Gut microbial communities are critical for the health of many insect species. However, little is known about how gut microbial communities respond to anthropogenic changes and how such changes affect host-pathogen interactions. In this study, we used deep sequencing to investigate and compare the composition of gut microbial communities within the midgut and ileum (both bacteria and fungi) in Bombus terrestris queens collected from natural (forest) and urbanized habitats. Additionally, we investigated whether the variation in gut microbial communities under each habitat affected the prevalence of two important bumblebee pathogens that have recently been associated with Bombus declines (Crithidia bombi and Nosema bombi). Microbial community composition differed strongly among habitat types, both for fungi and bacteria. Fungi were almost exclusively associated with bumblebee queens from the forest habitats, and were not commonly detected in bumblebee queens from the urban sites. Further, gut bacterial communities of urban B. terrestris specimens were strongly dominated by bee-specific core bacteria like Snodgrassella (Betaproteobacteria) and Gilliamella (Gammaproteobacteria), whereas specimens from the forest sites contained a huge fraction of environmental bacteria. Pathogen infection was very low in urban populations and infection by Nosema was only observed in specimens collected from forest habitats. No significant relationship was found between pathogen prevalence and microbial gut diversity. However, there was a significant and negative relationship between prevalence of Nosema and relative abundance of the core resident Snodgrassella, supporting its role in pathogen defense. Overall, our results indicate that land-use change may lead to different microbial gut communities in bumblebees, which may have implications for bumblebee health, survival and overall fitness.
Citation: Bosmans L, Pozo MI, Verreth C, Crauwels S, Wilberts L, Sobhy IS, et al. (2018) Habitat-specific variation in gut microbial communities and pathogen prevalence in bumblebee queens (Bombus terrestris). PLoS ONE 13(10): e0204612. https://doi.org/10.1371/journal.pone.0204612
Editor: James C. Nieh, University of California San Diego, UNITED STATES
Received: May 28, 2018; Accepted: September 11, 2018; Published: October 25, 2018
Copyright: © 2018 Bosmans et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was supported by KU Leuven (Postdoctoral Mandate, grant to L.B. (PDM/17/132)) and the Research Foundation Flanders (FWO) (grant to M.I.P. (12A0716N)). Biobest provided salary support to FW, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific role of FW is articulated in the ‘author contributions’ section.
Competing interests: We have the following interests. F. WACKERS is affiliated to Biobest. There are no patents, products in development or marketed products to declare. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors.
Insects represent one of the most species-rich animal groups on Earth [1, 2] and play an important role in ecosystem functioning . Insects provide numerous important ecosystem services such as pollination, crop protection and detrivory, nutrient cycling, and providing a food source for higher trophic levels [4, 5]. Recent studies have suggested that insects are drastically declining worldwide , particularly pollinators . Habitat loss and fragmentation, decline in food resources and nest availability, and increased use of pesticides, as well as climate change have been proposed as the most important factors leading to insect decline . Recently, increased prevalence of diseases and parasites has been suggested to contribute to pollinator decline as well . However, there is increasing consensus that there is likely no single factor that can explain this severe decline, but rather a complex interaction of many factors that act together .
While effects of anthropogenic disturbances on insect diversity have been extensively studied [10, 11], little is known about the effect of land-use change on the microbes that live in close association with insects and how anthropogenic changes affect pathogen infection in insects. It is known that diverse microbial communities inhabit insect guts (but exceptions exist, see ) and that they can provide critical functions for their hosts, such as nutrient acquisition, food digestion, regulation of immune responses, insecticide resistance and defense against pathogens and parasites [13–17]. They can also influence host development, behaviour and reproduction [14, 18, 19]. Disruption or changes of the gut microbiome may therefore have important consequences for the host, often resulting in adverse health and fitness effects . The exact species composition of the gut microbiome is influenced by many factors, including diet, host genetics, immune responses, stress, exposure to antimicrobials, interactions between members of the gut microbial community, and other factors such as flower availability, local environmental conditions and the pool of environmental bacteria that may invade and colonize the insect gut [19, 21–23]. It can therefore be hypothesized that land-use conversion affects the gut microbial community composition, and therefore also defense against pathogens and overall health .
The overall objective of this study was to determine whether gut microbial communities of wild bumblebee queens are influenced by urbanization, and whether potential differences in community composition translate in different pathogen prevalence. Previous research has shown that bumblebees harbour a specialized, species-poor gut microbial community, which is dominated by a limited set of bee-specific core bacteria, including Gilliamella (Gamma-1 phylotype; Gammaproteobacteria; Orbaceae), Snodgrassella (Beta; Gammaproteobacteria; Neisseriaceae), Lactobacillus (Firm-4/Lacto-2 and Firm-5/Lacto-1; Firmicutes; Lactobacillaceae) and bifidobacteria (Actinobacteria) [24–27]. Cariveau et al.  found that anthropogenic change only marginally affected the gut microbial communities of Bombus workers. However, in their study only the effect of agricultural activities was investigated, while effects of other anthropogenic disturbances such as urbanization remain to be studied, as well as effects on queens. We first examined and compared the composition of gut microbial communities within the midgut and ileum (both bacteria and fungi) in Bombus terrestris queens that were collected from natural (forest) and urbanized habitats. Next, we investigated whether habitat type and variation in gut microbial community composition affected the occurrence of two important bumblebee pathogens, Crithidia bombi (Kinetoplastida; Trypanosomatidae) and Nosema bombi (Microsporidia; Nosematidae). Together with a few other pathogens like Apicystis bombi (Neogregarinorida; Ophryocystidae), these two species are believed to have contributed to the observed declines of Bombus populations [28, 29]. Previous research has shown that infection by these pathogens may be related to the gut microbial community composition, although relationships were not always significant [17, 24, 26, 30].
