Experimental

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
The prevalence of Acarapis woodi in Spanish honey bee (Apis mellifera) colonies
Encarna Garrido-Bailón
a
, Carolina Bartolomé
b
, Lourdes Prieto
c
, Cristina Botías
a
,
Amparo Martínez-Salvador
d
, Aránzazu Meana
e
, Raquel Martín-Hernández
a,f
, Mariano Higes
a,⇑
a
Bee Pathology Laboratory, Centro Apícola Regional (CAR), Junta de Comunidades de Castilla La Mancha, 19180 Marchamalo, Spain
b
Departamento de Anatomía Patolóxica e Ciencias Forenses, Grupo de Medicina Xenómica, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain
c
Instituto Universitario de Investigación en Ciencias Policiales (IUICP), Comisaría General de Policía Científica, DNA Laboratory, Madrid, Spain
d
Epidemiology Consultant, C/Puente la Reina, 28050 Madrid, Spain
e
Animal Health Department, Facultad de Veterinaria, Universidad Complutense, 28040 Madrid, Spain
f
Instituto de Recursos Humanos para la Ciencia y la Tecnología (INCRECYT). Fundación Parque Científico y Tecnológico de Albacete, Spain
h i g h l i g h t s
" A new polymerase chain reaction
(PCR) was developed to identify
Acarapis woodi.
" A great prevalence was detected in
Spain during the 2006 and
2007 years.
" Series doubts about the current
classification of Acarapis species
have arised.
g r a p h i c a l a b s t r a c t
a r t i c l e i n f o
Article history:
Received 3 February 2012
Received in revised form 25 July 2012
Accepted 28 August 2012
Available online 10 September 2012
Keywords:
Acarapis woodi
Apis mellifera
PCR
Colony-loss
Temperate area
a b s t r a c t
Acarapis woodi is an internal obligate parasite of the respiratory system of honey bees which provokes
significant economic losses in many geographical areas. The main aim of this study was assess the A.
woodi role in the ‘‘higher honey bee colony losses phenomenon’’ in Spain. Therefore, a new polymerase
chain reaction (PCR) was developed to amplify the mitochondrial cytochrome oxidase I gene (COI) and so
the actual prevalence of A. woodi in Spanish honey bee colonies in 2006 and 2007 was determined as part
of a wider survey. The results revealed a greater prevalence than expected in most of the geographical
areas studied where has been generally underestimated One problem encountered in this study was to
distinguish between A. woodi and other species (Acarapis dorsalis and Acarapis externus) at the molecular
level. Furthermore, the patterns of genetic divergence across sequences raised serious doubts about the
current classification of these organisms.
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
Acarapisosis is a disease of the adult honey bee (Apis mellifera)
caused by the microscopic tracheal tarsonemid mite, Acarapis woo-
di. This parasite was first identified on the Isle of Wight (England)
in 1919 and since then, it has been associated with high levels of
bee mortality and poor winter survival (McMullan and Brown,
2009; Otis and Scott-Dupree, 1992; Villa and Rinderer, 2008a,b).
A. woodi is an internal obligate parasite of the respiratory system,
0014-4894/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.exppara.2012.08.018
⇑
Corresponding author. Address: Bee Pathology Laboratory, Centro Apícola
Regional, Camino San Martín s/n, 19180 Marchamalo, Spain. Fax: +34 949 250 176.
E-mail addresses: [email protected] (E. Garrido-Bailón), carolina.bartolo-
[email protected] (C. Bartolomé), [email protected] (L. Prieto), [email protected]
(C. Botías), [email protected] (A. Martínez-Salvador), [email protected] (A.
Meana), [email protected] (R. Martín-Hernández), [email protected] (M. Higes).
Experimental Parasitology 132 (2012) 530–536
Contents lists available at SciVerse ScienceDirect
Experimental Parasitology
j our nal homepage: www. el sevi er . com/ l ocat e/ yexpr
Author's personal copy
which lives and reproduces primarily in the large prothoracic tra-
chea of the bee, although it can also be found in the head, and in
thoracic and abdominal air sacs (Giordani, 1965). This organism
feeds on the haemolymph of its host and it is a vector of several
honey bee viruses (Bailey, 1975, 1985, 1999; Collison, 2001; Kome-
ili and Ambrose, 1991). The pathogenic effects of A. woodi on indi-
vidual bees depend on the number of parasites within the tracheae
and they can be attributed to both mechanical injury and to phys-
iological disorders resulting from the obstruction of the air ducts,
lesions in the trachea walls and the depletion of haemolymph. In-
creased parasitic load causes the tracheal walls, which are nor-
mally white and translucent, to become opaque and discolored,
with blotchy black areas that are thought to be due to melanin
crusts (Giordani, 1964).
This parasite has been reported worldwide (Matheson, 1993),
except for Sweden Norway, Denmark, New Zealand, Australia and
the state of Hawaii (Sammataro et al., 2000), and it is recognized
as the cause of significant economic losses. However, little is
known regarding its actual prevalence and impact on honey bee
colonies in Mediterranean countries. The traditional diagnostic
methods are very time consuming and they are based on direct
visualization of A. woodi or its lesions in the tracheas (OIE, 2008).
