The studied nest aggregation occupied an area of 5.90 m x 1.61 m. Females started foraging around 7h00 am and continued leaving their nests until at least early afternoon. Observations on flowers recorded P. graminea females collecting pollen or nectar on flowers of 14 plant species within the campus (Table 1, Fig. S2). From this total, five species were also identified in the pollen samples removed from females' scopae and other 11 pollen types were also recorded in the samples (Table 1, Fig. 1). Samples from March 2019 had only five pollen types whilst 13 types were identified in samples from March 2022, with only two species, Pleroma granulosum (Desr.) D. Don (Melastomataceae) and Solanum palinacanthum Dunal (Solanaceae), recorded in samples from both studied years (Table 1). Additionally, the proportion of pollen types collected by females in 2019 was markedly characterised by the dominance of P. granulosum, with 73.19% of the pollen grains counted, followed by Solanum sisymbriifolium Lam. (Solanaceae) (15.03%), S. palinacanthum (6.8%), and Bauhinia variegata L. (Fabaceae) (3.48%). In 2022, the percentages of the four species more visited by females were less uneven, with Senna spectabilis (DC.) H.S.Irwin & Barneby (Fabaceae) (33.21%) as the main pollen source, followed by P. granulosum (25.16%), S. palinacanthum (13.47%), and Chamaecrista rotundifolia (Pers.) Greene (Fabaceae) (11.93%). It was detected a strong preference for pollen collection on flowers with poricidal anthers, with eight species belonging to seven genera and five families (Cassia, Chamaecrista, Senna, Solanum, Apeiba, Pleroma, and Bixa) (Table 1). From all 12 pollen sources used by P. graminea females, only four plant species have no poricidal anthers: Psidium guajava L. (Myrtaceae), Mimosa skinneri Benth., B. variegata, and Type Paloue (Fabaceae), and were visited for pollen gathering by only one female each. The most important pollen sources identified in pollen samples, adding data from both studied periods, were: P. granulosum (46.51%), S. spectabilis (18.45%), and S. palinacanthum (10.51%). The achiote, Bixa orellana L. (Bixaceae), was the lone pollen source identified only through observations on flowers. Thirteen nectar sources were recorded for P. graminea; only five of them through pollen analysis. Almost all pollen sources consist of native plant species, but P. guajava, B. variegata and Cassia fistula L. (Fabaceae), while the nectar sources species, in turn, are mostly exotic (Table 1). In most of the nectar sources, especially the plants with long-tubed corolla flowers, females robbed nectar by piercing the base of corolla to collect this flower resource (Fig. S1). Even in plants with funnel-shaped corolla such as Turnera subulata Sm. (Turneraceae) (Fig. S1), females collected nectar without touching the anthers or the stigma; thus, not performing pollination.
Table 1
Plant species, their origin, their codes used in interaction networks, their floral resources collected by females and the relative frequency of occurrence of their pollen grains counted on the pollen samples removed from Pseudaugochlora graminea females' scopae at the Santa Mônica campus in Uberlândia, Brazil in March 2019 and March 2022. The asterisk indicates plant species that females were observed collecting resources on flowers.
Plant species
|
Proportion (%)
|
Codes
|
Resource(s) collected
|
Plant origin
|
2019
|
2022
|
Total
|
Acanthaceae
|
|
|
|
|
|
|
Asystasia gangetica*
|
-
|
-
|
-
|
-
|
nectar
|
naturalised
|
Asteraceae
|
|
|
|
|
|
|
Centratherum punctatum
|
-
|
0.01
|
0.006
|
P
|
nectar
|
native
|
Bignoniaceae
|
|
|
|
|
|
|
Fridericia platyphylla*
|
-
|
-
|
-
|
-
|
nectar
|
native
|
Tecoma stans*
|
-
|
0.04
|
0.022
|
N
|
nectar
|
naturalised
|
Bixaceae
|
|
|
|
|
|
|
Bixa orellana*
|
-
|
-
|
-
|
-
|
pollen
|
native
|
Cucurbitaceae
|
|
|
|
|
|
|
Luffa cylindrica
|
-
|
0.02
|
0.011
|
M
|
nectar
|
cultivated
|
Fabaceae
|
|
|
|
|
|
|
Bauhinia variegata
|
3.48
|
-
|
1.54
|
A
|
pollen
|
naturalised
|
Calliandra brevipes*
|
-
|
-
|
-
|
-
|
nectar
|
native
|
Cassia fistula
|
-
|
4
|
2.22
|
G
|
pollen
|
cultivated
|
Cenostigma pluviosum
|
-
|
0.19
|
0.106
|
L
|
nectar
|
native
|
Chamaecrista rotundifolia*
|
-
|
11.93
|
6.63
|
H
|
pollen
|
native
|
Mimosa skinneri
|
-
|
3.87
|
2.15
|
J
|
pollen
|
native
|
Senna spectabilis*
|
-
|
33.21
|
18.45
|
F
|
pollen
|
native
|
Type Paloue
|
1.51
|
-
|
0.67
|
B
|
pollen
|
?
