See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/228779871

Bacterial communities in floral

Article in Environmental Microbiology Reports · February 2012 DOI: 10.1111/j.1758-2229.2011.00309.x

CITATIONS READS 127 1,534

4 authors, including:

Ido Izhaki Yoram Gerchman University of Haifa Oranim college and University of Haifa, Israel

358 PUBLICATIONS 5,587 CITATIONS 188 PUBLICATIONS 2,930 CITATIONS

SEE PROFILE SEE PROFILE

Malka Halpern University of Haifa

187 PUBLICATIONS 2,694 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

“Understanding the ecology and virulence of Legionella spp. populations in freshwater systems in Germany, Palestine and Israel” View project

The effects of habitat structure in managed forests on bird community composition and behavior View project

All content following this page was uploaded by Ido Izhaki on 17 October 2017.

The user has requested enhancement of the downloaded file. Environmental Microbiology Reports (2012) 4(1), 97–104 doi:10.1111/j.1758-2229.2011.00309.x

Bacterial communities in floral nectaremi4_309 97..104

Svetlana Fridman,1 Ido Izhaki,1 Yoram Gerchman2 the dominance of sugar (> 90% dry weight) in nectar, and Malka Halpern1,2* non-sugar compounds play an important role. These com- 1Department of Evolutionary and Environmental Biology, pounds (< 10% dry weight) include amino acids, organic Faculty of Natural Sciences, University of Haifa, Mount acids, lipids, essential oils, polysaccharides, vitamins, Carmel, 31905 Haifa, Israel. antioxidants, minerals and secondary metabolites (Baker 2Department of Biology and Environment, Faculty of and Baker, 1983; Dafni, 1992; Carter et al., 2006). Natural Sciences, University of Haifa, Oranim, 36006 Nectar consumers such as insects, birds and bats were Tivon, Israel. suggested to transfer microflora among flowers, and between flowers and other organs (Sandhu and Waraich, 1985). However, floral nectar was suggested to Summary be not suitable as bacterial habitat and was even demon- Floral nectar is regarded as the most important strated to have antimicrobial properties (Sasu et al., reward available to animal-pollinated to attract 2010). These properties could be due to several chemical pollinators. Despite the vast amount of publications components that were suggested to limit growths of on nectar properties, the role of nectar as a natural microflora in the nectar: (i) high sugar concentration in bacterial habitat is yet unexplored. To gain a better may impose high osmotic pressure that con- understanding of bacterial communities inhabiting strains microbial growth (Pusey, 1999; Brysch-Herzberg, floral nectar, culture-dependent and -independent 2004); (ii) nectar-associated proteins were suggested to (454-pyrosequencing) methods were used. Our find- function as a defensive mechanism against microorgan- ings demonstrate that bacterial communities in ism infections by producing reactive oxygen molecules nectar are abundant and diverse. Using culture- (Carter and Thornburg, 2004a–c; González-Teuber et al., dependent method we showed that bacterial commu- 2009; Harper et al., 2010); and (iii) secondary metabolites nities of nectar displayed significant variation among such as phenolics have also been suggested to play an three plant : Amygdalus communis, Citrus antimicrobial role in nectar (Hagler and Buchmann, 1993). paradisi and Nicotiana glauca. The dominant class in Based on these constraints (e.g. Minorsky, 2007), one the nectar bacterial communities was Gammaproteo- may expect that floral nectar is populated by only a few bacteria. About half of the isolates were novel species microbiota groups that are adapted to live in such extreme (< 97% similarities of the 16S rRNA gene with known environment. species). Using 454-pyrosequencing we demon- Although some publications have indicated the pres- strated that nectar microbial community are distinct ence of microorganisms in floral nectar, most of them for each of the plant species while there are no sig- described fungi and yeasts (Sandhu and Waraich, 1985; nificant differences between nectar microbial commu- Lachance et al., 2001; Brysch-Herzberg, 2004; Manson nities within nectars taken from different plants of the et al., 2007; Herrera et al., 2008; 2009; Pozo et al., 2009; same species. Primary selection of the nectar bacte- 2011) and only one addressed the presence of bacteria ria is unclear; it may be affected by variations in the (Gilliam et al., 1983). A significant negative correlation chemical composition of the nectar in each plant. The was found between yeast density and sugar content, as role of the rich and diverse nectar microflora in the well as yeast density and nectar concentration in a Wat- attraction–repulsion relationships between the plant sonia species (de Vega et al., 2009). Herrera and Pozo and its nectar consumers has yet to be explored. (2010) described a phenomenon whereby the sugar catabolism of yeast populations inhabiting floral nectar can increase its temperature and thus modify the thermal Introduction microenvironment within the flower. However, despite Floral nectar is considered the most important reward the vast amount of publications on nectar properties, the animal-pollinated plants furnish to attract pollinators role of nectar as a natural habitat for microorganisms (Forcone et al., 1997; Bernardello et al., 1999). Despite and specifically for bacteria is yet unexplored. Microbial communities in nectar may affect the nectar’s chemical Received 22 July, 2011; accepted 27 October, 2011. *For correspon- dence. E-mail [email protected]; Tel. (+972) 4 9838727; profile, thus directly controlling nectar consumption by Fax (+972) 4 9838911. flower visitors such as pollinators and nectar thieves, and

