 17 628 line) and outside (dashed red line) the Centris burgdorfi nests throughout the day, over 18 19 20 629 12 days. The symbols represent the mean and the bars are the standard deviations of the 21 22 630 measurements. The humidity of the environment (dotted green line) are also presented. 23 24 631 25 26 27 632 Table captions 28 29 633 Table 1. Number of days spent by each Centris burgdorfi female to construct one 30 31 32 634 brood cell in a dune region in northeastern, Brazil. 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 26 61 62 63 64 65 )LJXUHD )LJXUHE )LJXUHF )LJXUHG )LJXUHD )LJXUHE )LJXUHF )LJXUHG )LJXUH )LJXUHD )LJXUHE )LJXUHF )LJXUHG )LJXUH )LJXUH 7DEOH

Year Nest identification Number of cells built Activity days Days/cell 2014 CB1 4 11 2.75 CB2 6 17 2.83 CB3 3 7 2.33 CB4 3 10 3.33 CB5 7 14 2 CB6 4 9 2.25 CB7 4 10 2.5 CB8 5 14 2.8 CB9 3 10 3.33 CB10 5 12 2.4 CB11 1 3 3 CB12 5 9 1.8 2015 CB13 6 17 2.83 CB14 5 15 3 CB15 7 16 2.29 CB16 5 15 3 CB17 4 10 2.5 CB18 6 16 2.67 CB19 6 15 2.5 CB20 2 5 2.5 Average 4.55 11.75 2.58 Standard deviation 1.60 4.01 0.40