Determination of the status of the etiological agent of American foulbrood, Paenibacillus larvae, in Swaziland

ORIGINAL RESEARCH ARTICLE Determination of the status of the etiological agent of

American foulbrood, Paenibacillus larvae, in Swaziland

Dermot Cassidy1*, Teresa Goszczynska2, John Burnet3, Ulrike Hirschauer4, Solomon Gebeyehu1, Gudeta W Sileshi5 and Lise Korsten6 1USAID, P.O. Box 11218, Silver Lakes, Pretoria 0054, South Africa. 2Agricultural Research Council, Plant Protection Research Institute, Private Bag X 134, 0121 Queenswood, Pretoria, South Africa. 3Eswatini Swazi Kitchen Honey, P.O. Box 1137, Manzini, Swaziland. 4TechnoServe, Inc., P.O. Box 663, Ezulwini, Swaziland. 5World Agroforestry Centre (ICRAF), SADC-ICRAF Agroforestry Programme, Chitedze Agricultural Research Station, P.O. Box 30798, Lilongwe, Malawi. 6Department of Microbiology and Plant Pathology, University of Pretoria, 0002, Pretoria, South Africa. Received 28 January 2010, accepted subject to revision 1 June 2011, accepted for publication 7 July 2011. *Corresponding author: Email: [email protected]

Summary American foulbrood (AFB) is a cosmopolitan disease affecting both larval and pupal stages of honey bees. There are considerable doubts

about the true status of AFB in Africa and there is, indeed, some evidence that sub-Saharan Africa, until recently, was largely free of AFB.

Requirements for honey imports into South Africa are governed by a concern for the potential introduction of AFB. The study describes a cost

effective and simple methodology for science-based trade in honey from Swaziland into South Africa that complies with the guidelines

developed by the World Organisation for Animal Health (OIE).

Determinación del estado del agente etiológico de la loque

americana, Paenibacillus larvae, en Suazilandia Resumen

La loque americana (LA) es una enfermedad cosmopolita que afecta a los estadios de larva y pupa de la abeja de la miel. Hay dudas

considerables acerca del estado real de la LA en África y hay, de hecho, alguna evidencia de que África sub-sahariana estaba, hasta hace

poco, libre de LA. Los requisitos para la importación de miel a Sudáfrica están regidos por una cierta preocupación por la potencial

introducción de LA. Este estudio describe una metodología simple y efectiva para el comercio basado en la ciencia de miel desde Suazilandia

hasta Sudáfrica que cumple con los preceptos desarrollados por la Organización de Salud Animal.

Keywords: American foulbrood, honey, survey design, market access, sub-Saharan Africa

Journal of Apicultural Research 50(4): 284-291 (2011) © IBRA 2011 DOI 10.3896/IBRA.1.50.4.05

Introduction

American foulbrood (AFB), caused by the Gram positive sporulating

bacterium Paenibacillus larvae, is an infectious, highly contagious,

cosmopolitan disease affecting the larval and pupal stages of the

honey bee Apis mellifera and other Apis spp., occurring in most

countries where bees are kept (World Organization for Animal Health,

2009). Taxonomic studies of the two subspecies of Paenibacillus

larvae, P. larvae subsp. larvae and P. larvae subsp. pulvifaciens, have

led to the reclassification of these subspecies into a single species, P.

larvae (Kilwinski et al., 2004; Genersch et al., 2006; OIE, 2009). In

this paper the designation AFB specifically refers to the ERIC I (aβ,

ab, and Ab) genotypes of Paenibacillus larvae and excludes ERIC II

(AB), III and IV genotypes (Kilwinski et al., 2004; Genersch et al.,

2006; OIE, 2009).