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd 98 S. Fridman, I. Izhaki, Y. Gerchman and M. Halpern consequently indirectly governing plant fitness (Herrera Culturable microbial communities structure in nectar of et al., 2008; 2009; Herrera and Pozo, 2010). different plant species The aim of the current study was to gain a better under- Nectar samples were collected aseptically from flowers of standing of floral nectar bacterial communities. To that A. communis, C. paradisi and N. glauca and were spread end, we studied the bacterial communities in flower nectar onto R2A agar (Himedia) and R2A agar supplemented of three plant species: Nicotiana glauca, Amygdalus com- with 20% sucrose. One hundred representative isolates munis and Citrus paradisi in Northern Israel. Here we were identified by amplifying and sequencing the 16S show that (i) bacteria are common inhabitants of floral rRNA gene (Appendix S1). nectar and (ii) each plant species floral nectar support a About 33%, 75% and 42% of A. communis, C. paradisi unique bacterial community. and N. glauca nectar isolates, respectively, were found to be novel species (< 97% similarities in the 16S rRNA gene Results and discussion sequences to known species) (Table 1). This demon- The presence of bacteria in floral nectar strates that indeed, the nectar is an unexplored bacterial niche. Representatives of the Gammaproteobacteria Yeasts have been shown to inhabit flower nectar (Manson class dominated all nectar samples, accounting for 59%, et al., 2007; Herrera et al., 2008; 2009; Manson, 2009). 82% and 45% of the nectar isolates in A. communis, However, as far as we know, nectar has not been consid- C. paradisi and N. glauca respectively (Table 1). Isolates ered a bacterial habitat. On the contrary, Minorsky (2007) belonging to the Bacilli class also occurred in the nectar raised the question why microbes do not grow in nectar, from all plant species. Actinobacteria was identified only and quoted Carter and colleagues (2007) who found that from A. communis and N. glauca. Representatives of the nectarins, proteins which accumulate in the nectar of Alphaproteobacteria and Flavobacteria classes were ornamental tobacco plants, produce very high levels of found only in the A. communis nectar. Interestingly, the hydrogen peroxide (up to 4 mM). This might be the case most abundant species in the nectar were a novel uniden- for some bacterial strains, but as we show here, many tified Enterobateriaceae species in A. communis and others thrive in floral nectar. C. paradisi and Acinetobacter sp. in C. paradisi and N. glauca (Table 1). Nectar sampling and bacterial counts Significant differences were found between nectar bac- terial communities from different plant species (Table 1 Flower nectar from three plant species: N. glauca (Tree and Fig. 1). Figure 1 displays the results from the canoni- Tobacco), A. communis (Almond) and C. paradisi (Grape- cal correspondence analysis using the CANOCO computer fruit) were collected from flowers of each sampled plant program. The distribution of the bacterial species along (five different plants for each plant species) between March the ordinates was not random according to the Monte and June 2009. All the sampled plants were located within Carlo test (F = 1.14, P < 0.05) and thus can be explained a radius distance of up to 10 kilometres in Northern Israel. by their different plant species source. Plant species These three plant species were chosen because their floral explained 42% of the variation in the bacterial community nectar is known to contain secondary metabolites which composition whereas the horizontal and the vertical axes are considered as an antimicrobial agent: N. glauca con- explained 24% and 18% of the variation respectively tains nicotine and anabasine, A. communis contains (Fig. 1). amygdalin and C. paradisi contains caffeine (Detzel and Wink, 1993; Kretschmar and Baumann, 1999; London- Shafir et al., 2003; Tadmor-Melamed et al., 2004). 454-pyrosequencing of 16S rRNA genes Using DAPI, we found that bacterial counts per millilitre in the nectar samples from N. glauca, A. communis and Bacterial diversity in all nectar samples was surveyed by C. paradisi were 1.4 ¥ 107 (Ϯ 1.6 ¥ 106), 1.7 ¥ 107 454-pyrosequencing of 16S rRNA genes (five samples (Ϯ 4.8 ¥ 106) and 3.1 ¥ 107 (Ϯ 2.0 ¥ 106) respectively. per plant species, 15 samples in total). A total of approxi- Bacterial cfu ml-1 in the nectar samples from the different mately 10 000 sequences per sample were obtained. plants were approximately 50%, 25% and 10% from the Nevertheless, after chloroplasts and Archaea sequences DAPI counts respectively. These results demonstrate that were removed from the analysis, about 3200–7000 bacteria thrive in floral nectar. These high abundances of sequences per sample were analysed (77 077 bacteria (> 106 cfu per millilitre nectar) were about two sequences, in total) (see also Appendix S1). magnitude higher compared with what was observed for Sequences were assigned to species-level operational yeasts in the floral nectar of Helleborus foetidus, taxonomic units (OTUs) using a 97% pairwise-identity vulgaris and Aquilegia pyrenaica cazorlensis (Herrera cut-off. In sum, 2197 OTU’s were obtained for all 15 et al., 2008). samples with an average of 401, 379 and 207 OTU’s