Other methods as enzyme-linked immunosorbent assays (ELISA)
(Fichter, 1988; (Grant et al., 1993; Ragsdale and Furgala, 1987;
Ragsdale and Kjer, 1989) or methods based on the visualization
of guanine under ultraviolet light, the main end product of nitro-
gen metabolism in mites (Mozes-Koch and Gerson, 1997), are little
used in routine diagnostic. Recently, new molecular diagnostic
techniques were developed to identify A. woodi using a nested
PCR to amplify the COI gene (Evans et al., 2007) or different set
of primers to amplify the same gene (Kojima et al., 2011).
The aim of the present study was to determine the prevalence
of A. woodi in honey bee colonies in professional apiaries in Spain
in 2006 and 2007, using a PCR protocol that detects A. woodi in a
single step. In addition, as new information on A. woodi and related
mites, such as Acarapis externus or Acarapis dorsalis, has recently
become available at GenBank, we analyzed genetic distances to
determine whether the sequences deposited in GenBank were phy-
logenetically accurate. The results of these studies have important
implications not only for our understanding of the prevalence of
this mite and its influence on colony losses but also, for the phylo-
genetic classification of this genus.
2. Methods
2.1. New primers design
The mitochondrial COI gene was selected for PCR amplification
from A. woodi. All the A. woodi COI sequences available as of March
2010 in GenBank (http://www.ncbi.nlm.nih.gov/sites/entrez?db=
taxonomy) were compiled (EU190886, FJ603294.1, FJ603296.1
and GQ916565.1), although sequence FJ603295.1 was excluded
due to its short length. A multiple alignment of the four available
sequences was carried out using ClustalW (http://www.ebi.ac.uk/
clustalw/), which allowed us to identify polymorphic points and
to avoid these when designing the primers.
Primers were selected visually, ensuring that the resulting
amplicon was short enough to allow amplification in adverse con-
ditions (e.g. poorly preserved samples). The selected primers were:
forward primer (AcarFor) 5
0
-CGGGCCCGAGCTTATTTTACTGCTG-3
0
;
reverse primer (AcarRev) 5
0
-GCGCCTGTCAATCCACCTACAGAAA-3
0
.
The CG tails added to primers are shown in bold and they are
underlined). The expected size of the amplicon was 162 bp.
Primer suitability (G + C content and melting temperature) was
evaluated using the IDT OligoAnalyzer software (http://www.idtd-
na.com/analyzer/Applications/OligAnalyzer), and G or GC tails
were added to the 5
0
end of each primer to standardize the melting
temperatures of the primer sets. Potential primer interactions
(hairpin, homodimer, and heterodimer structures for the two
primers) were tested using the AutoDimer program (http://
www.cstl.nist.gov/div831/strbase//AutoDimerHomepage/AutoDi-
merProgramHomepage.htm). Species specificity was determined
by conducting a search for nearly exact matches using BLAST
(http://www.ncbi.nlm.nih.gov/BLAST/).
Once the specific annealing of primers was verified, PCR ampli-
fication and sequencing of three samples were performed. A gradi-
ent PCR (58 ± 5 °C) was performed to empirically determine the
annealing temperatures of the primer pairs.
2.2. Sample collection
Determining the prevalence of A. woodi was part of a wider sur-
vey designed to study ‘‘higher honey bee colony losses phenome-
non’’ in Spain, involving the study of many different pathogens.
The number of colonies to be sampled was calculated in relation
to the number of apiaries registered in 2003 (source: Spanish Min-
istry of Agriculture), with an expected prevalence of honey bee
depopulation and losses of 40% (CAR, unpublished data), a preci-
sion rate of 10% and a confidence level of 95%. The number of sam-
ples was subsequently distributed in proportion to the number of
apiaries in each region in which colonies were selected randomly
(Fig. 1). This cross-sectional study was carried out between spring
2006 and autumn 2007 and it involved a total of 1943 adult bee
samples distributed as follows: 630 honey bee colonies in spring
2006; 458 in autumn 2006; 526 in spring 2007; and 329 in autumn
2007.
All samples were submitted to the Bee Pathology Laboratory
(CAR) by beekeepers or veterinary services in charge of the bee-
keeping associations’ apiaries. Bee samples were stored aseptically
at À20 °C prior to testing. The parasitized honey bee samples used
as controls were sent to CAR by Dr. McMullan (Ireland) and Dr. José
Villa (USA).
Each apiary was geo-referred and they were linked with the
bioclimate belt described by Rivas-Martínez (1987) in the phyto-
geographical regions of the Iberian Peninsula as previously de-
scribed (Martín-Hernández et al., 2012) in relation to climatic
and altitudinal variables and vegetation attributes of the territory
(five belts in Mediterranean region: thermo-,meso-, supra-, oro-
and cryoromediterranean, from lower to higher altitude; and four
belts in the Eurosiberian region: colline, montane, subalpine and
alpine).
The distribution of A. woodi was related to the agroclimatic
information obtained and treated with Geographical Information
Systems (GIS, v. 9.0). Pathogens distribution and proportions found
were compared through Pearson Chi
2
(v
2
).