From a scientific and phytosanitary perspective there are doubts

about the true status of AFB in Africa, and there is some evidence

that sub-Saharan Africa, until recently, was considered largely AFB

free (Matheson, 1996; Fries and Raina, 2003). Paenibacillus larvae

subsp. pulvifaciens, i.e. the now reclassified ERIC III and IV

genotypes of P. larvae, has previously been isolated in South Africa

and Zambia (Ash et al., 1993; WTO 2008). Hansen et al. (2003)

suggested that retail honey purchased in South Africa and Guinea

Bissau was contaminated with P. l. larvae spores, but other surveys

conducted for the detection of AFB in honey samples from Kenya,

Senegal, South Africa, Tanzania, Uganda, Zambia, and Zimbabwe

found no contamination with AFB spores in the honey, or clinical

symptoms of the disease in bee colonies (Fries and Raina, 2003).

Since 2000, South Africa has been extensively surveyed for the

presence of AFB and after initial negative results, its presence has

now been confirmed in the Western Cape Province of the country

(Baxter, 2009).

International trade in honey and honey bee products is regulated

by the World Trade Organization (WTO), which sets the rules of trade

between nations at a global level. The Agreement on the Application

of Sanitary and Phytosanitary Measures (SPS Agreement) sets

constraints on members' trade policies relating to a number of areas

including animal health. Under the SPS agreement, trade restrictions

are required to be science based and in the case of animals and

animal products, these are administered in detail by the World

Organisation for Animal Health (OIE) based in Paris. The OIE

publishes scientifically based rules and regulations on trade in animals

and animal products which are referred to as the Terrestrial Animal

Health Code (TAHC) for any given year. The TAHC used as the basis

of this study is that of 2009. The current requirements for honey imports into South Africa from

any part of the world, including sub-Saharan Africa, are irradiation of

10kGy for honey, or other bee products containing any of these

ingredients (NDA, 2008). Irradiation is required due to a current non-

declaration of freedom from AFB status in any part of sub-Saharan

Africa, and is used as a mitigation measure to prevent the spread of

AFB in honey sourced from potentially infected areas. Honey

subjected to this treatment loses its organic properties, has a reduced

shelf life, and often turns black. Under the rules laid out in the OIE-

TAHC, 2009 it is therefore necessary for Swaziland to develop and

implement an effective and efficient honey bee pest surveillance

programme as a necessary component of the science based system

governing trade in honey with South Africa. The OIE-TAHC, 2009 requirements for declaration of freedom

from AFB can follow two paths, either with the declaration of a

historically free status in compliance with the provisions of Chapter

1.4 of the OIE-TAHC 2009, or as the result of an eradication program.

Given negative results of a risk assessment within the provisions of

Chapter 1.4 of the OIE-TAHC 2009 there is technically no requirement

for any specific type of survey or surveillance programme to

determine presence or absence of AFB. Surveillance to demonstrate

freedom from disease or infection (Article 1.4.6.1 subsection ‘a’

Historically free) requires, however, that infection is not known to be

established in the wild and that annual surveys have been carried out.

In the case of Swaziland, there is no historical or scientific

evidence to show the status of AFB in the country. The objectives of

this study therefore were: 1. to develop a scientific methodology for

market access for honey without irradiation that complies with

Chapter 9.2.3 paragraph 1 of the OIE-TAHC 2009; and 2. to develop a

methodology as a model for honey market access and bee health

studies in other parts of sub Saharan Africa.

Materials and methods Field survey and sampling

A stratified sampling frame was constructed to ensure that the survey

covered all three geographical (Murdoch, 1968) and five ecological

(Acocks, 1988) zones within Swaziland. In addition, it was necessary

to account for the high prevalence of hives in the main highveld

production areas comprising the eucalyptus forests between Sandlane

and Nhlangano in roughly the area designated as Piet Retief sourveld,

as well as the eucalyptus forests in an area roughly centred on Piggs

Peak in the Piet Retief sourveld transition area. The sampling frame

was therefore a conventional stratified four-stage cluster design in

which the first stage consisted of sampling clusters (honey producing

company), the second stage consisting of vegetation type, the third

stage of district or producers association and the fourth stage of

individual hives owned by a beekeeper (Table 1). The total number of

samples collected was estimated to represent just over 6% of the

total of 1500 commercial hives in Swaziland.