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 Bacterial communities in floral nectar 99

Table 1. List of bacterial isolates from nectar of Amygdalus communis, Citrus paradisi and Nicotiana glauca.

Class Closest relative in GenBank database Amygdalus communis Citrus paradisi Nicotiana glauca

Alphaproteobacteria Asaia astilbes 1 (100) Bartonella rattaustraliani 1 (97.2) Gammaproteobacteria Acinetobacter baumannii 2 (94.7–96.0) Acinetobacter baylyi 6 (95.5–96.1) Acinetobacter bouvetii 1 (95.1) Acinetobacter 1 (96.2) calcoaceticus Acinetobacter gerneri 6 (95.8–96.3) 1 (96.1) Acinetobacter johnsonii 2 (95.2–95.5) 1 (96.6) Acinetobacter radioresistens 1 (96.1) 3 (95.9–96.1) Acinetobacter schindleri 2 (95.1–96.2) Erwinia amylovora 5 (96.9–97.4) Erwinia persicina 2 (98.5–99.6) Pantoea agglomerans 2 (99.3) Pantoea septica 1 (98.7) Pseudomonas flectensa 14 (96.2–97.1) 17 (96.5–97.1) 3 (94.5–96.7) Pseudomonas lutea 2 (98.8–99.9) Pseudomonas rhizosphaerae 1 (99.0) Pseudomonas trivialis 1 (100) Pseudomonas viridiflava 2 (99.3–100) Pseudomonas synxantha 1 (99.7) Actinobacteria Arthrobacter tumbae 1 (99.7) 1 (99.8) Arthrobacter pascens 1 (99.9) Curtobacterium flaccumfaciens 1 (98.5) Kocuria kristinae 1 (98.0) Bacilli Bacillus megaterium 3 (97.9–99.6) Bacillus safensis 1 (99.5) Paenibacillus illinoisensis 1 (94.8) Paenibacillus validus 2 (99.1) Staphylococcus epidermidis 4 (98.9–100) Staphylococcus warneri 1 (100) 1 (99.9) Flavobacteria Chryseobacterium indoltheticum 1 (98.5) a. Isolates that were identified as most closely related to Pseudomonas flectens do not belong to the Pseudomonas genus and are in fact novel species in a novel genus in the Enterobacteriaceae family (M. Halpern, S. Fridman and I. Izhaki, unpubl. data). See also Fig. 5. The number before the parentheses indicates the number of isolates, the number within the parentheses indicates the percentage of the 16S rRNA gene similarities to the closest known species. Isolates with less than 97.5% 16S rRNA gene similarities to known species are most likely novel species and the name of their closest relative species is marked in bold. The isolates were identified by comparing their 16S rRNA gene sequences to that of the GenBank database (EZtaxon version 2.1. http://www.eztaxon.org). Sequences lengths were at least 850 bp. Sequences length obtained for most Acinetobacter species and for all the isolates that were identified as closely related to Pseudomonas flectens were 1300– 1500 bp. Accession numbers of the 16S rRNA gene sequences are HQ284799–HQ284831, HQ284869–HQ284906 and HQ284948–HQ284970. per sample and coverage of 95.8%, 97.2% and 91.4% for the different plant species nectar (A. communis vs.