2.3. DNA extraction
The thorax and head of over 100 house honey bees were
crushed for 4 min at high velocity in 50 ml of MilliQ water using
a Stomacher machine (Stomacher 80 Biomaster – Seward) and
plastic bags equipped with a filter (to retain the honey bee exoskel-
eton). The macerates were centrifuged for 6 min at 800g, the super-
natant was discarded and the pellets were resuspended in 1 ml of
distilled water (PCR grade). An aliquot of each resuspended pellet
(150 ll) was placed in a 96-well plate (Qiagen) to which glass
beads had been added (2 mm diameter, Sigma). At least one blank
well with water alone was included for every 20 samples as a neg-
ative control of extraction. Positive and negative control samples
were included in the plates.
E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536 531
Author's personal copy
The plates were shaken in a TissueLyser machine (Qiagen) and
30 ll of ATL buffer (Cat. No. 19076, Qiagen) and 20 ll of Proteinase
K (Cat. No. 19131, Qiagen) were added to each well and incubated
overnight at 56 °C. DNA was subsequently extracted using a BS96
DNA Tissue extraction protocol in a BioSprint Workstation (Qia-
gen) and the plates were frozen at À20 °C until use.
2.4. Polymerase Chain Reaction
PCR amplification was performed with a Mastercycler ep gradi-
ent S (Eppendorf), under the following conditions: Fifty micro litter
reaction cocktail containing 25 ll of Fast Star PCR Master, (Cat. No.
04 710 452 001, Roche Diagnostic), 0.3 lM of each primer, 0.2 mg/
ml bovine serum albumin, 0.1% Triton X-100 and 5 ll of the DNA
template. The thermocycler program was set as follows: 95 °C
(10 min); 35 cycles of a 30 s denaturalization at 95 °C, a 30 s elon-
gation at 61.8 °C and a 45 s extension at 72 °C; and a final exten-
sion step at 72 °C for 7 min. For each PCR, templates
corresponding to positive (reference Acarapis spp., extracts as tem-
plate) and negative (H
2
O as template) controls were run along with
DNA extracted from the isolates.
PCR amplicons were analyzed in QIAxcel System (Qiagen), using
a QIAxcel DNA Screening Kit (Cat. No. 929004, Qiagen), analyzing
the negative controls (both extraction and PCR negatives) in paral-
lel to detect possible contamination.
2.5. Sequence analysis
All PCR positive products were initially purified with Qiaquick
PCR Purification Kit (Cat. No. 28104 Qiagen) in a QiaCube machine
(Qiagen) and they were later sequenced in both directions in an
Applied Biosystem Genetic Analyzer 3730 automated sequencer.
The sequence data obtained was checked visually using Chromas
1.43 software and it was then compared with the sequences depos-
ited in GenBank using BLAST.
2.6. Sequence divergence and phylogenetic analysis of Acarapis COI
data in GenBank
All overlapping Acarapis COI sequences available in GenBank as
on December 2011 were edited and aligned using Bioedit (http://
www.mbio.ncsu.edu/BioEdit/bioedit.html). In order to analyze
the maximum number of sites, three A. woodi (AB38409.1,
EU190886.1, FJ603295.1) and one A. externus (AB638410.1) se-
quences were excluded from the final alignment due to their short
length. To estimate the synonymous nucleotide divergence be-
tween sequences (K
S
) we applied the Nei–Gojobori method with
Jukes-Cantor correction (Nei and Gojobori, 1986) implemented in
MEGA4 (Tamura et al., 2007). A neighbor-joining tree was gener-
ated for synonymous sites with the same software using the
Nei–Gojobori method (Nei and Gojobori, 1986), with Jukes-Cantor
Fig. 1. Distribution of the prevalence of A. woodi in Spain, according to the bioclimatic belts described by Rivas-Martínez, 1987: Montane (C), Colline (D), supramediterranean
(G), mesomediterranean (H), termomediterranean (I).
532 E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536
Author's personal copy
correction for multiple hits. A bootstrap test (2000 replicates) was
performed to assess the reliability of the resulting phylogenetic
tree.
3. Results
3.1. PCR reproducibility, sensitivity and specificity
In this work, a new PCR protocol was developed on COI gene. In
temperature gradient PCR tests, the best amplification was ob-
tained at 59 °C. Indeed, this was the highest temperature at which
amplicons could be clearly seen and consequently, the risk of non-
specific amplification was minimized. Several primer concentra-
tions (from 0.2 to 0.6 lM) were also assessed and the best results
were obtained at 0.3 lM. All positive controls yielded amplicons of
the expected size (162 bp) and sequences.
Reproducibility was assessed by repeating the entire process
(from DNA extraction to PCR amplicons analyzed in QIAxcel Sys-
tem, Qiagen) in 10 samples 5 times. The same results were ob-
tained with each sample, demonstrating the reliability of the
method. In addition, infected bees (positive controls) from differ-
ent countries (Ireland and USA) were analyzed with our protocol,
and the expected amplicons and A. woodi COI sequences were ob-
tained in all cases.
