Journal of Apicultural Research 43(3): 101–104 (2004)
© IBRA 2004
ORIGINAL ARTICLE
An approach to the characterization of the
honey bee hive bacterial flora
CLAUDIA PICCINI,* KARINA ANTÚNEZ AND PABLO ZUNINO
Laboratorio de Microbiología, Instituto de Investigaciones Biológicas Clemente Estable, Avenida Italia 3318, 11600 Montevideo,
Uruguay
Received 29 December 2003, accepted subject to revision 19 May 2004, accepted for publication 17 June 2004
SUMMARY
In this study we describe an approach to the characterization of the cultivable heterotrophic bacterial flora of
honey bee hives. We assessed the bacterial abundance in brood combs and on hive floors at different times of
the year, and identified several isolated bacteria. The number of bacteria recovered from brood combs was
notably lower than the number obtained from hive floors. In addition, we observed an increase in bacteria
recovered from brood comb samples from winter to summer, probably reflecting the activity of the colony
during these seasons. In floor samples bacterial counts were not different among the different sampling times.
Most of the identified isolates belonged to the genera Bacillus and Corynebacterium. Spores of Paenibacillus larvae
subspecies larvae were only found in samples taken during summer of 2002, in accordance with an increase of
American foulbrood incidence in Uruguay in the last year. The presence of several micro-organisms such as
those identified as belonging to the genus Bacillus, potentially related to those that have biological control
capabilities, and the differences between brood combs and floor-recovered bacteria encourage further research
on the microbiology of honey bee hives and new strategies of pathogens control.
Keywords: bacteria, honey bees, hives, Bacillus, brood comb, beehive floor
INTRODUCTION
Apicultural economic development strongly relies on the health
status of honey bee colonies. Several bacterial diseases that affect
honey bees have been described, including American foulbrood
(AFB) and European foulbrood, caused by Paenibacillus larvae subspecies larvae (Heyndrickx et al., 1996) and Melissococcus plutonius (Allen & Ball, 1993), respectively. The presence of P. l. larvae
in Uruguay has been reported already in larvae, honey and adult
bees (Piccini & Zunino, 2001), but so far, studies related to the
bacterial flora present in honey bee hives are lacking. It has been
reported that the microbes most frequently found in honeycombs and adult bees belong to the Bacillus genus (Snowdon &
Cliver, 1996). It is also well known that several soil bacteria, in
particular Bacillus spp., produce antimicrobial substances (Kim et
al., 1997). Thus, information about the interactions that occur
between different bacterial species inside the hive and the
dynamics of the bacterial community could be important in
order to develop new approaches for disease control, avoiding
the use of commercial antimicrobial substances such as antibiotics. In order to gain insight into the microbial ecology of honey
bees, we determined the number of total viable aerobic and
microaerophilic bacteria from two different parts of the hive at
different seasons, and we also identified a number of frequently
found bacterial isolates. In the present study we chose the brood
combs and the floor of the hives as the sampling sites. The work
was done in this way in order to assess differences between the
place where the population resides and great activity takes place
(brood combs), and the place where the colony wastes accumulate (floor).
MATERIALS AND METHODS
Samples
Samples were obtained from an experimental apiary, containing
12 beehives, that belongs to the National Institute for Agricultural Research (INIA), close to the city of Colonia, Uruguay. Sam*Corresponding author: [email protected]
ples were collected from the brood combs and from the floor
of the hive during November (spring) of 2000 (S1), July (winter)
of 2001 (S2) and January (summer) of 2002 (S3).
Pieces of brood combs (0.9 g average weight ± 0.36 s.d.) were
collected from each hive by random sampling and stored at –20
°C until processed. In 24 brood combs samples out of 36
(66.6%) there was presence of honey or pollen, and in eight samples out of 36 (22.2%) there was presence of larvae.
To obtain floor samples, floor material (0.12 g average weight ±
0.045 s.d.) was scraped with an alcohol-disinfected spatula and
the collected samples were stored at –20 °C until processed.
Bacterial enumeration
Comb samples were divided into small fragments using sterile
scalpel blades and weighed. Then, 5 ml of sterile distilled water
(SDW) were added and the mix was vigorously vortexed for 3
min. One 100 µl aliquots of the resultant suspension were spread
in five replicates onto nutrient agar (NA) plates (Difco). When
bacterial growth was confluent and plate counting was not possible, we performed serial decimal dilutions of the original sample in SDW and aliquots of each dilution were spread onto NA
plates. Plates were incubated at 37 °C for 48 h and colony-forming units (cfu) were counted. This procedure was performed at
least three times for each sample. Since we wanted to assess
exclusively bacterial growth, we added 30 µg/ml of nystatin (Urufarma) to the medium in order to avoid the growth of yeasts.