A. communis, C. paradisi and N. glauca nectar bacterial C. paradisi, F1,9 = 3.97, P < 0.01; A. communis vs. communities, respectively. Chao1 richness estimator for N. glauca, F1,9 = 3.06, P < 0.01; C. paradisi vs. N. glauca

A. communis, C. paradisi and N. glauca nectar bacterial F1,9 = 3.97, P < 0.05). This indicates that nectar from communities was, 689, 629 and 418, respectively (see each plant species has a distinct microbial community also Appendix S1). (Fig. 2). This analysis revealed clustering of samples by their The majority of the sequences from all the nectar plant origin, except for one nectar sample from N. glauca samples were classified as Proteobacteria (> 83.0%). that overlapped with C. paradici samples (Fig. 2). Gammaproteobacteria was the dominant class and com- However, when AMOVA analysis was applied, significant prised 79.5%, 92.8% and 72.9% of the sequences in differences were found between the bacterial com- C. paradisi, N. glauca and A. communis respectively munities from the nectar samples that originated (Fig. 3, upper graph). The most prevalent families were from the three different plant species (F7,14 = 3.14; Moraxellaceae and Enterobacteriaceae. Acinetobacter P < 0.01). AMOVA post hoc analysis revealed significant was the dominant genus with the frequency of 49%, differences between the bacterial communities from 90% and 78% of the bacterial species in A. communis,

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 100 S. Fridman, I. Izhaki, Y. Gerchman and M. Halpern

Novel bacterial species in floral nectar

Acinetobacter. Acinetobacter species (Gammaproteo- bacteria) seemed to play an important role in the bacterial communities of flower nectar, as 25% of the cultivated nectar species were identified as novel Acinetobacter species (Table 1). Acinetobacter was also the dominant genus in the 454-pyrosequencing results (Fig. 3, upper part). To assure the novelty of the isolated Acinetobacter species, the Z1-Z2 region of the rpoB gene (coding for RNA polymerase B) of the isolates was amplified, sequenced and compared to known Acinetobacter species (Fig. 4). The different isolates shared 82–85% similarities with the following species: Acinetobacter gri- montii (10 isolates), Acinetobacter tjernbergiae (seven isolates), Acinetobacter gerneri (three isolates), Acineto- bacter baumannii and Acinetobacter ursingii (one isolate), demonstrating that they belong not only to novel species but most likely to novel genera (Fig. 4). The phylogenetic analyses of the rpoB gene sequences demonstrated that

Fig. 1. Ordination diagram (calculated with CANOCO software) Acinetobacter isolates belong to at least two different showing variation in the abundance of bacterial species isolates groups, both forming an out-group to all known Acineto- among the three plant species. This joint plot data analyse the bacter type strains (Fig. 4). relationship between bacterial species and plant species. The environmental variables are displayed as arrows radiating from the centre of the diagram. The length of the arrows represents the contribution of each plant species to the variation of the sample. The angle in between two arrows is a measure for the correlation between the two variables (small angle means high correlation), and the projection of a taxa point on an arrow is a measure for the relative value of that point; in other words, for the position of that point on the gradient described by the arrow. The green triangles represent different bacterial species. The identity of the species is as follows: group 1: Erwinia amylovora, Arthrobacter tumbae, Kocuria kristinae, Bacillus megaterium, Staphylococcus epidermidis; group 2: Acinetobacter radioresistens; group 3: Acinetobacter johnsonii; group 4: Arthrobacter tumbae, Staphylococcus warneri; group 5: Acinetobacter gerneri; group 6: Bacillus safensis, Pseudomonas synxantha, Acinetobacter schindleri, Acinetobacter bouvetii, Acinetobacter baumannii, Acinetobacter baylyi, Paenibacillus illinoisensis; group 7: Enterobacteriaceae nov. genus (the former Pseudomonas flectens); group 8: Acinetobacter calcoaceticus, Paenibacillus validus, Curtobacterium flaccumfaciens, Erwinia persicina, Asaia siamensis, Pantoea agglomerans, Pantoea septica, Bartonella rattaustraliani, Chryseobacterium indoltheticum, Pseudomonas rhizosphaerae, Pseudomonas trivialis, Pseudomonas viridiflava, Pseudomonas lutea.