Materials and methods
Study species and collection of specimens
Experiments were performed using queens of Bombus terrestris (L.) (Hymenoptera: Apidae). Bombus terrestris is one of the most numerous bumblebee species in Europe and known to be an important pollinator of several wild plants and crops . Specimens were collected in early spring (March/April) within 7 days when the first queens that emerged from hibernation were observed in the field. No permission was required to perform the sampling; the sampling dit not involve endangered or protected species. Indiviuals were sampled from two natural (forest) locations (further referred to as site “S1-F” (Heverlee) and “S2-F” (Viroin)) and three urbanized areas (“S3-U” (Sint-Katelijne-Waver), “S4-U” (Leuven) and “S5-U” (Leuven)) in Belgium (S1 Table). Sites were separated by a mean of 46 km (3.0–160 km). At the time of sampling, very few plant species such as Narcissus pseudonarcissus and Helleborus foetidus were flowering at the forest locations while Erica carnea and a few other species were flowering in gardens of the urban areas that were sampled. In total, ten individual queens were collected per location, except for forest location S1-F where eight queens were collected. Specimens were individually put in plastic 50 ml vials and transported in a cooling box to the laboratory to avoid any mortality incidences by stress.
Gut dissection, DNA extraction, PCR amplification and Illumina MiSeq analysis
Following rinsing with 70% ethanol, each specimen (alive, fresh specimens) was pinned to a polyacrylamide gel plate and immersed in sterile Ringer’s solution. Subsequently, the intestines were pulled out and the midgut and hindgut excluding rectum (i.e. the ileum region) were collected into a vial with 1 ml glycerol (40%) and homogenized by using zirconia beads and a Fast-Prep24 Instrument (MP Biomedicals, USA). The dissections were conducted under a binocular microscope (M420 Wild Makroskop, Switzerland). After every dissection, Ringer’s solution was replaced and the gel plate was cleaned and desinfected with 70% ethanol. Next, guts were crushed in 170 μl lysozyme solution (100 mg/ml) and DNA was extracted according to Meeus et al. . A negative control was included during extraction in which the lysozyme solution without gut material was used as starting material. DNA samples were then subjected to PCR amplification and Illumina MiSeq sequencing. Again a negative control was included (PCR amplification control), this time by replacing template DNA with sterile water. Results obtained for both types of negative controls were satisfactory, confirming that the experimental conditions were met to achieve robust data. DNA samples were amplified using sample-specific barcode-labeled versions of the primer sets 515F / 806R  and ITS86F / ITS4 , generating amplicons covering the V4 region of the bacterial 16S rRNA gene and the fungal internal transcribed spacer-2 (ITS-2) region, respectively (dual-index sequencing strategy ; S2 Table). Amplification was performed in a volume of 40 μl containing 1x Titanium Taq PCR buffer, 150 μM of each dNTP, 0.5 μM of each primer, 1x Titanium Taq DNA polymerase (Clontech, Saint-Germain-en-Laye, France) and 2 μl DNA (5 ng μl-1), using the following thermal profile: initial denaturation at 94°C for 2 min, followed by 30 cycles of denaturation at 94°C for 45 s, annealing at 59°C for 45 s and elongation at 72°C for 45 s, and terminated by a final elongation at 72°C for 10 min. PCR amplicons were then purified using Agencourt AMPure XP magnetic beads (Beckman Coulter Genomics GmbH, South Plainfield, UK) according to the manufacturer’s guidelines. Following quantification of the purified products using a Qubit High Sensitivity Fluorometer kit (Invitrogen, Carlsbad, CA, USA) amplicons were combined at equimolar concentrations into two amplicon libraries (one for bacteria and one for fungi). Subsequently, the libraries were subjected to an ethanol precipitation and loaded on an agarose gel. Next, the target bands (c. 250 bp for bacteria, c. 400 bp for fungi) were excised and the DNA was purified and measured again. Amplicon mixtures were then combined, diluted to 2 nM and sequenced using an Illumina MiSeq sequencer with v2 500 cycle reagent kit (Illumina, San Diego, CA, USA).
Sequences were received as a de-multiplexed FASTQ file. Paired-end reads were merged with a maximum of five mismatches using USEARCH (v10.0.240) to form consensus sequences  and truncated at the 250th base. Shorter reads or reads with a total expected error threshold above 0.05 (for 16S rRNA gene sequences) or above 0.10 (for ITS) were discarded using VSEARCH (v2.4.0) . Next, in order to remove potential contaminant sequences, sequences were classified with the Mothur (v1.36.1) implementation of the Silva database (v1.23) (for 16S rRNA gene sequences) and the RDP Warcup fungal ITS training set (v2)  (for ITS sequences) using the “classify.seqs” command. Subsequently, 16S rRNA gene sequences identified as “mitochondria,” “chloroplast,” “archaea,” “eukaryota,” or “unknown” were removed using the “remove.lineage” command. For ITS sequences, only sequences that could be identified to the fungal family level with a bootstrap confidence value of 100% were retained using the “get.lineage” command. Next, the number of sequences was rarefied to that of the sample with the lowest number of reads (c. 23,350 for bacteria and 2,000 for fungi), and sequences were grouped into operational taxonomic units (OTUs) based on a 3% sequence dissimilarity cut-off using the UPARSE greedy algorithm in USEARCH, during which chimeric sequences were also removed , as were global singletons (i.e. OTUs with only 1 sequence in the entire data set). The taxonomic origin of each OTU was determined with the SINTAX algorithm implemented in USEARCH , based on the Silva v1.23 database  and the RDP Warcup fungal ITS training set (v2)  for bacteria and fungi, respectively. In general, taxonomic assignments can be considered reliable when bootstrap confidence values exceed 0.80. Furthermore, BLAST searches were performed against type materials in GenBank verifying the identity of the most important OTUs. Additionally, for core bacteria, identifications were refined up to phylotypes by available information in the literature and GenBank [26, 27]. Raw sequence data were deposited in the Sequence Read Archive under BioProject accession PRJNA445658. Representative OTU sequence data (selected by UPARSE) were deposited in Genbank under the accession MH815148 to MH816704.