The inhibitory concentration of nystatin for yeasts in this kind
of sample was previously determined using different concentrations of the product for the same kind of samples (data not
shown).
In order to detect the presence of P. l. larvae, the remaining suspension was heated at 80 °C for 15 min and aliquots of 100 µl
were inoculated onto J agar (Hornitzky & Nicholls, 1993) supplemented with 9 µg/ml of nalidixic acid to avoid the growth of
Paenibacillus alvei (Sigma) (J/Nal) (Alippi, 1995). Plates were
Piccini, Antúnez, Zunino
102
A
Brood comb samples
10
log cfu/g
10
8
S1
6
S2
4
S3
2
P I larvae
0
1
2
3
4
5
6
7
8
9
10
11
12
beehives
Floor samples
B
log cfu/g
10
10
8
S1
6
S2
4
S3
2
P I larvae
0
1
2
3
4
5
6
7
8
9
10
11
12
beehives
FIG. 1. Colony-forming units per gram of sample (cfu/g) obtained from brood comb (A) and floor (B) samples of hives in
three sampling times. S1: spring of 2000, S2: winter of 2001 and S3: summer of 2002. Bars indicate cfu/g in brood combs or
floor samples for each beehive in one assay. Presence of Paenibacillus larvae larvae (only determined in samples
corresponding to S3) is indicated by dots.
incubated under microaerophilic conditions (5 to 10% CO2) at
37 °C for 72 h to detect P. l. larvae growth. The colonies were
also examined for detection of AFB clinical symptoms.
Floor samples were weighed, suspended in 1 ml of SDW and
vortexed for 3 min. Serial dilutions in SDW of the resultant suspension were done and aliquots of each dilution were spread on
NA in five replicas. Plates were incubated at 37 °C until growth
was observed and cfu were counted. The remaining suspension
was subjected to the same heating treatment and culture conditions already described for comb samples. After plate counting, the number of cfu per gram of sample (cfu/g) was calculated.
Isolate identification
Different isolates were observed to be pure prior to identification and were selected according to the following phenotypic
traits: colony morphology, cellular morphology and Gram staining and catalase reaction. Those phenotypes that were more frequently isolated were identified using the BBL-CRYSTAL GP kit
for Gram-positive bacteria (Becton Dickinson) and BBL-Enterotube II kit for Enterobacteriaceae (Becton Dickinson). When
identification was not possible using the kits we performed additional standard biochemical or bacteriological tests, including oxidase activity, glucose oxidation-fermentation, urea hydrolysis,
motility at 36 °C (Gerhardt et al., 1994), and growth on Rogosa
selective medium (de Waard et al., 2002). P. l. larvae was identi-
fied by colony shape and margins, microscopic examination of
Gram- stained smears and the next biochemical test: catalase
production, Voges-Proskauer reaction, gelatine liquefaction
(12%), indol production, starch hydrolysis, manitol usage and
growth on nutritive broth (Alippi, 1992).
Statistical analysis
Mann-Whitney one-tailed non-parametric analysis was used to
compare cfu/g from brood combs and hive floors among different seasons. Observed differences were considered significant
at P ≤ 0.05.
RESULTS
Bacterial enumeration
When the numbers of cfu/g were analysed, we found that bacterial numbers in brood combs samples were notably lower than
those observed in floor samples, showing ranges of approximately 101 to 105 in brood combs and from 103 to 108 in floor
samples (fig. 1).
A significant increase in cfu/g of brood comb samples was
observed from winter (S2) to summer (S3) (P = 0.0001) (fig. 1A).
Although we did not find a significant variation between brood
comb bacterial counts from spring (S1) to winter (S2) (P =
0.068), a decrease in cfu/g could be observed in nine out of 12
Microbiology of the honey bee hive
103
TABLE 1. Identification of isolates from brood combs and floor of hives.