Fig. 2. Nectar bacterial diversity clustering by plant species. C. paradisi and N. glauca nectar samples respectively Bacterial diversities of all nectar samples were surveyed by (Fig. 3). AMOVA analysis revealed significant differences 454-pyrosequencing of 16S rRNA genes. The first two principal between Acinetobacter species from the nectar bacterial coordinates (PC1 and PC2) from the principal coordinate analysis of unweighted UniFrac are plotted for each sample. Each symbol communities that originated from the different plant represents a sample, coloured by plant species (N. glauca, green; species (F7,13 = 3.03; P < 0.01), demonstrating that C. paradisi, red; A. communis, blue). The variance explained by the although Acinetobacter was the main genus in all nectar PCs is indicated on the axes. AMOVA analysis showed significant differences between the bacterial communities from the nectar samples, different plant nectar inhabits different Acineto- samples that originated from the different plant species bacter species. The other bacterial classes that were (F7,14 = 3.14; P < 0.01). The 454-pyrosequencing technique found in much lower abundances in the bacterial commu- demonstrates that nectar from each of the three plant species has a distinct microbial community, while there are no significant nities from the different plants are specified in Fig. 3 differences between nectar microbial communities within nectar (lower part). samples of the same plant species.

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 Bacterial communities in floral nectar 101

Fig. 3. Mean class abundances of bacterial communities from nectar samples of the different plant species A. communis, C. paradisi and N. glauca. Most of the sequences belonged to the Gammaproteobacteria (upper part of the figure). The prevalence of the genus Acinetobacter out of the Gammaproteobacteria sequences is indicated. All the rest of the classes are specified in the lower part of the figure.

It seems that a significant fraction of the bacteria in that family is required for definite taxonomic conclusion nectar (at least for C. paradisi) are culturable species as (Anzai et al., 2000). Chanbusarakum and Ullman (2008) seen in the food industry [e.g. raw milk, in which culturable isolated an unidentified bacterial strain from a western bacteria are considered more than 50% of the total bac- flower thrip. This unidentified species was closely related teria (Hantsis-Zacharov and Halpern, 2007)]. Interest- to our isolates (Fig. 5), thus, possibly indicating that thrips, ingly, Acinetobacter isolates grew well on LB or R2A agar which are tiny, slender insects, feeding on pollen, might be plates with the supplement of 20% sucrose but very the vectors of transmission of this species in the flower’s poorly without the addition of sucrose. Furthermore, iso- nectar. lates could not be transferred more than 10 times from their first isolation (data not shown), suggesting that they Secondary metabolites and nectar bacterial isolates are lacking some nutrition from the flower’s nectar. Another interesting phenomenon was that most of the Given that the tested nectar is known to contain second- Acinetobacter isolates seemed to produce a mucus ary metabolites, antagonistic interactions of these matrix. The floral nectar contains high sucrose concentra- metabolites with the nectar isolates were tested. The sec- tions – the mean value of 64.4% sucrose, for example, ondary metabolites concentrations in the different plant was found in the nectars of 278 plant species pollinated species are: N. glauca nicotine (0.56 Ϯ 0.12 ppm) and by hummingbirds (Nicolson and Fleming, 2003). It is pos- anabasine (5.4 Ϯ 0.90 ppm), A. communis amygdalin sible that the nectar’s sucrose is used by the bacteria to (4–10 ppm) and C. paradisi caffeine (94.26 Ϯ 2.90 ppm) produce polysaccharides. However, it is unclear what the (Detzel and Wink, 1993; Kretschmar and Baumann, chemical composition of these polysaccharides is, and 1999; London-Shafir et al., 2003; Tadmor-Melamed et al., how the bacteria or the plant may benefit from them. 2004). Representative isolates from different plant species were spread onto R2A supplemented with 10% Enterobateriaceae gen. nov. sp. nov. Another bacterial sucrose. Secondary metabolites in different concentra- species that showed high prevalence in the nectar with 34 tions [amygdalin (5, 50 and 1000 ppm), caffeine (95, 200 isolates from all three plant species were novel Enteroba- and 1000 ppm), nicotine (0.6, 5 and 1000 ppm) and ana- teriaceae species. The novel species showed the highest basine (5, 10 and 1000 ppm)] were added onto paper similarity (but less than 97%) to Pseudomonas flectens disks which were placed in the middle of the agar plates. which is misclassified as a member of the genus No antagonistic interactions were found between the sec- Pseudomonas (Table 1, Fig. 5). Pseudomonas flectens ondary metabolites and the bacterial isolates. was included in the family Enterobacteriaceae, but an Plant secondary metabolites in floral nectar are mainly extensive study comparing this species with others from recognized as deterrents and toxins for a variety of

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 102 S. Fridman, I. Izhaki, Y. Gerchman and M. Halpern