3.2. A. woodi prevalence in Spanish samples collected in 2006 and
2007
A total of 274 house worker bee samples out of 1943 tested po-
sitive for A. woodi infection in 2006 (13%) and 2007 (15.5%) and no
significant differences were found between the two sampled years
(v
2
= 2,31; p = 0,13). When the seasonal average prevalence of A.
woodis was assessed, the Autumn prevalence in each year (Table 1)
was significantly high when compared with spring of the corre-
sponding year (Spring/Autumn 2006 v
2
= 5,86, p = 0,015; Spring/
Autumn 2007 v
2
= 7,57, p = 0,006), and no differences statistical
significant were found between spring or autumn of both years
(Spring 2006/2007, v
2
= 0,92; p = 0,337; Autumn 2006/2007,
v
2
= 1,93; p = 0,165).
A. woodi was detected at least once in all the regions included in
the study (Fig. 1). The prevalence was significantly higher in the
montane and supramediterranean belts (Table 2), corresponding
to the colder regions the in Spain, where the longest periods when
frost are possible. However, A. woodi as well was detected in hotter
areas of the country where the climate tends to be drier and hotter
than in northern areas.
All the amplicons generated were sent for sequencing and the
sequences obtained matched the A. woodi COI sequence. Indeed,
two example sequences were submitted to GenBank (accession
numbers HM213853 and HM213854). No other Acarapis spp., were
detected in these samples.
3.3. Suitability of the GenBank COI sequences
The analysis of the patterns of genetic differentiation across se-
quences raised serious doubts about the current classification of
these organisms (Table 3). In some cases the level of synonymous
divergence (K
s
) between sequences attributed to the same species
was greater than that observed between sequences assigned to dif-
ferent species. For example, the K
S
values for two A. dorsalis
(GQ916568.1 and GQ916567.1) and two A. externus (HQ243442.1
and GQ916566.1) COI sequences were 23% and 30%, respectively.
These values were greater than the synonymous divergence be-
tween GQ916566.1 (A. externus) and any of the three A. woodi se-
quences (22%). Indeed, K
S
values as large as those observed here
(in the order of 10% or greater) strongly suggest that these se-
quences belong to five different lineages.
Table 1
Prevalence and confidence interval (95%) of Acarapis woodi in Spain in the transverse study relative to the 1943 samples.
A. woodi (%)
Year n Annual Prevalence CI 95% Seasonal Prevalence CI 95%
2006 1089 13 10,99–15,09 Spring 11 8,4–13,5
Autumn 15,9 12,5–19,4
2007 854 15,5 12,97–17,94 Spring 12,7 9,8–15,7
Autumn 19,8 15,3–24,2
Table 2
Distribution of A. woodi in Spain according to the bioclimatic belts (Rivas-Martínez, 1987).
A. woodi
Bioclimatic belt Climatic characteristics n Positive Prevalence (%) v
2
p v
2
p
Mesomediterranean (H) T 17 to 13 °C, m 4 to À1 °C, M 14 to 9 °C, 338 27 8 – –
IT 350 to 210, H: X–IV. Semi-arid to hyper-humid
Termomediterranean (I) T 19 to 17 °C, m 10 to 4 °C, M 18 to 14 °C, 173 21 12,1 2,4
a
0,125
a
– –
IT 470 to 350, H: XII–II. Arid to humid
Colline (D) T > 12 °C, m>2 °C, M > 10 °C, 169 22 13 3,3
a
0,068
a
0,06
b
0,8054
b
It > 240, H: XI–IV. Sub-humid to Hyper-Humid
Montane (C) T 12 to 6 °C, m 2 to -4 °C, M 10 to 3 °C, 228 35 15,4 7,5
a
0,006
a
– –
IT 240 to 50, H: IX–VI. Sub-humid to Hyper-Humid
Supramediterranean(G) T 13to 8 °C, m À1 to À4 °C, M 9 to 2 °C, 263 47 17,9 13,
a
0,000
a
0,58
c
0,4454
c
IT 210 to 60, H: IX–VI. Semi-arid to Hyper-humid
Bold indicates statistical significance.
T = annual average temperature, m = average of minim temperature on the colder month. M = average of maximum temperatures on the colder month. IT: Termic
index = (T + m + M) Â 10. H: months (in Roman numbers) when frost is statistically probable to happen. Ombroclimate classification according to rainfall: Arid < 200 mm,
Semi-arid 200–350 mm, Dry 350–600 mm, Sub-humid 600–1000 mm, Humid 1000–1600 mm, Hyper-humid > 1600 mm.
a
Prevalence of A. woodi on each bioclimatic belt was compared with the lowest level (Mesomediterranean H).
b
Prevalence of A. woodi on Colline (D) was compared with Termomediterranean (I).
c
Prevalence in Supramediterranean (G) belt compared with Montane (C).
E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536 533
Author's personal copy
In contrast to the conflicting classification of the A. dorsalis and
A. externus datasets, A. woodi sequences (FJ603294.1, FJ603296.1
and GQ916565.1) in GenBank were reliably grouped, showed no
divergence (K
S
= 0.0) and were identical to those obtained in this
study (data not shown).