The sampling strategy was based on the honey robbing practices

of bees where it has been shown that colonies near to an AFB

collapsed colony have a high chance of rapidly contracting AFB, at

least sub-clinically. Studies show that 75% of colonies located 500m

or less from an AFB collapsed colony contracted AFB and died of the

disease, 50% of colonies 1km from the AFB collapsed colony died of

AFB disease (De Graaf et al., 2001; Pernal and Melanopoulos, 2006;

Lindstrom et al., 2008a). Approximately half the hives surveyed were of the straight sided

Tanzania Top Bar design and contained combs with mixed brood and

honey, the remainder being conventional Langstroth hives. Based on

the findings of Lindström et al. (2008a; 2008b) AFB, if present in the

swarm, would be also be detected in the honey. A further assumption

is based on recent work which indicates that where colonies are

known to be infected, analysis of honey samples may be of lesser

sensitivity as only 86% of such samples may be positive for AFB

(Gillarda et al., 2008). Brood comb and honey was sampled by

examining the brood comb for dead or discoloured brood and then

cutting a 20cm2 section of the brood comb for packing in paper

sample bags. About 60 to 80 g of honey was collected from each hive

American foulbrood in Swaziland 285

and placed in capped plastic bottles. The colour of the honey was

noted, ranging from light yellow to dark brown, and consistency from

liquid to solid. The survey was conducted in September to November

2008 during the main honey harvesting season.

Statistical analyses

The possible distribution of AFB was assumed to follow either a

binomial or negative binomial distribution, i.e. binomial distributions

when presence or absence was recorded (binomial sampling), and the

negative binomial when counts (integer values) are recorded (Sileshi,

2008). Therefore in order to determine the validity of the results,

probability mass functions (pmf) and cumulative distribution functions

(cdf) were calculated for both negative binomial and binomial

distributions. In terms of OIE standards, the required rate of detection

is at a 95% confidence rate given a 5% prevalence rate. The basic

assumption is that AFB occurs in 5% of the population of bee hives.

The 97 honey samples were picked at random using the sampling

frame in Table 1.

A simulation study was also conducted to examine the

performance of the single-season site occupancy model (MacKenzie et

al., 2002). A wide range of scenarios was considered, as there was a

286 Cassidy, Goszczynska, Burnet, Hirschauer, Gebeyehu, Sileshi, Korsten

great deal of uncertainty as to the likely probabilities of occurrence

and detection that would be encountered in the field. A simple model

that assumes both the probability of occupancy and detection

probability was constant across all sites was used to generate the

data. Given the constraints on available resources, it was assumed

that two repeated surveys could be conducted on each farm. For the

simulation study, one thousand sets of data were generated for each

scenario, i.e. detection probability (0.2, 0.4, 0.6, 0.8 and 1.0), true

occurrence probability (0.2, 0.4, 0.6, 0.8 and 1.0) and number of sites

(N = 40, 60 and 83). The simulation study was done using software

PRESENCE (Hines et al., 2006).

Taxonomic tests on field collected honey and brood

The samples collected were tested for the presence of AFB. For

isolation of viable P. larvae spores from honey, the method of Alippi

(1995) was used with modifications. Honey container samples were

sealed in a plastic bag and placed in a water bath (50°C) for 5-10 min

(or until the solid honey liquefied), shaking gently to distribute any

spores present.

Each honey sample was diluted with an equal volume of sterile

distilled water and mixed gently. The honey samples were centrifuged

Strata level 1 2 3 4

Strata type Company Vegetation type District/producers association Hive numbers

Producer 1 Eswatini Swazi Kitchen (ESK)

Eucalyptus in Piet Retief sourveld Lujulwemvelo coop 10

Lowveld sour bushveld Imphalu 7

Lowveld sour bushveld Malkerns 5

Zululand thornveld Shewula 7

Citrus in Lowveld sour bushveld Ngononi 8

Lowveld Sidvokodvo 10

Sub-Total – ESK 47

Producer 2

Swazi Honey (Mavimbela brothers)

Eucalyptus forests in Northeastern

mountain sourveld Hhohho/Piggs Peak region 15

Eucalyptus forests in Northeastern

mountain sourveld Bulembu 5

Sub-Total – Mavembela 20

Producer 3 Bulembu Honey (Bulembu Ministries)