qPCR to estimate bacterial and fungal abundance
Quantitative real-time PCR (qPCR) was used to estimate total bacterial and fungal abundance in the gut samples, as well as the abundance of the core bacteria Gilliamella and Snodgrasella. Specifically, the universal bacterial primers 519F and 907R were used to amplify total copies of the16S rRNA gene. Additionally, Beta-1009-qtF and Beta-1115-qtR  and LWI1_1 (5’-TAC GGA GGG TGC GAG CGT T-3’) and Gamma1-648-qtR  were used to amplify the 16S rRNA genes of Snodgrassella and Gilliamella, respectively. Further, ITS86F and ITS4 were used to amplify total fungal ITS-2 copies. qPCR amplifications were performed in MicroAmp Fast 8-Tube Strips (Life Technologies, Carlsbad, CA, USA) using a StepOnePlus real-time PCR system (Applied Biosystems, Carlsbad, CA, USA), and each reaction was performed in duplicate. Each reaction contained 1.0 μl DNA (5 ng), 10.0 μl of the iTaq Universal SYBRGreen supermix (Bio-Rad, Hercules, CA, USA), 0.2 μl of each primer (20 μM) and 8.6 μl sterile water. Thermal cycling conditions consisted of 2 min at 95°C, followed by 40 amplification cycles of 15 s at 95°C, 30 s at 59°C (519F/907R; ITS86F/ITS4) or 55°C (Beta-1009-qtF/Beta-1115-qtR; LWI1_1/Gamma1-648-qtR) and 30 s at 60°C. Fluorescence (520 nm) was detected at the end of the elongation phase for each cycle. To evaluate amplification specificity, a melting curve analysis was performed at the end of each PCR run. In each analysis, a positive and negative control (template DNA replaced by sterile water) was included. Quantification was based on standard curves from amplification of cloned target sequences in a TOPO-TA vector (Invitrogen).
In order to evaluate the prevalence of N. bombi and C. bombi, qPCR was performed using the primers developed by Huang et al. . Amplifications were performed as described above, with the exception of the annealing temperature which was 64.5°C. Each reaction was performed in duplicate, and in each analysis, a positive (pathogen DNA) and negative control (template DNA replaced by sterile water) was included. Additionally, six gut samples from commercially reared B. terrestris queens (Biobest Group, Westerlo, Belgium), that were known to be free of pathogens (screened by microscopy), were included as a negative control, and were also found to be free of pathogens using the qPCR (S1 Table). As previous analyses had shown that the primers may yield aspecific products (data not shown), all amplicons obtained were sequenced to confirm the identity of the pathogens. A sample was considered positive when both the CT value was below that of the negative control (sterile water) and amplicon sequencing confirmed pathogen identity.
Rarefaction curves were constructed for the bacterial and fungal communities using the Vegan package (v2.4–6) for R (R Development Core Team, 2013). Additionally, for each sample, OTU richness (S), the Chao1 estimator, Shannon-Wiener diversity (H) and Pielou’s evenness (J = H/ln(S)) were calculated using USEARCH (v10.0.240)  and compared using a mixed model analysis of variance with habitat type (forest or urban environment) as fixed factor and population as random factor. To this end, Shannon diversity was exponentially transformed (exp(H)) to obtain a linear variable . Non-metric multidimensional scaling (NMDS) was used to visualize the level of similarity in community composition between the different samples based on Bray-Curtis similarities (relative abundance data). Further, permutational multivariate analysis of variance (PERMANOVA) using the adonis function in the vegan package was performed to test for significant differences in microbial gut community composition between the different locations.
To test the hypothesis that disease prevalence was associated with habitat type, we used a logistic regression analysis, with habitat type as fixed factor and presence/absence of the disease as dependent variable. Analyses were performed for each disease separately. Secondly, we investigated whether disease occurrence was related to gut microbial diversity. Finally, we tested the hypothesis that pathogen occurence was related to the abundance of core bacteria in the gut that were previously suggested to protect the bees against gut pathogens and parasites [24, 26, 30]. We used logistic regression analyses with the relative abundance of the two main core gut bacteria (Snodgrassella and Gilliamella) as independent variable and the presence/absence of N. bombi and C. bombi as dependent variable. Analyses were performed for each pathogen separately.
Bacterial community composition
Analysis of the gut microbiota revealed a total of 1,450 bacterial OTUs belonging to diverse phyla (among which the most abundant were Acidobacteria, Bacteroidetes, Firmicutes and Proteobacteria) (S3 Table and S1 Fig). Based on Chao1, the mean sampling coverage was 86.8% (S1 Table), suggesting that the most abundant bacterial community members were covered, as can also be observed from the rarefaction curves that approached saturation (S2 Fig). Observed OTU richness varied between 14 and 589 bacterial OTUs (mean of 239 OTUs) per bumblebee in the forest locations, while sampled individuals from the urbanized locations contained between 4 and 153 bacterial OTUs (mean of 56 OTUs) (S1 Table). Furthermore, gut microbial communities were more evenly distributed (F = 1.87, P = 0.029) for bumblebee queens from the forest locations (Table 1; S1 Table). Results from the qPCR analysis indicated that log transformed total bacterial 16S rRNA gene copy numbers were signifcantly higher for the urban specimens (F = 1.63, P = 0.018). Gene copy numbers in the forest populations were 1 to 2 log orders lower compared to the urban populations (Fig 1).
Bumblebee queens (Bombus terrestris) were collected from five different locations, representing two habitat types, including forest (S1-F and S2-F) and urbanized habitats (S3-U, S4-U and S5-U). No bars represent concentrations below the limit of detection (100 copies).