To illustrate this finding, we constructed a phylogenetic tree
(Fig. 2) that reflects both the consistency of the A. woodi clade,
showing identical sequences, and the incongruent names given
to the remaining species. Two very divergent groups of ‘‘A. exter-
nus’’ and ‘‘A. dorsalis’’ could be distinguished, indicating the exis-
tence of at least four different lineages represented by four
clades (HQ243433.1, GQ916566.1, HQ243435.1 and FJ603293.1).
Although GQ916567.1 and GQ916568.1 differ slightly from the
other sequences of the HQ243433.1 and HQ243435.1 clades,
respectively, the synonymous pairwise divergence (K
S
= 1.3%) be-
tween each one of them and their respective groups is compatible
with these sequences belonging to these clusters. However, the
possibility of being distinct entities cannot be excluded.
4. Discussion
This is the first extensive study of the prevalence of A. woodi in
Spain, and the only such study in a European country in the last
decade that uses an accurate and sensitive method that can be eas-
ily applied in the laboratory setting without the need for special-
ized or expensive equipment. In this first study, we do not
determine the percentage of parasitized worker bees in each posi-
tive colony due to the design of the study and the number of pos-
itives samples found, much higher than expected. This is of great
importance to assess the mite impact on the colony and the diag-
nostic method developed in this work we will do so in future stud-
ies of this parasite.
The most common method to detect A. woodi has been to visu-
alize mites in the tracheal trunk under a microscope following dis-
section of individual honeybees (Giordani, 1974). Several
modifications to this classic technique have been proposed over
the years, such as the incubation of discs cut from the thorax in
10% KOH (Shimanuki and Knox, 1991) or lactic acid, staining with
dyes like methylene blue (Peng and Nasr, 1985) and thiazol blue
tetrazolium to distinguish live from dead mites (Liu, 1995), the flo-
tation of mites (Camazine, 1985), large sampling trials using ELISA
(Fichter, 1988; Grant et al., 1993; Ragsdale and Furgala, 1987;
Ragsdale and Kjer, 1989), or detection of the presence of guanine
(Mozes-Koch and Gerson, 1997). However the molecular tech-
niques are often more sensitive, especially in cases of low parasite
loads.
One of the challenges in the present study was to distinguish A.
woodi fromother species of the same genus which also infect honey
bees using a molecular approach. In February 2012, there were six
COI sequences of A. woodi deposited in GenBank-without including
ours- (AB638409.1, EU190886.1, FJ603294.1, FJ603295.1,
FJ603296.1, GQ916565.1), six of A. externus (AB638410.1,
FJ603293.1; GQ916566.1; HQ243440.1 to HQ243442.1) and nine
of A. dorsalis (GQ916567.1, GQ916568.1, HQ243433.1 to
HQ243439.1). There were four additional COI sequences of A. woodi
(AB634837.1, AB634838.1, HQ162656.1, HQ162657.1), one of A.
dorsalis (HQ162658.1) and five of A. externus (AB634839.1,
HQ162659.1 to HQ162662.1), that did not overlap with the previ-
ous ones and were discarded from further analysis.
Sequencing of all the amplified products confirmed the infec-
tion by A. woodi, and these products were identical to all the COI
sequences found in GenBank from this organism. There were no
polymorphisms (K
S
= 0.0 among A. woodi COI sequences), and in-
deed, this is the only species of this genus that has been clearly
and correctly identified by several authors (Delmiglio et al. unpub-
lished; Evans et al., 2007; Kojima et al., 2011). By contrast, the ge-
netic distances between sequences attributed to A. externus
(GQ916566.1, HQ243440.1-HQ243442.1and FJ603293.1, Delmiglio
et al. unpublished data and Ward et al. unpublished data) and A.
dorsalis (GQ916567.1, GQ916568.1, HQ243433.1–HQ243439.1,
Delmiglio et al. unpublished data and Ward et al. unpublished
data), respectively, were greater than those observed between se-
quences assigned to different species. This indicates that the
nomenclature of these organisms is incompatible with their cur-
rent classification. As such, it is necessary to further clarify this is-
sue and to establish the true origin of these sequences, which
almost certainly belong to at least four different species.
The method described here is very accurate and sensitive. The
short size of our amplicon allows A. woodi to be detected even in
samples that are not perfectly preserved (e.g. those that contain
degraded DNA) and the limit of detection is close to that obtained
using nested PCR. We found that the prevalence of A. woodi in hon-
ey bee colonies was higher than expected in all Spanish geograph-
ical areas, including areas with climatic conditions not considered
suitable for the development of A. woodi. Moreover, the sequencing
of all the products amplified from our samples confirmed infection
by A. woodi alone, with no other mites detected in any of the sam-
ples studied.
Table 3
Synonymous divergence between Acarapis sequences in GenBank calculated using the Nei–Gojobori method with Jukes-Cantor correction.