Eucalyptus forests in Northeastern

mountain sourveld Bulembu 30

Sub-Total - Bulembu 30

TOTAL (hives) 97

Table 1. Sampling frame constructed for the AFB survey in Swaziland.

at 10,000 g for 25-30 minutes at 25°C. The supernatant was

discarded and the pellets were suspended in 0.5-1 ml of sterile

distilled water. The re-suspended pellets were transferred into

separate Eppendorf tubes and were placed in a heating block at 80°C

for 10 min in order to kill vegetative cells of bacteria and yeasts. The

pellet suspensions were plated on J-agar medium made in the

laboratory supplemented with 30µg/ml nalidixic acid and 20µg/ml

pipemedic acid (both obtained from Sigma Chemical Co.; St. Louis,

MO, USA) and 5% CO2 atmosphere which inhibits the growth of many

microorganisms present in honey (Alippi, 1995). A volume of 0.1 ml of

the suspension was streak-plated on the J-agar medium, so as the

total volume of suspensions was 1 to 1.5 ml, each suspension was

plated on 10 to 15 J-agar plates. The type strains of Ab, ab and aβ genotypes and AB genotype P.

larvae were plated on J-agar to serve as positive controls (P. larvae

LMG 9820T (synonym P. larvae subsp. P. larvae, type strain), P. larvae

LMG 15974T (synonym P. larvae subsp. pulvifaciens, type strain) and

P. alvei LMG 13253T, type strain. Inoculated plates were incubated in

5% CO2 atmosphere in the incubator at 37°C for 4-7days.

Representative colonies from each sample were selected for initial

identification assessing colony shape and margins, microscopic

characterization, and standard biochemical tests (Alippi, 1995).

Bacteria were isolated from all samples. Representative suspect

colonies grown on J-agar plates were examined for initial identification

assessing colony shape and margins, microscopic characterization and

standard biochemical tests, growth at 20°C, Gram stain, catalase test,

starch hydrolysis, milk hydrolysis, acid production from mannitol, D-

glucose and salicin (Alippi et al., 2002). Some non-target bacteria

may, however, still be recovered. For example P. larvae LMG 9820T

produces small, white, circular colonies on J-agar within 4-5 days.

Another genotype of P. larvae LMG 15974T also grows on J-agar.

Colonies are very similar to that of P. larvae LMG 9820T. Some strains

of P. larvae produce orange to brown colonies, but this is not

considered a diagnostic feature (Genersch et al., 2006).

Biochemical characterization of selected isolates

The isolated strains were tested with the API 50CHE system

(bioMérieux Clinical Diagnostics; La Balme les Grottes, Montalieu

Vercieu, France) using the procedure recommended by the

manufacturers. The results of the API 50CHE tests were recorded

after 72 h of incubation at 37oC.

Molecular detection and identification by PCR

The method of isolation of P. larvae from honey, primers and the PCR

conditions were performed as specified by Bakonyi et al. (2003).

Paenibacillus larvae specific confirmation of AB genotype suspected

colonies was carried out by PCR, using the primers AF6f 5’- GCA AGT

CGA GCG GAC CTT GT -3’ and AF7r 5’- GCA TCG TCG CCT TGG TAA

American foulbrood in Swaziland 287

GC -3’ which amplifies a fragment (237 bp) of the 16S rRNA gene of

the Ab, ab and aβ genotypes of Paenibacillus larvae. DNA was

extracted from the selected colonies as described by Bakonyi et al.

(2003) with the same reaction conditions.

BioPCR

The method used was that of Schaad et al. (1995) which enhances

the sensitivity of PCR reaction. Bio-PCR detects living cells of

pathogens, those that could cause a disease, as bacterial colonies are

washed up from agar plates preceding the PCR reaction. A honey

extract was plated on J-agar supplemented with 30µg/ml nalidixic acid

and 20µg/ml pipemedic acid (both obtained from Sigma Chemical Co.;

St. Louis, MO, USA). Plates were incubated for five to seven days at

37oC in 5% CO2 atmosphere which inhibits the growth of many

microorganisms present in honey (Alippi, 1995). After incubation the

bacterial growth was removed from the agar plates and suspended in

sterile distilled water. This bacterial suspension was used as a

template in the PCR with primers targeting P. larvae.