NMDS ordination of the gut bacterial community composition showed no clear separation among the different sampled areas (Bray-Curtis; stress = 0.16; Fig 2A). Indeed, also adonis analysis revealed no significant differences in bacterial community composition among queens from the different sites (n = 5; R = 0.259, P = 0.055). By contrast, significant differences were found between habitat types (n = 2; R = 0.957, P = 0.034). Gut bacterial communities in queens collected from the forest habitats showed a lower relative abundance of the core bacteria Snodgrassella (OTU1; Beta) (P = 0.041) and Gilliamella (OTU2; Gamma-1) (P = 0.011) (Fig 3), as was also confirmed by absolute quantifications using qPCR (data not shown). While being found in every specimen investigated, relative abundance of Snodgrassella and Gilliamella varied between 0.01 and 55.0% (mean: 5.4%) and between 0.03 and 94.1% (mean: 17.9%) for the forest specimens, while this was between 0.01 and 74.5% (mean: 35.9%) and between 0.03 and 78.7% (mean: 32.4%) for the urban specimens, respectively (Fig 3). When both taxa were taken together, mean relative abundance varied between 3.5% for the forest bees and 34.1% for the urban bee populations. Instead, forest specimens contained a higher relative amount of other bacteria, particularly members of Bacilaceae, Flavobacteriaceae, Planococcaceae and Pseudomonaceae (Fig 3; S3 Fig). For example, guts of forest specimens were particularly enriched with OTUs corresponding to Bacillus niacini (OTU3; Bacilaceae), Cellvibrio sp. (OTU53; Pseudomonaceae), Flavobacterium sp. (e.g. OTU30, 48 and 130; Flavobacteriaceae), Pseudomonas sp. (e.g. OTU 263, 326 and 1438; Pseudomonaceae), and Sporosarcina sp. (OTU12; Planococcaceae). Furthermore, an Apibacter OTU (Apibacter mensalis; OTU16; Flavobacteriaceae) represented a considerable fraction of the reads in some forest specimens (Fig 3; S3 Table).
Non-metric multidimensional scaling (NMDS) based on Bray-Curtis dissimilarities of the gut bacterial (A) (stress value = 0.16) and fungal community composition (B) (stress value = 0.21) of Bombus terrestris queens from five different locations. Sampled locations represent two habitat types, including forest (S1-F (yellow) and S2-F (orange)) and urbanized habitats (S3-U (purple), S4-U (dark blue) and S5-U (light blue)). The distance between different points on the plot reflects their similarity level: the more similar the communities, the smaller the distance between the points.
Sampled locations represent two habitat types, including forest (S1-F and S2-F) and urbanized habitats (S3-U, S4-U and S5-U). Only the most abundant OTUs (i.e. with a mean sequence relative abundance > 1% over the entire dataset) are represented in the figure.
Lactobacilli were found in 37 of the 48 investigated specimens (14 forest and 23 urban specimens). Nevertheless, despite their common occurrence, OTUs belonging to Lactobacillaceae occured at low relative abundance. Among these, the Lactobacillus phylotypes Firm-4/Lacto-2 (Lactobacillus bombi; OTU6; found in 21 specimens), Lacto-5 (Lactobacillus sp.; OTU20; found in 21 specimens) and Firm-5/Lacto-1 (Lactobacillus bombicola; OTU23; found in 16 specimens) were most prevalent (Table 2). Additionally, an OTU corresponding to Lactobacillus iners (OTU136) was frequently found (detected in 9 forest and 5 urban specimens). Further, a number of other lactobacilli were detected, albeit more eradically (Table 2). Likewise, bifidobacteria were sporadically detected, and were found to have low relative abundances (range: 0.001–2.0%). Specifically, we found bifidobacteria in 4 forest specimens (all from S1-F) and 11 urban specimens, with phylotype Bifido-3 (OTU24; Bombiscardovia coagulans) as the most prevalent OTU (present in 7 specimens) (Table 2). The phylogenetic position of the lactobacilli and bifidobacteria found in this study with regard to their closest family members is shown in S4 and S5 Figs, respectively. The Enterobacteriaceae core phylotypes Gamma-E1 (OTU9; Buttiauxella sp.) and Gamma-E2 (OTU19; Hafnia sp.) were commonly detected in our study. More particularly, the Gamma-E1 OTU was found in every specimen with the exception of two forest and two urban specimens. Phylotype Gamma-E2 was especially found in the forest specimens (17 out of 18 forest specimens versus 18 out of 30 urban specimens) (Table 2).
Fungal community composition
Whereas bacterial OTUs were found in every specimen, fungal taxa (both true fungi and yeasts) were only detected in B. terrestris queens from the forest locations S1-F and S2-F (fungi present in every specimen analyzed), and in two individuals from site S3–U (S1 Table). Illumina sequencing revealed a total of 107 fungal OTUs belonging to three phyla, including Ascomycota, Basidiomycota and Zygomycota. No differences were observed in OTU richness, evenness and community structure between bees from both forest locations (P = 0.063) (Fig 2B; S6 Fig), and a mean fungal richness of 15 OTUs per specimen was recorded (range between 6 and 25 OTUs) (S1 Table). The most prevalent fungi and yeasts detected included widespread taxa such as Saccharomyces (Ascomycota; Saccharomycetaceae), Trichoderma (Ascomycota; Hypocreaceae) and Mucor (Zygomycota; Mucoraceae) (S4 Table).
Pathogen assessment and association with habitat type and specific bacteria
Pathogen screening using qPCR revealed the presence of N. bombi and C. bombi in 4 (8.3%) and 20 (43.8%) of the 48 investigated queens, repectively. Two specimens were found to contain both pathogens (S1 Table). Nosema was only found in forest specimens. Further, incidence of Crithidia was significantly higher (P = 0.03) in specimens from the forest locations than from urbanized locations (Fig 4). The occurrence of both pathogens in the gut of bumblebees was not significantly (P > 0.05) related to bacterial diversity in the gut. On the other hand, presence of Nosema was negatively and significantly (P < 0.001) related to the relative abundance of Snodgrassella (OTU1) (Fig 5).