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]
[1]
[2] 0.00
[3] 0.00 0.00
[4] 0.17 0.17 0.17
[5] 0.25 0.25 0.25 0.21
[6] 0.23 0.23 0.23 0.23 0.01
[7] 0.15 0.15 0.15 0.01 0.23 0.21
[8] 0.15 0.15 0.15 0.01 0.23 0.21 0.00
[9] 0.25 0.25 0.25 0.21 0.00 0.01 0.23 0.23
[10] 0.25 0.25 0.25 0.21 0.00 0.01 0.23 0.23 0.00
[11] 0.15 0.15 0.15 0.01 0.23 0.21 0.00 0.00 0.23 0.23
[12] 0.15 0.15 0.15 0.01 0.23 0.21 0.00 0.00 0.23 0.23 0.00
[13] 0.25 0.25 0.25 0.25 0.08 0.09 0.27 0.27 0.08 0.08 0.27 0.27
[14] 0.25 0.25 0.25 0.25 0.08 0.09 0.27 0.27 0.08 0.08 0.27 0.27 0.00
[15] 0.25 0.25 0.25 0.25 0.08 0.09 0.27 0.27 0.08 0.08 0.27 0.27 0.00 0.00
[16] 0.25 0.25 0.25 0.25 0.08 0.09 0.27 0.27 0.08 0.08 0.27 0.27 0.00 0.00 0.00
[17] 0.22 0.22 0.22 0.22 0.23 0.22 0.20 0.20 0.23 0.23 0.20 0.20 0.30 0.30 0.30 0.30
A. woodi: [1] FJ603294.1, [2] FJ603296.1, [3] GQ916565.1, [4] GQ916567.1; A. dorsalis. [5] HQ243438.1, [6] GQ916568.1, [7] HQ243439.1, [8] HQ243437.1, [9] HQ243436.1,
[10] HQ243435.1, [11] HQ243434.1, [12] HQ243433.1, [13] HQ243442.1; A. externus: [14] HQ243441.1, [15] HQ243440.1, [16] FJ603293.1 and [17] GQ916566.1.
534 E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536
Author's personal copy
The reports about prevalence of tracheal mites in Spain are
scarce. In a study performed between 1974 and 1979, years previ-
ous to the entry of Varroa destructor in the Iberian Peninsula, the
prevalence of A. woodi on clinical samples was estimated in an
18.4% and it was even considered as an endemic pathogen in same
areas (Gómez Pajuelo and Fernández Arroyo, 1979). In a study per-
formed in the South of Spain some years later (1990–1991), when
V. destructor had spread around the country, A. woodi was reported
in four out of 35 (11%) of colonies studied but only in the 2.5% of
samples analyzed (Orantes and García-Fernández, 1997). The low-
er prevalence in this study was linked to the acaricide treatments
against varroa for more than 10 years, and for that reason in last
decades tracheal mites were considered almost to be disappeared
in Spain (Centro Apícola Regional, data not shown). Despite these
previous studies were performed using direct observation of the
parasite and they are undoubtedly not as sensitive as molecular
detection, we can see that prevalence in same moments of our sur-
vey (e.g. spring 2007) is close to values previous to the entry of Var-
roa in Spain.
A Greek study using a similar method reported that the acara-
pisosis problem was self-correcting under their experimental con-
ditions, and a long-term decrease in the incidence of A. woodi from
1986 to 1995 was reported (Bacandritsos and Saitanis, 2004). In
agreement with previous data (Ruijter and Eijnde, 1997), these
authors confirmed the higher prevalence of the parasite during au-
tumn months, and a lower prevalence during hot and dry periods
(summer), a pattern that was exacerbated in conditions of high
temperature and low humidity. Similar results were found in our
study were the prevalence in autumn was higher than in spring.
In our study, although the higher prevalence was found in colder
regions, interestingly, we detected A. woodi DNA in samples from
warmer climatic belts of Spain, where the climate is traditionally
characterized by high temperatures and short rainy periods.
Parasitismof honey bees by tracheal mites continues to produce
problems for beekeepers (Villa and Rinderer, 2008a,b). Susceptible
colonies are often weakened or killed when tracheal mite popula-
tions augment (De Guzmán et al., 2006), and increased winter mor-
tality has been associated with high levels of tracheal mite
infestation in the autumn (Furgala et al., 1989). There are no reli-
able clinical signs to diagnose acarapisosis since the signs of infec-
tion are not specific and bees behave in much the same way as
those affected by other diseases or disorders (OIE, 2008), crawling
in the front of the hive and climbing blades of grass, unable to fly.
Dysentery may occur (OIE, 2008), causing beekeepers to confuse
these symptoms with those of nosemosis due to Nosema apis.
Although clinical outbreaks of acarapisosis were not frequent in
any of the areas sampled here, the detection of a higher prevalence
of mites than expected (regardless of the higher sensitivity of the
PCR) indicates that parasites apparently remain undetected by
classical microscopy methods. The National Spanish Control Pro-
gram for varroosis has enhanced Acarapis control through the reg-
ular application of miticides to honey bee colonies. Although no
reliable studies of the efficacy of miticide treatment exist, it ap-
pears that these treatments are sufficiently effective to prevent
clinical outbreaks, particularly since the prevalence we report
never exceeds 30%, considered to be the level at which colony
losses occur (Bailey, 1981).