Extraction of DNA

Genomic DNA was extracted from pure cultures by using the GenElute

Bacterial Genomic DNA kit (Sigma). The concentration of DNA was

measured with the NanoDroP. Extracted DNA was stored at -20oC

until needed.

Electrophoresis of PCR products

The PCR products were separated in a 1% agarose gel with ethidium

bromide, in 1 x TAE buffer at 100 V for 45-50 min. The bands were

visualized under UV light and photographed with a Kodak DS

electrophoresis documentation and analysis system, using the Kodak

Digital Science ID software program (Fig. 2).

Tests using the VITA® AFB diagnostic kits

Confirmatory tests of suspect colonies were carried out using VITA®

AFB diagnostic test kit (Vita (Europe) Ltd., Basingstoke, UK). Orange

colonies and brood from these hives tentatively identified as Ab, ab

and aβ genotypes of P. larvae were tested using the methodology as

described in the VITA® test kit leaflet by substituting colonies of

suspected Ab, ab and aβ genotypes of Paenibacillus larvae instead of

larvae with suspicious symptoms. Test colonies were deposited into

the extraction bottle and shaken vigorously for about 20 seconds to

mix them into the buffer. A VITA® test device was removed from the

foil pack and a sample pipetted from the bottle immediately after

shaking. Two drops were placed onto the sample well of the device

and the device kept horizontal until the extract was absorbed (c. 30

seconds) and a blue dye appeared in the viewing window. With the

appearance of the control line after 1-3 minutes the result was

recorded.

Cassidy, Goszczynska, Burnet, Hirschauer, Gebeyehu, Sileshi, Korsten

288

Fig. 2. The results of bio-PCR with eight Swazi honey samples from which bacteria were isolated. MWM, molecular weight marker; line 1,

positive control P. larvae LMG 9820T (synonym P. larvae subsp. P. larvae, type strain); line 2, positive control P. larvae LMG 15974T (synonym

P. larvae subsp. P. pulvifaciens, type strain); line 3, negative control P. alvei LMG 13253T , type strain; line 4, honey No 22; line 5, honey No

30; line 6, honey No 21; line 7, honey No 24; line 8, honey No 29; line 9, honey No 38; line 10, honey No 49; line 11, honey No 55; NC,

negative control, water.

Fig. 1. Average estimates of occurrence probability (solid line) and lower and upper 95% confidence bounds (dashed lines) from a simulation

study using various levels of detection probability, true occurrence probability (0.2, 0.4, 0.6 and 0.8) and number of sites (N) visited twice

(N = 40 first column), (N = 60 second column) and (N = 83 third column).

Results Detection simulations

The detection simulations were run using a residual valid sample

number corrected for this reduced sensitivity i.e. assuming that a total

of only 83 samples were collected. With the assumed prevalence rates

returns the required detection confidence rate of 95% at sample 58

for a negative binomial distribution.

The calculated confidence of detection at this prevalence rate is

thus well within the required OIE-TAHC 2009 norms. Alternatively

using the binomial equation based detection simulation software

inputted with the OIE-TAHC norms the following detection

probabilities were produced using PRESENCE2 software method as

described by Hines (Hines, 2006). A probability P =0.012 is returned

for 0 detections in 83 samples based on the assumption of an AFB

prevalence of 5%. A 95% confidence of detection rate requires 59

samples which are close to those calculated for the negative binomial

distribution above.