11NS3 Fig. 4. Neighbour-joining phylogenetic tree of 10NS2 bacterial partial rpoB gene sequences. The tree presents isolates that are affiliated to the 10N4 genus Acinetobacter (Gammaproteobacteria) 13N3 and type strains in this genus. All 17N2 Acinetobacter isolates represent novel 76 10NS6 species. The phylogenetic tree was 10N2 constructed in MEGA 4.1 software. Bootstrap 97 values greater than 50% are shown at the 12NS2 branch points. Pseudomonas aeruginosa 97 14N2 LMG 1242T was used as an out-group. The 86 14N1 bar indicates 10% sequence divergence. 87 9N6 The letter N in the isolate’s name indicates its nectar origin. Isolate names with no special 12N1 mark: C. paradisi nectar isolates; bold, 18N3 97 underlined letters: N. glauca nectar isolates. 66 11NS4 13NS1 11N2 68 11N1 54 9N1 Acinetobacter ursingii NIPH 137T 58 Acinetobacter lwoffii CCM 5581T Acinetobacter schindleri NIPH 1034T 51 Acinetobacter bouvetii CCM 7196T Acinetobacter berezinae LMG 1003T 81 Acinetobacter baylyi CCM 7195T Acinetobacter gerneri CCM 7197T Acinetobacter towneri CCM 7201T Acinetobacter grimontii DSM14968 100 Acinetobacter baumannii ATCC 19606T Genomic sp. 13TU LMG 993 54 Acinetobacter johnsonii LMG 999T Acinetobacter beijerinckii NIPH 838T Acinetobacter parvus NIPH 384T 78 Acinetobacter tjernbergiae CCM 7200T 81 Genomic sp. 16 ATCC 17988 100 Genomic sp. 15BJ LUH 1729 Acinetobacter haemolyticus CCM 2358T Pseudomonas aeruginosa LMG 1242T

0.1

organisms (Wink, 1999). Hagler and Buchmann (1993) pressure, nectarins and secondary metabolites may limit suggested that secondary metabolites in nectar could bacterial growth. have an antimicrobial function although currently there is no published data evaluating the antimicrobial effect of Conclusions nectar secondary metabolites (Adler, 2000). In our experi- ments we did not find any antagonistic effect of amygda- Here we demonstrate, for the first time, that floral nectar lin, caffeine, nicotine or anabasine (all found in the nectar is a unique, specialized and diverse bacterial habitat. of the tested plant species) on growth of the bacterial Using culturable and molecular methods we showed isolates from the different plant species. We tested the that different plant species have distinctive bacterial effect of the secondary metabolites in vitro. However, it community composition. Significant differences were might be the case that these secondary compounds affect found between nectar bacterial communities from bacteria differently in the presence of the other chemical different plant species but not between different plants of constituents of the nectar (in vivo). Our study indicates the same plant species (Figs 1 and 2). Primary selection that floral nectar is a rich medium for microbial growth of the nectar bacteria is unclear; it may be affected by despite the notion that several constrains such as osmotic the different chemical compositions of the nectar in

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 Bacterial communities in floral nectar 103

