Queens become workers: pesticides alter caste differentiation in bees
www.nature.com/scientificreports
OPEN
Queens become workers: pesticides
alter caste differentiation in bees
Charles F. dos Santos1, André L. Acosta2,3, Andressa L. Dorneles1, Patrick D. S. dos Santos1 &
Betina Blochtein1,4
received: 24 February 2016
accepted: 21 July 2016
Published: 17 August 2016
Bees are important for the world biodiversity and economy because they provide key pollination
services in forests and crops. However, pesticide use in crops has adversely affected (decreased)
queen production because of increased mortality among larvae. Here, we demonstrated that in vitroreared queens of a neotropical social bee species (Plebeia droryana) also showed high larval mortality
after exposure to an organophosphate pesticide (chlorpyrifos) via larval food. Moreover, most of the
surviving larvae that were destined to develop into queens became workers more likely because they
ate less food than expected without pesticide skewing thus caste differentiation in this bee species.
This adverse effect has not been previously reported for any other social insects, such as honeybees or
bumblebees. Queens are essential for breeding and colony growth. Therefore, if our data are applicable
to other pantropical social bee species across the globe, it is likely that these bees are at a serious risk of
failure to form new colonies.
The bee population is declining worldwide, and pesticides are among the controversial factors behind this
phenomenon1–3. Although pesticides can effectively combat agricultural pest insects, they may have adverse
sub-lethal effects on beneficial insects including bees compromising thus ecological service of pollination1–3.
For example, pesticides can severely compromise cognition, foraging, navigation, homing, and memory abilities
of honeybee and bumblebee workers4–6. Similarly, queens may also develop serious problems when exposed to
pesticides: damage to ovarian tissues, high mortality, and workers’ rejection as well as difficulties with emerging,
mating, and laying eggs (Supplementary Table S1).
To date, most of the toxicological research (on pesticides) has been focused on the vulnerability of honeybees and bumblebees (Supplementary Fig. S1) because they are prime pollinators in the global agriculture.
Nevertheless, these bees are not always the most effective pollinators in many regions of the world7,8. For instance,
on pantropical region the stingless bees play an important role as pollinators there7,9,10. They are ca. 45-fold and
twice richer in species than honeybees and bumblebees, respectively, encompassing approximately 450 species11,12
(Supplementary Fig. S1). Only in Brazil it is believed that there are more than 300 stingless bee species12 which
add high economic value (billions of US dollars)13 to crops commercialized worldwide7,9,10,13.
Stingless bees are eusocial insects forming colonies with two distinct female castes, i.e. sterile workers and
reproductive queens14. As such, they share features observed in caste determination system of both honeybees
and bumblebees as, for example, larvae destined to become queens receive/consume larger amount of larval
food than those that become workers14–16. Furthermore, stingless bees (as also bumblebees) do not receive royal
jelly during larval development like seen for honeybees14. Additionally, caste determination in stingless bees is
characterised by massive differential larval nourishment14,15, except the genus Melipona which has an alternative
pathway17. Therefore, female larvae of stingless bees that are destined to develop into queens are reared in larger
brood cells (royal cells) and hence receive more food than female larvae that will become workers14,15. This situation most probably causes corpora allata to synthesise greater amounts of juvenile hormone (JH) during larval
development, thereby inducing queen characteristics18,19.
At present, it is unknown whether queen larvae of stingless bees may suffer any sub-lethal effects as a result
of possible exposure to pesticides. Floral resources collected by bees to rear new individuals have been found to
1
Departamento de Biodiversidade e Ecologia, Faculdade de Biociências, Pontifícia Universidade Católica do Rio
Grande do Sul, Av. Ipiranga, 6681, 90619-900 Porto Alegre, RS, Brazil. 2Departamento de Ecologia, Universidade de
São Paulo, Rua do Matão, 321, Travessa 14, 05508-090 São Paulo, SP, Brazil. 3Núcleo de Pesquisa em Biodiversidade
e Computação - BioComp. Escola Politécnica, Av. Prof. Luciano Gualberto, Trav. 3, n. 380, 05508-010 São Paulo,
São Paulo, Brazil. 4Instituto do Meio Ambiente, Pontifícia Universidade Católica do Rio Grande do Sul, Av. Ipiranga,
6681, 90619-900 Porto Alegre, RS, Brazil. Correspondence and requests for materials should be addressed to C.F.d.S.
(email: ) or B.B. (email: )
Scientific Reports | 6:31605 | DOI: 10.1038/srep31605
1
www.nature.com/scientificreports/
Live
GLMM parameters
Treament
Survival probability plus
(μg a.i./bee)
95% CI
Dead
Workers
Intercastes
Queens
Total
Estimate
Std. Error
z value
Pr(>|z|)
control
0.76 [0.664–0.875]
15
0
0
48
63
−0.3125
2.1684
−0.144
0.885414
0.0088
0.62 [0.510–0.751]
24
13
0
26
63
13.2258
3.4019
3.888
0.000101
0.0176
0.32 [0.201–0.532]
36
8
0
19
63
11.7151
4.1211
2.843
0.004473
0.0264
0.20 [0.123–0.332]
50
5
1
7
63
10.6836
3.1231
3.421
0.000624
0.0352
0.13 [0.064–0.281]
51
8
0
4
63
5.6065
2.8915
1.939
0.048510
0.044
0.09 [0.031–0.262]
54
5
1
3
63
10.5075
3.0898
3.401
0.000672
0.088
0.00 [0.008–0.012]
62
1
0
0
63
8.2827
2.4862
3.331
0.000864
Total
292
40
2
107
441
Table 1. The results of the GLMM for factors (chlorpyrifos doses) affecting survival probability of bees
and its skewing in caste differentiation. a.i./bee: active ingredient per bee; 95% CI: 95% confidence interval;
Dead: amount of immature bees (mostly larvae) that did die (i.e. not emerged) per treatment; Live: amount of bees
that did emerge per treatment and its respective caste after we did perform a morphological analysis for presence/
absence of spermatheca and corbiculae; Total: total number of larvae transferred for rearing plates per treatment;
Std. Error: Standard Error; z-value: Standard score, i.e. standard deviations from their means. Negative values
when raw score is below the mean, positive when above; Pr(>|z|): probability to find z-scores by chance.
be contaminated with numerous agrochemicals including organophosphate pesticide chlorpyrifos (CPY)20–26.
(Supplementary Table S2). While adverse effects on pesticide exposed bees have repeatedly been reported20–25,27 it
is currently unknown how queen production in stingless bees may be affected, if at all.
Currently, CPY is widely used to control agricultural pest insects worldwide28,29 (Supplementary Fig. S2). In
Brazil, for example, cross-checking of adjacent crops (where CPY is recommended for one of the crops) resulted
in a grave warning about p (...truncated)