Identified isolates
Source (number of isolates)a
Brevibacillus brevis
BC (3), F (5)
Bacillus cereus
BC (2)
Bacillus licheniformis
F (1)
Bacillus megaterium
BC (2), F (1)
Bacillus pumilus
BC (1), F (1)
Bacillus sphaericus
BC (1)
Bacillus subtilis
BC (2) F (2)
Bacillus spp.*
BC (1)
Lactobacillus spp.*
BC (2), F (1)
Leifsonia aquatica
BC (1), F (1)
Corynebacterium propinquum
BC (3), F (7)
Corynebacterium pseudotuberculosis
F (1)
Corynebacterium sp.*
BC, F (2)
Paenibacillus alvei
F (1)
Proteus sp.*
F (1)
Enterobacteriaceae * (Enteric group 60)
BC (2)
Sources of samples were brood comb (BC) and floor (F). The number of isolates is indicated between brackets.
* ID was not at species-level.
a
hives (fig. 1A). No differences between bacterial counts of spring
(S1) and summer (S3) samples were observed (P = 0.817).
We did not find significant differences in bacterial counts for floor
samples among the different sampling times (fig. 1B).
In a preliminary assay we found that the use of other culture
media (Luria-Bertani and Brain-Heart Infusion agar) did not significantly change the recovery of bacteria (data not shown). We
also verified that the bacterial growth was not affected by the
chosen concentration of nystatin used to avoid the presence of
yeasts.
Although storage at –20 °C could result in loss of some of the
less robust organisms, we did not notice variation in the number of total recovered bacteria when control samples were
frozen and thawed twice (data not shown).
Isolate identification
Forty-four different isolates were selected according to their different phenotypic traits and were identified using the commercial kit or complemented with standard tests. Most of the identified isolates (14 out of 44) belonged to the Bacillus genus (table
1), but we also detected bacteria belonging to the genera Brevibacillus, Corynebacterium, Lactobacillus, Paenibacillus, Proteus and
Leifsonia (Evtushenko et al., 2000).
After cultures of heated combs and floor samples on J/Nal, we
isolated P. l. larvae from seven out of 12 colonies from S3 (fig. 1).
P. l. larvae was not recovered from any sample corresponding to
S1 or S2. Among the P. l. larvae positive samples, five were from
brood combs and three were from floor samples. No AFB symptoms were seen in any of the 12 colonies at any time.
DISCUSSION
There are several studies about bacteria present in different
parts of honey bees and bee larvae (Snowdon & Cliver, 1996;
Jeyaprakash et al., 2003). In this study, the cultivatable heterotrophic bacterial flora sampled from two different places in
the beehive, brood comb and hive floor, was assessed.
We observed that the number of bacteria recovered from the
material located on hive floors was markedly higher than the
number of bacteria present in brood comb samples. As it has
been described, fatty acids present in pollen, especially linoleic,
linolenic, myristic and lauric acids, have inhibitory properties
against several micro-organisms including P. l. larvae and B. cereus
(Manning, 2001). In addition, Gilliam (1979) found that Bacillus
subtilis, which produces antibiotics that are active against Grampositive bacteria, was the most common bacterial species
trapped in almond pollen. Therefore, our findings could be
explained by the presence of antimicrobial substances in pollen
and honey (Tichy & Novak, 2000), which were in high amounts
(almost in 70% of the samples) in brood combs. A further possibility that cannot be ruled out is that hygienic behaviour of adult
bees may also be also responsible for the lower number of bacteria detected in brood combs (Spivak & Gilliam, 1998). We also
found a decrease in the number of total bacteria obtained in
floor samples taken during winter that could be related to the
decline of honey bee populations in this season. Besides, bacterial numbers increased significantly from winter to summer in
brood comb samples. This could be related to bee population
recovery during the warmer seasons.
It is important to consider that in a preliminary assay we found
that the use of other culture media (Luria-Bertani and BrainHeart Infusion agar) did not significantly change the bacterial
counts (data not shown). We also verified that bacterial growth
was not affected by the chosen concentration of nystatin used
to avoid the presence of yeasts.
Identification of isolates using commercial kits was not completely effective, probably because these kinds of kits are
designed for clinical strains. For this reason, some isolates could
be identified only to genus level.
The prevalence of Bacillus species among the identified isolates
is in agreement with previous reports, where it is described that
bacteria isolated from different organs of the queen honey bee
are mostly B. megaterium, B. cereus, B. subtilis and B. brevis (Gilliam,
1978). Furthermore, we also found P. alvei, a common inhabitant
of the beehive (Djordjevic et al., 2000), Proteus, a genus previously
104
associated with honey bees (Snowdon & Cliver, 1996) that
includes an insect pathogen species (P. myxofaciens, Rozalsky et
al., 1996), and members of the Enterobacteriaceae family belonging to the Enteric Group 60. Enteric Groups are defined as
groups containing biochemically similar strains whose classification needs further study. Particularly, Enteric Group 60 was first
thought to be inactive strains of Morganella, but this hypothesis
was ruled out and there is no clue to its correct taxonomic position (Farmer et al., 1985).