8N6 Fig. 5. Neighbour-joining phylogenetic tree of 1010NS3 bacterial partial 16S rRNA gene sequences. 4N3 The tree presents isolates that are affiliated to 1313NSNS3 the family Enterobacteriaceae and type 8N4 strains in this family. All isolates represent a 9N9N5 novel genus. The phylogenetic tree was 8NS6 constructed in MEGA 4.1 software. Bootstrap 5NS5 10N310N3 values greater than 50% are shown at the 1313N2 branch points. The bar indicates 1% 9NS9NS3 sequence divergence. Pseudomonas EnteEnterobacteriaceaeobacteriaceae bacteriumbacterium BFo-2BFo-2 ((EU029106)EU029106) aeruginosa LMG 1242T was used as an 9NS9NS5 out-group. 5NS1 The letter N in the isolate’s name indicates its 1919NS3 nectar origin. Isolate names with no special 4N1 mark: C. paradisi nectar isolates; bold, 4NS3 62 5NS2 underlined letters: N. glauca nectar isolates; 8N2 white letters on grey background: EnteEnterobacteriaceaeobacteriaceae bacteriumbacterium M515M515 ((AB461801)AB461801) A. communis nectar isolates. 11NS5 8N3 9NS9NS6 8NS1 12N312N3 91 9N2 4NS5 64 1919NS2 9NS9NS1 1010NSNS5 96 12N212N2 1010NSNS1 1313NSNS4 88 1919NS1 8NS7 64 10N110N1 UnidentifiedUnidentified thripthrip gutgut bacteriubacterium (AF024609)(AF024609) PseudomonasPseudomonas flectensflectens ATCCTCC 1277512775T 80 Tatutumellaella citcitrea LMGLMG 2204922049T Yersrsiniaia massmassilieiliensnsis CCUGCCUG 534453443T BudviciaBudvicia aquaticaaquatica DSMDSM 50755075T CedeceCCedeceaedecea ddavisaeavisae DSDSM 45684568T PantoeaPantoea eucalyptieucalypti LMGLMG 2241984198T Eschericichia coli KCTCCTC 24412441T 57 Cronobacteronobacter turicensisturicensis Z3032Z3032T HafniaHafnia alveialvei ATCCTCC 1333713337T Raoultellltella plananticticola DSMDSM 30693069T KluyveraKluyvera cryoccryocrescensescens ATCCTCC 3343533435T 58 SeSerrrratiaatia ureilyticeilytica NiNiVa 5511T 86 LLecleecleeclercicia adaadecarboxylatadececarboarboxxylylata GTGTC 12671267T 99 EnteEnterobacterobacter ludwigiiludwigii DSMDSM 1668816688T MoelleMoellerellaella wisconsensiwisconsensis DSMDSM 56765676T PseudomonasPseudomonas aeruginosaaeruginosa LMGLMG 11242242T 0.010.01 the different plant species. The effect of the bacteria References on the nectar is also unknown; these bacteria may Adler, L.S. (2000) The ecological significance of toxic nectar. have a crucial influence on the chemical profile of the Oikos 91: 409–410. nectar. For instance, they may produce volatile com- Anzai, Y., Kim, H., Park, J.Y., Wakabayashi, H., and Oyaizu, pounds which can impact visitation rate by nectar con- H. (2000) Phylogenetic affiliation of the pseudomonads sumers such as pollinators and nectar thieves, thus based on 16S rRNA sequence. Int J Syst Evol Microbiol 4: affecting plant pollination and fitness. Further study on 1563–1589. the effect of pollinators or specific vectors that consume Baker, H.G., and Baker, I. (1983) A brief historical review of chemistry of floral. In The Biology of Nectaries. Bentley, B., the nectar, on bacterial community composition in and Elias, T.S. (eds). New York, USA: Columbia University nectar, distribution and transfer between flowers is Press, pp. 126–152. needed. Bernardello, G., Galetto, L., and Forcone, A. (1999) Floral nectar chemical composition of some species from Patago- Acknowledgements nia. II. Biochem Syst Ecol 27: 779–790. Brysch-Herzberg, M. (2004) Ecology of yeasts in plant- This study was supported by a grant from the Israel Science bumblebee mutualism in Central Europe. FEMS Microbiol Foundation (ISF, grant no. 189/08). Ecol 50: 87–100.

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104 104 S. Fridman, I. Izhaki, Y. Gerchman and M. Halpern