|
Malvaceae
|
|
|
|
|
|
|
Apeiba tibourbou
|
-
|
3.95
|
2.194
|
I
|
pollen
|
native
|
Melastomataceae
|
|
|
|
|
|
|
Pleroma granusolum*
|
73.19
|
25.16
|
46.51
|
E
|
pollen
|
native
|
Myrtaceae
|
|
|
|
|
|
|
Psidium guajava
|
-
|
4.14
|
2.3
|
K
|
pollen
|
naturalised
|
Rubiaceae
|
|
|
|
|
|
|
Ixora casei*
|
-
|
-
|
-
|
-
|
nectar
|
cultivated
|
Mussaenda philippica*
|
-
|
-
|
-
|
-
|
nectar
|
cultivated
|
Sapindaceae
|
|
|
|
|
|
|
Serjania sp.
|
-
|
0.01
|
0.006
|
O
|
nectar
|
?
|
Solanaceae
|
|
|
|
|
|
|
Solanum palinacanthum*
|
6.8
|
13.47
|
10.51
|
D
|
pollen
|
native
|
Solanum sisymbriifolium
|
15.03
|
-
|
6.68
|
C
|
pollen
|
native
|
Turneraceae
|
|
|
|
|
|
|
Turnera subulata*
|
-
|
-
|
-
|
-
|
nectar
|
native
|
Verbenaceae
|
|
|
|
|
|
|
Duranta erecta*
|
-
|
-
|
-
|
-
|
nectar
|
naturalised
|
Lantana camara*
|
-
|
-
|
-
|
-
|
nectar
|
naturalised
|
All plant species recorded as floral sources were located within the campus and most of them nearby the nest aggregation (around 250 m from the nests) (Fig. 2). The two main pollen sources, P. granulosum and S. spectabilis, were represented by only eight and two individuals, respectively (Fig. 2). While S. palinacanthum individuals occurred in a great spot within the campus, S. sisymbriifolium was recorded by only one specimen nearby the nest aggregation until February 2022, when this single plant was killed by a mowing conducted in the same month. Likewise, specimens belonging to other Solanum species, to M. skinneri, and to the three Chamaecrista species (Fabaceae) identified in the study area were constantly mowed during the rainy season. Individuals of some nectar sources such as T. subulata, Luffa cylindrica (L.) M.Roem. (Cucurbitaceae) and Centratherum punctatum Cass. (Asteraceae) were also mowed. According to the University landscaping department, mowing occurs from 10 to 15 times to control the grasses in the campus, especially during the rainiest months (from December to March). Mowing was performed both with trimming machines and with a riding lawn mower, which was used on the large lawn areas within the campus. These gardening practices affect the flora composition at least for some months, since several herbs, prostate vines, weeds and small shrubs are permanently eliminated or may sprout after some weeks. On the other hand, the plant replacement performed in the campus can be beneficial for P. graminea females since species such as Asystasia gangetica (L.) T.Anderson (Acanthaceae) and Lantana camara L. (Verbenaceae), both visited by the bees for nectar collection, are used to cover areas with bare soil when seedlings are available. Nonetheless, the species Syngonium angustatum Schott (Araceae), which is not used by the females for floral resources gathering, is the most used in gardening due to the ease of its propagation and resistance. Regarding tree species, the plant compensation occurred mostly to replace dead trees or to form a green belt in large lawns and all tree seedlings used are donations from the municipal garden. The plantings were performed only during the rainy season. Among the tree species donated were native species: P. granulosum, Cenostigma pluviosum (DC.) Gagnon & G.P.Lewis (Fabaceae) Handroanthus chrysotrichus (Mart. ex DC.) Mattos, Tabebuia roseoalba (Ridl.) Sandwith (Bignoniaceae) but also exotic trees such as Callistemon viminalis (Sol. ex Gaertn.) G.Don (Myrtaceae), B. variegata, and Tecoma stans (L.) Juss. ex Kunth (Bignoniaceae).
All interaction networks between P. graminea females and plant species were highly specialised (C = 0.25, H'2 = 0.997 for March 2019; C = 0.142, H'2 = 0.987 for March 2022; C = 0.099, H'2 = 0.979 for the total network). From the 20 females sampled in 2019, 16 (80%) collected pollen on flowers of a single plant species (Fig. 3). This percentage was much lower in 2022, with 11 females (44%) visiting only one plant species (Fig. 3). For the total network, the plants with the highest numbers of interactions, were the main pollen sources: P. granulosum (D = 28), S. spectabilis (D = 9), and S. palinacanthum (D = 7). As for females, in both years, 11 bees interacted with two plant species; six females interacted with three plants, and only one P. graminea female (I43) interacted with four floral sources (Fig. 3). Among these 18 females, most of them had very low proportion of pollen grains belonging to other plant species (less than 12 grains), which might indicate pollen contamination rather than active pollen collection. Only four females (I4, I14, I32, and I36) (Fig. 3) presented similar proportions of two pollen sources, what suggests a concomitant pollen gathering on flowers of two plant species.