The environment that surrounds honey bee hives could provide
important sources of bacteria, especially those commonly found
in soil. In the case of honey, the primary sources of microbial
contamination include pollen, dust, air, soil and the digestive tract
of honey bees (Snowdon & Cliver, 1996). This may explain the
presence of Corynebacterium pseudotuberculosis, the causative
organism of caseous lymphadenitis in sheep and goats (Connor
et al., 2000) that was isolated in this study. Its presence in the
beehive samples could be explained by the fact that the apiary
is located in a farming area. Thus, the source of this organism
was probably the surrounding vegetation or water resources,
being carried inside the hive by foraging honey bees.
The genus Lactobacillus comprises the largest group of rodshaped organisms within the lactic acid bacteria. These bacteria
are widespread in nature and are frequently found in association
with the intestinal tracts and mucous membranes of a wide range
of animal species (Lawson et al., 2001). Lactobacillus and related
bacteria have been previously identified in Apis mellifera capensis
and A. m. scutellata and even from other honey bee subspecies
(Jeyaprakash et al., 2003). Lactobacillus spp. are being intensively
studied for probiotic usage. The frequent presence of native Lactobacillus strains associated with honey bees confirmed in this
study could encourage research related to the use of probiotics
or related products for apiculture.
Along with bacteria from environmental sources, bacterial
pathogens can also be found inside the hive. In Uruguay, P. l. larvae has been detected in both larvae and adult bee samples (Piccini & Zunino, 2001) and even in honey (Antúnez et al., 2004).
In this study, P. l. larvae spores were detected in S3 samples, corresponding to the summer of 2002. This finding correlates with
an observed increase of AFB incidence in Uruguay during the
last year, particularly in the region where the sampled hives were
located.
Knowledge about the microbial community of the beehive can
be important for the development of biological control strategies based on the use of native bacterial strains. The high incidence of Bacillus species observed among our isolates is an interesting finding that could be taken in account in order to reach
this objective. A preliminary survey using different Bacillus spp.
strains from this study showed that some of them have inhibitory activity against P. alvei (data not shown). Moreover, it could
be important not only to study in more detail the relationship
among the different bacterial species, but also the response of
the bacterial community to different environmental factors.
Acknowledgements
We are very grateful to Eduardo Corbella for his assistance with different aspects of
this work. We also thank Laura Zunino (B-Z Laboratory, Uruguay) for her assistance
with some aspects of bacterial identification. This study was partially supported by
Grant INIA-FPTA 089, Uruguay.
REFERENCES
ALIPPI, A M (1992) Characterization of Bacillus larvae White, the causative agent of
American Foulbrood of honey bees. First record of its occurrence in Argentina. Revista Argentina de Microbiología 24: 67–72.
ALIPPI, A M (1995) Detection of Bacillus larvae spores in Argentinian honeys by using
a semi-selective medium. Microbiología SEM 11: 343–350.
ALLEN, M F; BALL, B V (1993) The cultural characteristics and serological relationships of isolates of Melissococcus pluton. Journal of Apicultural Research 32:
80–88.
Piccini, Antúnez, Zunino
ANTÚNEZ, K; D´ALESANDRO, B; PICCINI, C; ZUNINO, P (2004) Paenibacillus larvae larvae spores in honey samples from Uruguay: a nationwide survey. Journal of Invertebrate Pathology 86: 56–58.
CONNOR, K M; QUIRIE, M M; BAIRD, G; DONACHIE, W (2000) Characterization
of United Kingdom isolates of Corynebacterium pseudotuberculosis using
Pulsed-Field Gel Electrophoresis. Journal of Clinical Microbiology 38: 2633–2637.
DJORDJEVIC, S P; FORBES, W A; SMITH, L A; HORNITZKY, M A (2000) Genetic
and biochemical diversity among isolates of Paenibacillus alvei cultured from
Australian honeybee (Apis mellifera) colonies. Applied and Environmental Microbiology 66: 1098–1106.
EVTUSHENKO, L I; DOROFEEVA, L V; SUBBOTIN, S A; COLE, J R; TIEDJE, J M (2000)
Leifsonia poae gen. nov., sp. nov., isolated from nematode galls on Poa annua,
and reclassification of Corynebacterium aquaticum Leifson 1962 as Leifsonia
aquatica (ex Leifson 1962) gen. nov., nom. rev., comb. nov. and Clavibacter xyli
Davis et al. 1984 with two subspecies as Leifsonia xyli (Davis et al. 1984) gen.
nov., comb. nov. International Journal of Systematic and Evolutionary Microbiology 1: 371–80.