Carter, C., and Thornburg, R.W. (2004a) Tobacco nectarin V Lachance, M.A., Bowles, J.M., Kwon, S., Marinoni, G., is a flavin-containing berberine bridge enzyme-like protein Starmer, W.T., and Janzen, D.H. (2001) Metschnikowia with glucose oxidase activity. Plant Physiol 134: 460– lochheadii and Metschnikowia drosophilae, two new yeast 469. species isolated from insects associated with flowers. Can Carter, C., and Thornburg, R.W. (2004b) Is the nectar redox J Microbiol 47: 103–109. cycle a floral defense against microbial attack? Trends London-Shafir, I., Shafir, S., and Eisikowitch, D. (2003) Plant Sci 9: 320–324. Amygdalin in almond nectar and pollen – facts and pos- Carter, C., and Thornburg, R.W. (2004c) Tobacco nectarin III sible roles. Plant Syst Evol 238: 87–95. is a bifunctional enzyme with monodehydroascorbate Manson, J.S. (2009) The ecological consequences and reductase and carbonic anhydrase activities. Plant Mol Biol adaptive function of nectar secondary metabolites. PhD 54: 415–425. Dissertation. Toronto, CA, Canada: University of Toronto. Carter, C., Shafir, S., Yehonatan, L., Palmer, R.G., and Manson, J.S., Lachance, M.A., and Thomson, J.D. (2007) Thornburg, R. (2006) A novel role for proline in plant floral Candida gelsemii sp. nov., a yeast of the Metschnikowi- nectars. Naturwiss 93: 72–79. aceae clade isolated from nectar of the poisonous Carolina Carter, C., Healy, R., O’Tool, N.M., Naqvi, S.M., Ren, G., jessamine. Anton Leeuw Int J G 92: 37–42. Park, S., et al. (2007) Tobacco nectaries express a novel Minorsky, P.V. (2007) Why Don’t Microbes Grow in Nectar? NADPH oxidase implicated in the defense of floral repro- Plant Physiol 143: 1–2. ductive tissues against microorganisms. Plant Physiol 143: Nicolson, S.W., and Fleming, P.A. (2003) Nectar as food for 389–399. birds: the physiological consequences of drinking dilute Chanbusarakum, L., and Ullman, D. (2008) Characterization sugar solutions. Plant Syst Evol 238: 139–153. of bacterial symbionts in Frankliniella occidentalis (Per- Pozo, M.I., de Vega, C., Canto, A., and Herrera, H.C. (2009) gande), Western flower thrips. J Invert Pathol 99: 318–325. Presence of yeasts in floral nectar is consistent with the Dafni, A. (1992) Pollination Ecology: A Practical Approach. hypothesis of microbial-mediated signaling in plant- Oxford, UK: Oxford University Press. pollinator interactions. Plant Signal Behav 4: 1102–1104. Detzel, A., and Wink, M. (1993) Attraction, deterrence or Pozo, M.I., Herrera, H.C., and Bazaga, P. (2011) Species intoxication of bees (Apis mellifera) by plant allelochemi- richness of yeast communities in floral nectar of southern cals. Chemoecol 4: 8–18. Spanish plants. Microb Ecol 61: 82–91. Forcone, A., Galetto, L., and Bernardello, L. (1997) Floral Pusey, P.L. (1999) Effect of nectar on microbial antagonists nectar chemical composition of some species from Patago- evaluated for use in control of fire blight of pome fruits. nia. Biochem Syst Ecol 25: 395–402. Phytopathology 89: 39–46. Gilliam, M., Moffett, J.O., and Kauffeld, N.M. (1983) Exami- Sandhu, D.K., and Waraich, M.K. (1985) Yeasts associated natin of floral nectar of citrus cotton, and Arizona desert with pollinating bees and flower nectar. Microb Ecol 11: plants for microbes. Apidologie 14: 299–302. 51–58. González-Teuber, M., Eilmus, S., Muck, A., Svatos, A., and Sasu, M.A., Wall, K.L., and Stephenson, A.G. (2010) and Heil, M. (2009) Pathogenesis-related proteins protect Antimicrobial nectar inhibits a florally transmitted pathogen extrafloral nectar from microbial infestation. Plant J 58: of a wild Cucurbita pepo (Cucurbitaceae). Am J Bot 97: 464–473. 1025–1030. Hagler, J.R., and Buchmann, S.L. (1993) Honey bee Tadmor-Melamed, H., Markman, S., Arieli, A., Distl, M., Wink, (Hymenoptera; Apidae) foraging responses to phenolic- M., and Izhaki, I. (2004) Limited ability of Palestine sun- rich nectars. J Kansas Entomol Soc 66: 223–230. birds (Nectarinia osea) to cope with pyridine alkaloids in Hantsis-Zacharov, E., and Halpern, M. (2007) Culturable psy- nectar of tree tobacco (Nicotiana glauca). Funct Ecol 18: chrotrophic bacterial communities in raw milk and their 844–850. proteolytic and lipolytic traits. Appl Environ Microbiol 73: de Vega, C., Herrera, C.M., and Johnson, S.D. (2009) Yeasts 7162–7168. in floral nectar of some South African plants: quantification Harper, A.D., Stalnaker, S.H., Wells, L., Darvill, A., Thorn- and associations with pollinator type and sugar concentra- burg, R., and York, W.S. (2010) Interaction of Nectarin 4 tion. S Afr J Bot 75: 798–806. with a fungal protein triggers a microbial surveillance and Wink, M. (1999) Functions of Plant Secondary Metabolites defense mechanism in nectar. Phytochemistry 71: 1963– and Their Exploitation in Biotechnology. Boca Raton, FL, 1969. USA: CRC Press. Herrera, C.M., and Pozo, M.I. (2010) Nectar yeasts warm the flowers of a winter-blooming plant. Proc R Soc B 277: 1827–1834. Supporting information Herrera, C.M., García, I.M., and Pérez, R. (2008) Invisible floral larcenies: microbial communities degrade floral Additional Supporting Information may be found in the online nectar of bumble bee-pollinated plants. Ecology 89: 2369– version of this article: 2376. Appendix S1. Experimental procedures. Herrera, C.M., de Vega, C., Canto, A., and Pozo, M.I. (2009) Yeasts in floral nectar: a quantitative survey. Ann Bot 103: Please note: Wiley-Blackwell are not responsible for the 1415–1423. content or functionality of any supporting materials supplied Kretschmar, J.A., and Baumann, T.W. (1999) Caffeine in by the authors. Any queries (other than missing material) Citrus flowers. Phytochemistry 52: 19–23. should be directed to the corresponding author for the article.

© 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Reports, 4, 97–104

View publication stats