Fig. 1 presents results of the simulation study. These suggest that

when the true probabilities of occurrence is low (0.2) and detection

probability is also low (0.2-0.4), occurrence will be overestimated by

30-70%, even with large number of sites (83 sites). Similarly, when

the true probabilities of occurrence is 0.4 and detection probability is

also low (0.2-0.4), occurrence will be overestimated by 10-41%. At

these low levels of occurrence and detection probability, the standard

errors will also be large and erratic (Fig. 1). The simulation results

seem generally stable for true occurrence probabilities of 0.8 and

detection probability of 0.6 or more. When the true occurrence

probability is 0.6 or more, even with low detection probability levels

and fewer sites (as few as 40) can give reasonable estimates of the

average occurrence probability. These results highlight the fact that at

low levels of AFB prevalence (and if the method used cannot ensure

up to 60% detection), there will be huge false positive results. When

detection probability is high (1.0), there was also about 1.0% increase

in false negative values.

Taxonomic tests

All suspect colonies were purified on J-agar and identity confirmed

using biochemical tests. Bacterial growth in 39 samples was observed

on J-agar. From three samples, however, an almost pure culture of

yellow, round colonies was isolated, and nine colonies from these

samples were purified on J-agar for additional testing. None of these

bacterial colonies produced a biochemical profile similar to P. larvae

LMG 9820T. Three representative strains that were catalase negative

were, however, used in the API 50 CHE biochemical strips. The three

isolates produced biochemical profiles that did not resemble that of P.

larvae LMG 9820T. Bacterial growth of 39 samples from J-agar plates

showing presumptive pathogen growth was tested by bio-PCR with

negative results. Electrophoresis results of PCR products were

American foulbrood in Swaziland 289

negative for AFB (Fig. 2). Tests using the VITA® AFB diagnostic kits

were negative when tested with suspect colonies and gave weakly

positive, but definite reactions, to the type strains of AFB.

Discussion Despite the adoption and publication of AFB surveillance standards by

the OIE, there is no evidence in the literature that any current honey

bee disease surveillance programme in the world has ever been

designed to meet these standards. New Zealand finalized a Pest

Management Strategy (PMS) for AFB, and reported the results of the

national programme but the design does not explicitly follow OIE

guidelines (Anonymous, 2008). The Swaziland survey for AFB is

therefore the first such formal exercise ever conducted according to

OIE guidelines for the application of OIE-TAHC 2009 to establish a

path for future trade in honey, particularly within sub-Saharan Africa

where the current status of AFB is a concern.

The results of the survey reported here demonstrate that AB

genotype P. larvae is absent from Swaziland using the criteria

stipulated by the OIE-TAHC-2009. This confirms studies that adult

bees from wild colonies in areas without intensive beekeeping rarely

contain detectable spore levels (Lindström, 2006). Furthermore, until

the recently reported outbreak in the Western Cape of South Africa

(Baxter, 2009), clinical cases of AFB have never been found in wild

honey bees south of the Sahara (Fries and Raina, 2003).

The reasons for the freedom of sub-Saharan Africa from AFB are

the subject of some speculation. A significant part of the problem is

the lack of long term honey bee systematic surveillance programmes

on the continent outside South Africa. Even in South Africa, AFB

surveillance has been conducted only since 2000, and the results have

not been published in peer-reviewed journals. Given the lightly

regulated nature of African trade in honey, bees and used bee

keeping equipment, it is likely that AFB has been introduced many

times to the continent. The recent outbreak of AFB in the Western

Cape of South Africa shows, however, that the continent’s bees are

not resistant to this disease and that appropriate surveys followed by

ongoing surveillance based on Article 1.4 of the OIE-TAHC, 2009 are

urgently necessary, not only in Swaziland but in other parts of sub-

Saharan Africa where bees are economically important as well as

being a key part of the ecology. As follow up activities to these survey results, a full review of

existing Swazi legislation, statutory instruments and regulations as

they address bee matters and bee health is underway. In addition the

design and implementation of a survey and surveillance programme

using the guidelines in the OIE TAHC, 2009 coupled with an

appropriate AFB management strategy if and when the disease arrives

in Swaziland, has been devised. A survey of Swaziland’s potential for

bees and honey production (forage, habitat etc) is urgently needed.

This together with information on land use patterns, human attitudes

(security of hives), the existing use of moveable frame hives, and

deficiencies in apicultural management knowledge in the country is

hoped to form the foundation for the sustainable management of

honey bee pests and diseases and development of the apicultural

industry.