FARMER III, J J; DAVIS, B R; HICKMAN-BRENNER, F W; MCWHORTER, A; HUNTLEY-CARTER, G P; ASBURY, M A; RIDDLE, C; WATHEN-GRADY, H G;
ELIAS, C; FANNING, G R; STEIGERWALT, A G; O’HARA, C M; MORRIS,
G K; SMITH, P B; BRENNER, D (1985) Biochemical identification of new
species and biogroups of Enterobacteriaceae isolated from clinical specimens.
Journal of Clinical Microbiology 21: 46–76.
GERHARDT, P; MURRAY, R G E; WOOD, W A; KRIEG, N R (1994) Methods for general and molecular bacteriology. American Society for Microbiology Press;
Washington, DC, USA; 791 pp.
GILLIAM, M (1978) Bacteria belonging to the genus Bacillus isolated from selected
organs of queen honey bees, Apis mellifera. Journal of Invertebrate Pathology 31:
389–391.
GILLIAM, M (1979) Microbiology of pollen and bee bread: the genus Bacillus. Apidologie 10: 269–274.
HEYNDRICKX, M; VANDEMEULEBROECKE, K; HOSTE, B; JANSSEN, P; KERSTERS,
K; VOIS, P DE; LOGAN, N A; ALI, N; BERKELEY, R C W (1996) Reclassification of Paenibacillus (formerly Bacillus) pulvifaciens (Nakamura 1984) Ash et
al. 1994, a later subjective synonym of Paenibacillus (formerly Bacillus) larvae
(White 1906) Ash et al. 1994, as a subspecies of P. larvae, with amended
description of P. larvae as P. larvae subsp. larvae and P. larvae subsp. pulvifaciens. International Journal of Systematic Bacteriology 46: 270–279.
HORNITZKY, M A Z; NICHOLLS, P J (1993) J medium is superior to sheep blood
agar and brain heart infusion agar for the isolation of Bacillus larvae from
honey samples. Journal of Apicultural Research 32: 51–52.
JEYAPRAKASH, A; HOY, M A; ALLSOPP, M H (2003) Bacterial diversity in worker
adults of Apis mellifera capensis and Apis mellifera scutellata (Insecta:
Hymenoptera) assessed using 16S rRNA sequences. Journal of Invertebrate
Pathology 84: 96–103.
KIM, D S; COOK, R J; WELLER, DM (1997) Bacillus sp. L324-92 for biological control
of three root diseases of wheat grown with reduced tillage. Phytopathology
87: 551–558.
LAWSON, P A; WACHER, C; HANSSON, I; FALSEN, E; COLLINS, M D (2001) Lactobacillus psittaci sp. nov., isolated from a hyacinth macaw (Anodorhynchus
hyacinthinus). International Journal of Systematic and Evolutionary Microbiology 51:
967–970.
MANNING, R (2001) Fatty acids in pollen: a review of their importance for honey
bees. Bee World 82: 60–75.
PICCINI, C; ZUNINO, P (2001) American foulbrood in Uruguay: isolation of Paenibacillus larvae subspecies larvae from larvae with clinical symptoms and adult
honeybees, and susceptibility to oxytetracycline. Journal of Invertebrate Pathology 78: 176–177.
ROZALSKY, A; SIDORCYK, Z; KOTELKO, K (1996) Potential virulence factors of Proteus bacilli. Microbiology and Molecular Biology Reviews 61: 65–89.
SNOWDON, J A; CLIVER, D O (1996) Microorganisms in honey. International Journal
of Food Microbiology 31: 1–26.
SPIVAK, M; GILLIAM, M (1998) Hygienic behaviour of honey bees and its application
for control of brood diseases and varroa. Bee World 79: 124–134.
TICHY, J; NOVAK, J(2000) Detection of antimicrobials in bee products with activity
against viridans streptococci. Journal of Alternative and Complementary Medicine 6: 383–389.
WAARD, R DE; SNEL, J; BOKKEN, G C A M; TAN, P S T; SCHUT, F; HUIS IN’T VELD,
J H J (2002) Comparison of faecal Lactobacillus populations in experimental
animals from different breeding facilities and possible consequences for probiotic studies. Letters of Applied Microbiology 34: 105–109.