Sampled locations represent two habitat types, including forest (S1-F and S2-F) and urbanized habitats (S3-U, S4-U and S5-U).
Knowledge about the health of bees and bumblebees is important as they are among the most important pollinators for many native plants as well as several crops [42, 43]. Wild bee and bumblebee populations have been in a steady decline worldwide , and among several factors, land-use change and increased pathogen prevalence have been proposed to contribute to bee decline [8, 9, 44]. As anthropogenic use of land is likely to increase, we investigated the impact of urbanization on the gut microbial community composition of wild B. terrestris queens, and investigated whether relationships could be established with pathogen infection.
Gut microbial community analysis revealed several bacterial taxa known to be associated with bees and bumblebees. Snodgrassella, Gilliamella, Lactobacillus Firm-4/Lacto-2 and Firm-5/Lacto-1 and Bifidobacterium have been described as core bacteria of Apis and Bombus hosts [26, 27, 45, 46]. The former two were also highly prevalent in the B. terrestris queens investigated in this study, confirming their strong association with (bumble)bees [27, 47]. These microbes are vertically transmitted from a mother colony to the offspring and have not been detected outside of bees . Additionally, the bee gut symbiotic lactobacilli Firm-4/Lacto-2, Firm-5/Lacto-1 and Lacto-5 were commonly detected next to a number of environmental lactobacilli with a more erratic occurrence. Recently, Bifidobacteria were proposed as core bacteria in B. terrestris . However, our data does not seem to support this scenario. Bifidobacteria occurred at low relative abundance, and were only detected in 15 out of the 48 investigated queens. Most probably, the low prevalence and low relative abundance of bifidobacteria can be explained by the fact that we focused on midgut and ileum, while other studies generally investigated the microbiome of whole guts, including rectum. Previous research has shown that the midgut of social bees only contains few bacteria, while the ileum and rectum are strongly colonized by bacteria, totaling up to 108 and 109 bacterial cells, respectively . Furthermore, while the ileum is dominated by Snodgrassella, Gilliamella, and the lactobacilli Firm-4/Lacto-2 and Firm-5/Lacto-1, the rectum is dominated by lactobacilli and bifidobacteria . Therefore, as the rectum was not taken into account in our study, this may explain the low prevalence and abundance of bifidobacteria. Additionally, it cannot be excluded that age differences contributed to this discrepancy , but further research is needed to confirm this. By contrast, the Enterobacteriaceae OTUs corresponding to Gamma-E1 and Gamma-E2 had a high prevalence (present in 44 and 35 specimens, respectively), corroborating previous results .
Impact of habitat type
The gut microbial community composition from B. terrestris queens differed strongly between bumblebees from forests and urbanized habitats. First, fungi were almost exclusively associated with bumblebee queens from forests. Within these fungal communities, Saccharomyces yeasts were dominant (mean average relative abundance of 17.3%) and occurred in every specimen investigated. In previous studies, insect-associated yeasts such as Saccharomyces were reported to play an important role in food digestion and detoxification of toxic plant metabolites in the insect host . Further, environmental fungi like Trichoderma that are commonly encountered in forest soils and airborne fungi like Mucor and Penicillium were found. So far it is not clear why fungi were strongly associated with forest specimens, but lower bacterial concentrations in these samples may (partially) explain this observation (less competition). Secondly, there was a striking difference in gut bacterial community composition. Bacterial richness was highest for the bumblebee queens occurring in the forest locations (on average 239 OTUs per specimen); queens from the urbanized areas contained c. 75% less bacterial OTUs (on average 56 OTUs per specimen). Gut communities of the latter were predominantly composed of the core bacteria Snodgrassella and Gilliamella, representing a combined relative abundance up to 68.2% per specimen. Forest specimens had a lower relative abundance of both bacteria, as were absolute levels. Instead, forest specimens harboured a huge diversity of non-core, environmental bacteria. Recent research has shown that flowers represent an important hot spot of transmission of environmental microbes to bee guts , and it is reasonable to assume that the likelihood of transmitting novel microbes to the bee gut is higher in natural environments. Natural habitats like forest environments are generally characterized by dense patches of flowers that are frequently visited and shared by diverse insects , facilitating transmission of new and additional microbes to the gut microbiome, e.g. by ingestion of microbe-contaminated nectar or pollen [23, 53, 54]. Guts of forest specimens were enriched with opportunistic bacteria like pseudomonads and Flavobacteriacae of which several members have been found in flowers [54–56]. Several Pseudomonaceae and Flavobacteriaceae OTUs showed high 16S rRNA gene sequence identity (up to 100%) with strains that have been found in nectar or on pollen (data not shown). At the time of sampling, however, there were not many plant species flowering, and flowers were not investigated for microbial presence. Further investigations are needed to explain the differences in microbial gut community composition between both types of habitat.
So far, only very little is known about the functional potential of these environmental microbes in the bee gut. However, in a recent study it was found that several of such microorgansms may have antimicrobial activity and can help protect against microbial pathogens and parasites . Further research is needed to unravel the possible functions of these bacteria, and to find out whether or not they impact on bee fitness. However, it has to be noted that our data do not imply that all opportunistic taxa found are able to replicate and to stably colonize the gut. Possibly they are dead microbes consumed with pollen or nectar, or microorganisms that became inactivated in the harsh environment of the bee gut. Further, it is unknown whether the investigated bumblebee specimens had the same age or experienced different levels of stress, two factors that may affect gut communities .