Tests of presumptive colonies of AFB using the VITA® AFB

diagnostic kits have been demonstrated as a quick method of

screening laboratory cultures to determine the presence of AFB.

Acknowledgements Funding and technical assistance for this study was made possible by

a grant from USAID under their SPS support programme for the

Southern African Development Community (SADC). This is through

their Participatory Agency Services Agreement (PASA) with the United

States Department of Agriculture - Foreign Agriculture Service (USDA-

FAS).

References ACOCKS, J P H (1988) Veld types of South Africa. Memoirs of the

Botanical Society of South Africa 57. ISBN 0621 11394 8.

ALIPPI, A M (1995). Detection of Bacillus larvae spores in Argentinean

honeys by using a semi-selective medium. Microbiologica SEM 11:

343–350.

ALIPPI, A M; LOPEZ, A C; AGUILAR, O M (2002) Differentiation of

Paenibacillus larvae subsp. P. larvae, the cause of American

foulbrood of honey bees, by using PCR and restriction fragment

analysis of genes encoding 16S rRNA. Applied and Environmental

Microbiology 68(7): 3655–3660.

ANONYMOUS (2006) Situation analysis of beekeeping industry. Total

Transformation Agribusiness (Pty) Ltd. http://

www.beekeeping.com/articles/us/beekeeping_regional_situational

-analysis.pdf (accessed on 04 August 2011).

ANONYMOUS (2008) Report on the American foulbrood National Pest

Management Strategy. MAF Biosecurity New Zealand Discussion

Paper No: 2008/07, 51pp ISBN 978-0-478-32183-8 ISSN 1176-

8398.

ANONYMOUS (2009a) The outbreak of American Foul Brood (AFB) in

the Republic of South Africa, 25 March 2009, http://

www.info.gov.za/speeches/2009/09040109451001.htm, (accessed

on 04 April 2009).

ANONYMOUS (2009b) Animal health monitoring and surveillance

status of reportable diseases in the United States. http://

www.aphis.usda.gov/vs/nahss/disease_status.htm (accessed on

03 February 2009).

290 Cassidy, Goszczynska, Burnet, Hirschauer, Gebeyehu, Sileshi, Korsten

ANONYMOUS (2009c) Development of the VITA ® AFB Diagnostic Kit,

http://www.vita-europe.com/Map_enscript/frmbuilder.php?

dateiname=/en/products/afb_dignosyic.htm (accessed on 02

February 2009).

ASH, C; PROEST, F; COLLINS, M (1993) Molecular identification of

rRNA group 3 bacilli using a PCR probe test. Antonie van

Leeuwenhoek 64: 253-260.

BAKONYI, T; DERAKHSHIFAR, I; GRABENSTEINER, E; NOVOTNY, N

(2003) Development and evaluation of PCR assays for the

detection of Paenibacillus larvae in honey samples: comparison

with isolation and biochemical characterization. Applied and

Environmental Microbiology 69: 1504-1510.

BAXTER, A (2009). American foulbrood (AFB) in the Western Cape:

advisory notice. South African Bee Journal 81(1): 8-9.

BRØDSGAARD, C J; HANSEN, H (2003). Tolerance mechanisms

against American Foulbrood in honey bee larvae and colonies.

Apiacta 38: 114–124.

DE GRAFF, D C; VANDEKERCHOVEA, D; DOBBELAERE, W; PEETERS,

J E; JACOBS, F J (2001) Influence of the proximity of American

foulbrood cases and apicultural management on the prevalence of

Paenibacillus larvae spores in Belgian honey. Apidologie 32: 587-599.

FRIES, R; RAINA, S (2003) American foulbrood and African honey

bees (Hymenoptera: Apidae) Journal of Economic Entomology 96

(6): 1641-1646.

GENERSCH, E; FORSGREN, E: PENTIKÄINEN, J; ASHIRALIEVA, A;

RAUCH, S; KILWINSKI, J; FRIES, I (2006) Reclassification of

Paenibacillus larvae subsp. P pulvifaciens and Paenibacillus larvae

subsp. P. larvae as Paenibacillus larvae without subspecies

differentiation. International Journal of Systematic Evolutionary

Microbiology 56: 501-511.