The presence of A. woodi in colonies should not be underesti-
mated, especially when the percentage of parasitized bees is high
(Villa, personal communication). Indeed, honey bee colonies para-
sitized with Varroa destructor and A. woodi sustain considerably
greater mortality during winter months than uninfected colonies,
or those with just one mite species (Downey and Winston, 2001).
This scenario may also be applicable to colonies parasitized with
other pathogens such as Nosema ceranae, which is highly prevalent
in Spain (Higes et al., 2010; Martín-Hernández et al., 2007; Orantes
and García-Fernández, 1997), and that requires notification of the
OIE. Compared with the well-known effects of pathogens such as
V. destrutor or Nosema spp., A. woodi has been somewhat neglected,
as witnessed by the few research projects into Acarapis spp., that
receive funding and the limited publication of new scientific data
in this field. The development of new tools to enhance our under-
standing of this prevalent parasite will help to define the interac-
tions that occur between pathogens, as well as aiding the
identification of epidemiological factors influencing its prevalence
and determine the consequences of climate change on A. woodi.
5. Conclusions
Our results show that the prevalence A. woodi in our country is
comparable with that recorded before the massive application of
Fig. 2. Neighbor-joining tree (synonymous sites) of Acarapis sequences deposited in GenBank, generated with the Nei–Gojobori method with Jukes-Cantor correction. The
numbers on the branches indicate the bootstrap values.
E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536 535
Author's personal copy
acaricides for control of the varroosis. This result suggests a pro-
found reflection on the reasons that have caused this situation
and their direct influence on the health status of colonies of bees
in many areas of a Mediterranean country. Since the prevalence
of this pathogen has been generally underestimated, further stud-
ies should be conducted to determine if these levels are cause for
concern The patterns of genetic differentiation and phylogenetic
analysis of Acarapis sequences deposited in Genbank seriously
question the species assignment of these organisms, except in
the case of A. woodi, that is concordant in all studies.
Acknowledgments
The authors wish to thank the Spanish Beekeeper Association
for supplying the samples, and Dr. McMullan and Dr. Villa for the
control parasitized honey bee samples. Dr. Villa also made a con-
structive criticism of this work. This study was supported by fund-
ing from the Junta de Comunidades de Castilla-La Mancha
(Consejería de Agricultura and Consejería de Educación), INIA
(RTA2005-152 and RTA 2008-00020-C02-01, FEDER funding) and
INCRECYT, European Social Funds. We would like to thank to Vir-
ginia Albendea, Carmen Abascal, Carmen Rogerio and Teresa Cor-
rales for their technical support.
References
Bacandritsos, N.K., Saitanis, C.J., 2004. A field study on the long-term incidence of
Acarapis woodi in Greece. J. Apic. Res. 43, 21–26.
Bailey, L., 1975. Recent research on honeybee viruses. Bee World 56, 55–64.
Bailey, L., 1981. Honey Bee Pathology. Academic Press, London, UK.
Bailey, L., 1985. Acarapis woodi: a modern appraisal. Bee World 66, 99–104.
Bailey, L., 1999. The century of Acarapis woodi. Am. Bee J. 139, 541–542.
Camazine, S., 1985. Tracheal flotation: a rapid method for the detection of honey
bee acarine disease. Am. Bee J. 125, 104–105.
Collison, C.H., 2001. The pathological effects of the tracheal mite on its host. In:
Webster, T.C., Delaplane, K.S. (Eds.), Mites of the Honeybee. Dadant and Sons,
Hamilton, Illinois, pp. 57–71.
De Guzmán, L.I., Rinderer, T.E., Bigalk, M., Tubbs, H., Bernard, S.J., 2006. Russian
honey bee (Hymenoptera: Apidae) colonies: Acarapis woodi (Acari:
Tarsonemidae) infestations and overwintering survival. J. Econ. Entomol. 98,
1796–1801.
Downey, D.L., Winston, M.L., 2001. Honey bee colony mortality and productivity
with single and dual infestations of parasitic mite species. Apidologie 32, 567–
575.
Evans, J.D., Pettis, J.S., Smith, I.B., 2007. A diagnostic genetic test for the honey bee
tracheal mite. J. Apic. Res. 46, 195–197.
Fichter, B.L., 1988. ELISA detection of Acarapis woodi. In: Needham, G.R., Page, R.E.,
Jr., Delfinado-Baker, M., Bowman, C.E. (Eds.), Africanized Honey Bees and Bee
Mites. Ellis Horwood, Chichester, UK, pp. 526–529.
Furgala, B., Duff, S.R., Aboulfaraj, S., Ragsdale, D.W., Hyser, R.A., 1989. Some effects
of the honey bee tracheal mite (Acarapis woodi) on non-migratory honey bee
colonies in east central Minnesota. Am. Bee J. 129, 195–197.
Giordani, G., 1965. Laboratory research on Acarapis woodi Rennie, the causative
agent of acarine disease of honeybees (Apis mellifera L.) Note 2. Bull. Apic. 6,
185–203.
Giordani, G., 1974. Méthodes de diagnostic des maladies des abelles adultes.
Dianostic de l
´
acariose. Bull. Apic., 17 (In French).
Giordani, G., 1964. Rechercher au laboratoire sur Acarapis woodi (Rebbue), agent de
l
´
acariose des abeilles (Apis mellifera L.) Note 4.. Bull. Apic. 7, 43–60 (In French).