Association with pathogen infection
Gut microbiota are critical for the health of many insect species . In this regard, overall microbiota richness has been assumed to be important in microbial gut function. For example, in locusts it has been demonstrated that increasing diversity of gut bacteria reduces susceptibility to a pathogen . Likewise, using experimental transplantations in Bombus impatiens, Mockler et al.  showed that lower Crithidia infection loads were associated with high microbiome diversity and large gut bacterial populations. Similar results were obtained by Praet et al.  although bacterial gut isolates from wild bumblebees were screened in vitro in pectin degradation assays and pathogen growth inhibition assays. By contrast, Koch et al.  found in their in vivo assays that microbiota richness was positively associated with Crithidia infection in B. terrestris. Our results, however, did not show a significant relationship between pathogen infection and microbial diversity. On the other hand, we found a significant and negative relationship between Nosema prevalence and relative abundance of the core resident Snodgrassella. This is in line with previous research that has shown Snodgrassella to be negatively associated with N. bombi in B. terrestris . Potentially, this can be explained by the fact that Snodgrassella (as well as Gilliamella) is able to form biofilm-layers on the host epithelium of the gut, by which pathogen infection may be restricted . In contrast to other studies [24, 26, 30, 61], no relationship was found between the prevalence of cre gut residents and Crithidia infection.
Overall, we observed a striking difference between the gut communities within the midgut and ileum of B. terrestris queens living in urban environments compared to those in forest environments, both for fungi and bacteria. Furthermore, our study provides evidence that the core resident Snodgrassella may have a protective effect against pathogens, and suggests that pathogens may be more prevalent in specimens from natural environments, which may be contradictory to the general assumption of a forest as a source of “pristine”, healthy specimens. Further research with more locations, however, is needed to draw strong conclusions regarding the effect of urbanization on the structure of bumblebee gut commities and bumblebee health and fitness. Possible factors include: landscape structure, pesticide exposure, quantity/ quality and connectivity of food sources, exposure to other arthropod fauna, and other symbiont or pathogen transmission routes. Furthermore, it remains to be investigated whether the same trends will be observed when the rectum is taken into account. Also, additional research using microbial isolates is needed to unravel the precise function of the environmental opportunistic microbes found, and to assess their role in bee health and overall bee fitness.
S1 Table. Overview of samples (Bombus terrestris queens) investigated in this study, including sampling information and diversity measures.
S2 Table. Primer design and sample-specific barcodes.
S3 Table. Identification of bacterial operational taxonomic units (OTUs) according to the Silva v1.23 database and distribution over the investigated samples.
S4 Table. Identification of fungal operational taxonomic units (OTUs) according to the RDP Warcup fungal ITS training set (v2) and distribution over the investigated samples.
S1 Fig. Relative abundance (%) of the bacterial phyla found in the guts (midgut and ileum) of bumblebee queens (Bombus terrestris) from five different locations.
Sampled locations represent two habitat types, including forest (S1-F and S2-F) and urbanized habitats (S3-U, S4-U and S5-U).
Rarefaction curves showing the number of gut bacterial (A) and fungal (B) operational taxonomic units (OTUs) per bumblebee queen (Bombus terrestris) from five different locations. Sampled locations represent two habitat types, including forest (S1-F (yellow) and S2-F (orange)) and urbanized habitats (S3-U (purple), S4-U (dark blue) and S5-U (light blue)). Rarefaction curves reached saturation, suggesting that the most abundant community members were covered by our sequencing depth.
S3 Fig. Relative abundance (%) of bacterial families in the guts (midgut and ileum) of bumblebee queens (Bombus terrestris) from five different locations.
Sampled locations represent two habitat types, including forest (S1-F and S2-F) and urbanized habitats (S3-U, S4-U and S5-U).
S4 Fig. Maximum-likelihood tree derived from the V4 region of the 16S rRNA gene sequence (250 bp) of Lactobacillaceae occurring in the digestive tract of Bombus and Apis, showing the phylogenetic position of the lactobacilli found in this study (OTU6, 20, 23, 37, 136, 189, 194, 282, 729, 1304, 1601 and 1697).
Type strains of the closest relatives were also included in the tree.
S5 Fig. Maximum-likelihood tree derived from the V4 region of the 16S rRNA gene sequence (250 bp) of Bifidobacteriaceae occurring in the digestive tract of Bombus, showing the phylogenetic position of the bifidobacteria found in this study (OTU24, 38 and 786).
Type strains of the closest relatives were also included in the tree.
S6 Fig. Fungal community composition at the level of operational taxonomic units (OTUs) within the midgut and ileum in bumblebee queens (Bombus terrestris) having fungi.
Specimens with fungi were caught in the forest habitats S1-F and S2-F (18 specimens) and the urbanized habitat S3-U (2 specimens). All other investigated bumblebees were negative for fungi. Only the most abundant OTUs (i.e. with a mean sequence relative abundance > 1% over the entire dataset) are represented in the figure.
- 1. Gaston KJ. The magnitude of global insect species richness. Conservation Biology. 1991; 5:283–296.
- 2. Mora C, Tittensor DP, Adl S, Simpson AG, Worm B. How many species are there on Earth and in the ocean? PLoS Biology. 2011; 9:e1001127. pmid:21886479
- 3. Yang LH, Gratton C. Insects as drivers of ecosystem processes. Current Opinion in Insect Science. 2014; 2:26–32.
- 4. Weisser WW, Siemann E. The various effects of insects on ecosystem functioning. In Insects and ecosystem function (Springer, Berlin, Heidelberg). 2008; 1:3–24.
- 5. Noriega JA, Hortal J, Azcárate FM, Berg MP, Bonada N, Briones MJ, et al. Research trends in ecosystem services provided by insects. Basic and Applied Ecology. 2018; 26:8–23.
- 6. Hallmann CA, Sorg M, Jongejans E, Siepel H, Hofland N, Schwan H, et al. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLoS One. 2017; 12:e0185809.