GILLARDA, M; CHARRIEREB, J D; BELLOY, L (2008) Distribution of

Paenibacillus larvae spores inside honey bee colonies and its

relevance for diagnosis. Journal of Invertebrate Pathology 99(1):

92-95.

GOVAN, V A; ALLSOPP, M H; DAVISON, S (1999) A PCR detection

method for rapid identification of Paenibacillus larvae. Applied and

Environmental Microbiology 65(5): 2243–2245.

HANSEN, H; BRØDSGAARD, C J; KRYGER, P; NICOLAISEN, M (2003)

A scientific note on the presence of Paenibacillus larvae larvae

spores in sub-Saharan African honey. Apidologie 34: 471–472.

HINES, J E; MACKENZIE, D I (2006) PRESENCE2 – Software to

estimate patch occupancy and related parameters. USGS-PWRC.

http://www.mbr-pwrc.usgs.gov/software/presence.html (accessed

19 February 2009)

KILWINSKI, J; PETERS, M; ASHIRALIEVA, A; GENERSCH, E (2004)

Proposal to reclassify Paenibacillus larvae subsp. P. pulvifaciens

DSM 3615 (ATCC 49843) as Paenibacillus larvae subsp. P. larvae.

Results of a comparative biochemical and genetic study.

Veterinary Microbiology 104: 1-2, 30, 31-42.

LINDSTRÖM, A; KORPALA, S; FRIES, I (2008a) Horizontal

transmission of Paenibacillus larvae spores between honey bee

(Apis mellifera) colonies through robbing. Apidologie 39: 515-521.

LINDSTRÖM, A; KORPALA, S; FRIES, I (2008b). The distribution of

Paenibacillus larvae spores in adult bees and honey and larval

mortality, following the addition of American foulbrood diseased

brood or spore-contaminated honey in honey bee (Apis mellifera)

colonies. Journal of Invertebrate Pathology 99: 82–86.

MACKENZIE, D I; NICHOLS, J D; LACHMAN, G B; DROEGE, S; ROYLE,

J A; LANGTIMM, C A (2002) Estimating site occupancy rates when

detection probabilities are less than one. Ecology 83: 2248–2255.

MATHESON, A (1996) World bee health update. Bee World 77: 45-51.

MURDOCH, G (1968) Soils and land capability in Swaziland. Swaziland

Ministry of Agriculture. 360pp.

NDA (2008), South African National Department of Agriculture, Honey

Import Permit, Additional declaration.

OIE (2009) Terrestrial Animal Health Code (TAHC) 2009. http://

www.oie.int/eng/normes/Mcode/en_sommaire.htm (accessed 02

November 2009).

PERNAL, S F; MELANOPOULOS, A P; (2006) Monitoring for American

foulbrood spores from honey and bee samples in Canada. Apiacta

41: 99-109.

American foulbrood in Swaziland 291

SCHAAD, N W; TAMAKI, S; HATZILOUKAS, E; PANOPUOLAS, N (1995)

A combined biological amplification (BIO-PCR) technique to detect

Pseudomonas syringae pv. phaseolicola in bean seed extracts.

Phytopathology 85: 243-248.

SILESHI, G (2008) The excess-zero problem in soil animal count data

and choice of appropriate models for statistical inference.

Pedobiologia 52: 1-17.

WAITE, R; THOMPSON, H; BROWN, M; WATKINS, M; BEW, M (2003)

Preliminary studies into novel detection methods for honey bee

pathogens. Proceedings of XXXVIIIth International Apicultural

Congress, Ljubljana, Slovenia.

WORLD ORGANIZATION FOR ANIMAL HEALTH (OIE) (2009) World

Animal Health Information Database (WAHID) - Version: 1.3

http://www.oie.int/wahis (accessed 05 March 2009).

WORLD TRADE ORGANIZATION (2008) Information on various SPS

matters, Communication from Zambia, G/SPS/GEN/836 27 March

2008 (08-1369), Committee on Sanitary and Phytosanitary

Measures.