Gómez Pajuelo, A., Fernández Arroyo, M.P., 1979. Enfermedades de las abejas en
España. XXVII Congreso Internacional de Apicultura. Apimondia, Bucarest, pp.
357–361.
Grant, G., Nelson, D., Olsen, P., Rice, W.A., 1993. The ELISA detection of tracheal
mites in whole honey bee samples. Am. Bee J. 133, 652–655.
Higes, M., Martín-Hernández, R., Meana, A., 2010. Nosema ceranae in Europe: an
emergent type C nosemosis. Apidologie 41, 375–392.
Kojima, Y., Yoshiyama, M., Kimura, K., Kadowaki, T., 2011. PCR-based detection of a
tracheal mite of the honey bee Acarapis woodi. J. Invert. Pathol. 108, 135–137.
Komeili, A.B., Ambrose, J.T., 1991. Electron microscope studies of tracheae and flight
muscles of noninfested, Acarapis woodi infested, crawling honeybees (Apis
mellifera). Am. Bee J. 131, 253–257.
Liu, T.P., 1995. A rapid differential staining technique for live and dead tracheal
mites. Canadian Beekeeping 18, 155.
Martín-Hernández, R., Meana, A., Prieto, L., Martínez-Salvador, A., Garrido-Bailón, E.,
Higes, M., 2007. Outcome of colonization of Apis mellifera by Nosema ceranae.
Appl. Environ. Microbiol. 73, 6331–6338.
Martín-Hernández, R., Botías, C., Garrido-Bailón, E., Martínez-Salvador, A., Prieto, L.,
Meana, A., Higes, M., 2012. Microsporidia infecting Apis mellifera: coexistence or
competition. Is Nosema ceranae replacing Nosema apis? Environ. Microbiol. 14,
2127–2138.
Matheson, A., 1993. World bee health report. Bee World 74, 176–212.
McMullan, J.B., Brown, M.J.F., 2009. A qualitative model of mortality in honey bee
(Apis mellifera) colonies infested with tracheal mites (Acarapis woodi). Exp. Appl.
Acarol. 47, 225–234.
Mozes-Koch, R., Gerson, U., 1997. Guanine visualization: a new method for
diagnosing tracheal mite infestation of honey bees. Apidologie 28, 3–9.
Nei, M., Gojobori, T., 1986. Simple methods for estimating the numbers of
synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3,
418–426.
Office International des Epizooties: 2008 Manual of Standards for Diagnostic Test
and Vaccines. URL: <http://www.oie.int/eng/norms/mmanual/2008>.
Orantes, F.J., García-Fernández, P., 1997. A scientific note on the current low levels
of honey bee tracheal mite in southern Spain. Apidologie 28, 149–150.
Otis, G.W., Scott-Dupree, C.D., 1992. Effects of Acarapis woodi on overwintered
colonies of honey bees (Hymenoptera: Apidae) in New York. J. Econ. Entomol.
85, 40–46.
Peng, Y.S., Nasr, M.E., 1985. Detection of honey bee tracheal mites (Acarapis woodi)
by simple staining techniques. J. Invert. Pathol. 46, 325–331.
Ragsdale, D., Furgala, B., 1987. A serological approach to the detection of Acarapis
woodi parasitism in honey bees using an enzyme-linked immunosorbent assay.
Apidologie 18, 1–10.
Ragsdale, D., Kjer, K.M., 1989. Diagnosis of tracheal mite (Acarapis woodi Rennie)
parasitism of honey bees using a monoclonal based enzyme-linked
immunosorbent assay. Am. Bee J. 129, 550–553.
Rivas-Martínez, S., 1987. Memoria del mapa de series de Vegetación de España.
Serie Técnica. Ministerio de Agricultura, Pesca y Alimentación I.C.O.N.A.,
Madrid. URL: <http://www.marm.es/es/biodiversidad/servicios/banco-de-
datos-biodiversidad/informacion-disponible/index_vegetacion_pot.aspx>.
Sammataro, D., Gerson, U., Needham, G., 2000. Parasitic mites of honey bees: life
history, implications and impact. Annu. Rev. Entomol. 45, 519–548.
Shimanuki, H., Knox, D.A., 1991. Diagnosis of honey bee diseases. US Department of
Agriculture, Agriculture Handbook. No. AH-690, p. 53.
Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA 4: molecular evolutionary
genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599.
Villa, J.D., Rinderer, T., 2008a. Inheritance of resistance to Acarapis woodi (Acari:
Tarsonemidae) in crosses between selected resistant russian and selected
susceptible US honey bees (Hymenoptera: Apidae). J. Econ. Entomol. 101, 1756–
1759.
Villa, J.D., Rinderer, T.E., 2008b. Inheritance of resistance to Acarapis woodi (Acari:
Tarsonemidae) in crosses between selected resistant Russian and selected
susceptible US honey bees (Hymenorptera: Apidae). J. Econ. Entomol. 101,
1726–1759.
536 E. Garrido-Bailón et al. / Experimental Parasitology 132 (2012) 530–536