- 7. Lebuhin G, Droege S, Connor EF, Gemmill‐Herren B, Potts SG, Minckley RL, et al. Detecting insect pollinator declines on regional and global scales. Conservation Biology. 2013; 27:113–120. pmid:23240651
- 8. Goulson D, Nicholls E, Botías C, Rotheray EL. Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science. 2015; 347:1255957. pmid:25721506
- 9. Meeus I, Pisman M, Smagghe G, Piot N. Interaction effects of different drivers of wild bee decline and their influence on host-pathogen dynamics. Current Opinion in Insect Science. 2018; 26:136–141. pmid:29764653
- 10. DeFries RS, Foley JA, Asner GP. Land‐use choices: Balancing human needs and ecosystem function. Frontiers in Ecology and the Environment. 2004; 2:249–257.
- 11. Newbold T, Hudson LN, Hill SL, Contu S, Lysenko I, Senior RA, et al. Global effects of land use on local terrestrial biodiversity. Nature. 2015; 520:45. pmid:25832402
- 12. Hammer TJ, Janzen DH, Hallwachs W, Jaffe SP, Fierer N. Caterpillars lack a resident gut microbiome. Proceedings of the National Academy of Sciences. 2017; 114:9641–9646.
- 13. Evans JD, Lopez DL. Bacterial probiotics induce an immune response in the honey bee (Hymenoptera: Apidae). Journal of Economic Entomology. 2004; 97:752–756. pmid:15279248
- 14. Chow J, Lee SM, Shen Y, Khosravi A, Mazmanian SK. Host–bacterial symbiosis in health and disease. Advances in Immunology. 2010; 107:243–274. pmid:21034976
- 15. Kikuchi Y, Hayatsu M, Hosokawa T, Nagayama A, Tago K, Fukatsu T. Symbiont-mediated insecticide resistance. Proceedings of the National Academy of Sciences. 2012; 109: 8618–8622.
- 16. Hanning I, Diaz-Sanchez S. The functionality of the gastrointestinal microbiome in non-human animals. Microbiome. 2015; 3:51. pmid:26552373
- 17. Praet J, Parmentier A, Schmid‐Hempel R, Meeus I, Smagghe G, Vandamme P. Large‐scale cultivation of the bumblebee gut microbiota reveals an underestimated bacterial species diversity capable of pathogen inhibition. Environmental Microbiology. 2018; 20:214–227. pmid:29076622
- 18. Meeus I, Mommaerts V, Billiet A, Mosallanejad H, Van de Wiele T, Wäckers F, Smagghe G. Assessment of mutualism between Bombus terrestris and its microbiota by use of microcolonies. Apidologie. 2013; 44:708–719.
- 19. Billiet A, Meeus I, Cnockaert M, Vandamme P, Van Oystaeyen A, Wäckers F, et al. Effect of oral administration of lactic acid bacteria on colony performance and gut microbiota in indoor-reared bumblebees (Bombus terrestris). Apidologie. 2017; 48:41–50.
- 20. Raymann K, Moran NA. The role of the gut microbiome in health and disease of adult honey bee workers. Current Opinion in Insect Science. 2018;
- 21. Nicholson JK, Holmes E, Kinross J, Burcelin R, Gibson G, Jia W, et al. Host-gut microbiota metabolic interactions. Science. 2012; 1223813.
- 22. Graystock P, Goulson D, Hughes WO. Parasites in bloom: flowers aid dispersal and transmission of pollinator parasites within and between bee species. Proceedings of the Royal Society B. 2015; 282:20151371. pmid:26246556
- 23. McFrederick QS, Thomas JM, Neff JL, Vuong HQ, Russell KA, Hale AR, et al. Flowers and wild megachilid bees share microbes. Microbial Ecology. 2017; 73:188–200. pmid:27592345
- 24. Koch H, Schmid-Hempel P. Bacterial communities in central European bumblebees: low diversity and high specificity. Microbial Ecology. 2011; 62:121–133. pmid:21556885
- 25. Martinson VG, Danforth BN, Minckley RL, Rueppell O, Tingek S, Moran NA. A simple and distinctive microbiota associated with honey bees and bumble bees. Molecular Ecology. 2011; 20:619–628. pmid:21175905
- 26. Cariveau DP. Powell JE, Koch H, Winfree R, Moran NA. Variation in gut microbial communities and its association with pathogen infection in wild bumble bees (Bombus). The ISME Journal. 2014; 8:2369–2379. pmid:24763369
- 27. Meeus I, Parmentier L, Billiet A, Maebe K, Van Nieuwerburgh F, Deforce D, et al.16S rRNA amplicon sequencing demonstrates that indoor-reared bumblebees (Bombus terrestris) harbor a core subset of bacteria normally associated with the wild host. PLoS One. 2015; 10:e0125152. pmid:25923917
- 28. Baer B, Schmid-Hempel P. Experimental variation in polyandry affects parasite loads and fitness in a bumble-bee. Nature. 1999; 397:151.
- 29. Colla SR, Otterstatter MC, Gegear RJ, Thomson JD. Plight of the bumble bee: pathogen spillover from commercial to wild populations. Biological Conservation. 2006; 129:461–467.
- 30. Mockler BK, Kwong WK, Moran NA, Koch H. Microbiome structure influences infection by the parasite Crithidia bombi in bumble bees. Applied and Environmental Microbiology. 2018; 84:e02335–17.
- 31. Kleijn D, Winfree R, Bartomeus I, Carvalheiro LG, Henry M, Isaacs R, et al. Delivery of crop pollination services is an insufficient argument for wild pollinator conservation. Nature Communications. 2015; 6:7414. pmid:26079893
- 32. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Lozupone CA, Turnbaugh PJ, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proceedings of the National Academy of Sciences of the United States of America. 2011; 108:4516–4522. pmid:20534432
- 33. Turenne CY, Sanche SE, Hoban DJ, Karlowsky JA, Kabani AM. Rapid identification of fungi by using the ITS2 genetic region and an automated fluorescent capillary electrophoresis system. Journal of Clinical Microbiology. 1999; 37:1846–1851. pmid:10325335
- 34. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Applied and Environmental Microbiology. 2013; 79:5112–5120. pmid:23793624
- 35. Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nature Methods. 2013; 10:996–998. pmid:23955772
- 36. Rognes T, Flouri T, Nichols B, Quince C, Mahé F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016; 4:e2584. pmid:27781170
- 37. Deshpande V, Wang Q, Greenfiel P, Charleston M, Porras-Alfaro A, Kuske CR, et al. Fungal identification using a Bayesian classifier and the Warcup training set of internal transcribed spacer sequences. Mycologia. 2016; 108:1–5. pmid:26553774
- 38. Edgar R. SINTAX: a simple non-Bayesian taxonomy classifier for 16S and ITS sequences. BioRxiv. 2016; 074161.
- 39. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research. 2012; 41:D590–D596. pmid:23193283
- 40. Huang WF, Skyrm K, Ruiter R, Solter L. Disease management in commercial bumble bee mass rearing, using production methods, multiplex PCR detection techniques, and regulatory assessment. Journal of Apicultural Research. 2016; 54:516–524.
- 41. Jost L. Entropy and diversity. Oikos. 2006; 113:363–375.
- 42. Williams IH. The dependence of crop production within the European Union on pollination by honey bees. Agricultural Zoology Reviews (United Kingdom), 1994; 6:229–257.
- 43. Klein AM, Vaissiere BE, Cane JH, Steffan-Dewenter I, Cunningham SA, Kremen C, et al. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society B. 2007; 274:303–313. pmid:17164193
- 44. Cameron SA, Lozier JD, Strange JP, Koch JB, Cordes N, Solter LF, et al. Patterns of widespread decline in North American bumble bees. Proceedings of the National Academy of Sciences. 2011; 108:662–667.
- 45. Moran NA, Hansen AK, Powell JE, Sabree ZL. Distinctive gut microbiota of honey bees assessed using deep sampling from individual worker bees. PLoS One. 2012; 7:e36393. pmid:22558460
- 46. Li J, Powell JE, Guo J, Evans JD, Wu J, Williams P, et al. Two gut community enterotypes recur in diverse bumblebee species. Current Biology. 2015; 25:R652–R653. pmid:26241138
- 47. Powell E, Ratnayeke N, Moran NA. Strain diversity and host specificity in a specialized gut symbiont of honeybees and bumblebees. Molecular Ecology. 2016; 25:4461–4471. pmid:27482856
- 48. Koch H, Cisarovsky G, Schmid‐Hempel P. Ecological effects on gut bacterial communities in wild bumblebee colonies. Journal of Animal Ecology. 2012; 81:1202–1210. pmid:22708631
- 49. Kwong WK, Moran NA. Gut microbial communities of social bees. Nature Reviews Microbiology. 2016; 14:374–3840. pmid:27140688
- 50. Anderson KE, Ricigliano VA, Mott BM, Copeland DC, Floyd AS, Maes P. The queen’s gut refines with age: longevity phenotypes in a social insect model. Microbiome. 2018; 6:108. pmid:29914555
- 51. Vega FE, Dowd PF. The role of yeasts as insect endosymbionts. Insect-Fungal Associations: Ecology and Evolution. Oxford University Press, New York, USA. 2005; pp. 211–243.
- 52. Pozo MI, Herrera M, Alonso C. Spatial and temporal distribution patterns of nectar-inhabiting yeasts: how different floral microenvironments arise in winter-blooming Helleborus foetidus. Fungal Ecology. 2014; 11:173–180.
- 53. Hani B, Dalila B, Saliha D, Daoud H, Mouloud G, Seddik K. Microbiological sanitary aspects of pollen. Advances in Environmental Biology. 2012; 6:1415–1420.
- 54. Pozo MI, Lievens B, Jacquemyn H. Impact of microorganisms on nectar chemistry, pollinator attraction and plant fitness. In: Peck RL, editor. Nectar: production, chemical composition and benefits to animals and plants. Nova Science Publishers, Inc.; New York, USA. 2015; pp. 1–40.
- 55. Lievens B, Hallsworth JE, Pozo MI, Belgacem ZB, Stevenson A, Willems KA, et al. Microbiology of sugar‐rich environments: diversity, ecology and system constraints. Environmental Microbiology. 2015; 17:278–298. pmid:25041632
- 56. Manirajan B, Ratering S, Rusch V, Schwiertz A, Geissler‐Plaum R, Cardinale M, et al. Bacterial microbiota associated with flower pollen is influenced by pollination type, and shows a high degree of diversity and species specificity. Environmental Microbiology. 2016; 18:5161–5174. pmid:27612299
- 57. Tarpy DR, Mattila HR, Newton IL. Development of the honey bee gut microbiome throughout the queen-rearing process. Applied and Environmental Microbiology. 2015; 81:3182–3191. pmid:25724964
- 58. Engel P, Moran NA. The gut microbiota of insects–diversity in structure and function. FEMS Microbiology Reviews. 2013; 37:699–735. pmid:23692388
- 59. Dillon RJ, Vennard CT, Buckling A, Charnley AK. Diversity of locust gut bacteria protects against pathogen invasion. Ecology Letters. 2005; 8:1291–1298.
- 60. Wintermantel D. The effect of field-level clothianidin exposure on the prevalence and loads of pathogenic and non-pathogenic microbes in bumblebees (B. terrestris). Swedisch University of Agricultural Science. 2015.
- 61. Palmer-Young EC, Raffel TR, McFrederick QS. pH-mediated inhibition of a bumble bee parasite by an intestinal symbiont. BioRxiv. 2018; http://dx.doi.org/10.1101/336347.