ADDIS ABABA UNIVERSITY

FACULTY OF VETERINARY MEDICINE

AFRICAN HORSE SICKNESS: STUDY ON SEROPREVALENCE AND

DENTIFICATION OF RISK FACTORS IN EQUIDAE AT SELECTED SITES IN

ETHIOPIA

A thesis submitted to the School of Graduate Studies of Addis Ababa University in the
partial fulfillment of the requirements for the Degree of Master of Science in Tropical
Veterinary Epidemiology

BY
KASSA DEMISSIE ABDI

JUNE 2006
DEBRE-ZEIT, ETHIOPIA

I
TABLE OF CONTENTS

PAGE

TABLE OF CONTENTS .................................................................................................... I

LIST OF TABLES .............................................................................................................. V

LIST OF FIGURES .......................................................................................................... VI

LIST OF ANNEXES ........................................................................................................ VII

LIST OF ABBREVIATIONS ....................................................................................... VIII

ACKNOWLEDGEMENTS ............................................................................................. IX

ABSTRACT ....................................................................................................................... XI

1. INTRODUCTION ........................................................................................................... 1

2. LITERATURE REVIEW ............................................................................................... 6

2. 1. The Epidemiology of AHS ........................................................................................... 6

2.1.1. The disease AHS ...................................................................................................... 6
2.1.2. Agent ........................................................................................................................ 7
2.1.3. Vector....................................................................................................................... 9
2.1.4. Host ........................................................................................................................ 10
2.1.5. Reservoir host ........................................................................................................ 11
2.1.6. Environment........................................................................................................... 12

2. 2. Experimental production .......................................................................................... 12

2. 3. Economic importance ................................................................................................ 13

2. 4. Human disease............................................................................................................ 13

2. 5. Pathogenesis ............................................................................................................... 14

2. 6. Clinical signs............................................................................................................... 15
2.6.1. “Dunkop” form ...................................................................................................... 15
2.6.2. “Mixed” form ......................................................................................................... 16
2.6.3. “Dikkop” form ....................................................................................................... 16
2.6.4. Horse sickness fever (abortive form) ..................................................................... 17

2.7. Pathology ..................................................................................................................... 17

2. 8. Diagnosis ..................................................................................................................... 18
2.8.1. Identification of the agent ...................................................................................... 18

II
 2.8.1.1. Samples for virus isolation ............................................................................. 19
2.8.1.2. Cell culture ..................................................................................................... 19
2.8 1.3. Newborn mice inoculation .............................................................................. 20
2.8.1.4. Polymerase chain reaction (PCR) .................................................................. 20
2.8.1.5. New approaches .............................................................................................. 21
2.8.2. Serological test ....................................................................................................... 21

2. 9. Differential diagnosis ................................................................................................. 22

2. 10. Treatment ................................................................................................................. 22

2. 11. Prevention and control ............................................................................................ 22

2.11.1. Endemic Areas ..................................................................................................... 23
2.11.1.1. Vaccination ................................................................................................... 23
2.11.1.2. Vector Control .............................................................................................. 23
2.11.2. Non-endemic Areas ............................................................................................. 24

3. MATERIALS AND METHODS .................................................................................. 25

3.1. Description and physiography of the study areas .................................................... 25

3.2. Geo-reference (GPS database) ................................................................................... 26

3.3. Target Population ....................................................................................................... 27

3.4. Sampling strategy ....................................................................................................... 28

3.5. Sample size determination ......................................................................................... 28

3.6. Study design................................................................................................................. 29

3.7. Study Methodologies................................................................................................... 29

3.7.1. Blood Collection and Assay Methodology ............................................................ 29
3.7.1.1. Seroprevalence survey .................................................................................... 29
3.7.1.2. Virus isolation and identification ................................................................... 32
3.7.1.2.1. Indirect-ELISA ............................................................................................. 33
3.7.2. Questionnaire survey ............................................................................................. 34

3.8. Data Management ....................................................................................................... 34

3.9. Data Analysis ............................................................................................................... 34

4. RESULTS ....................................................................................................................... 35

4.1. Seroprevalence survey ................................................................................................ 35

4.2. Virus Isolation and Identification ............................................................................. 41

4.3. Questionnaire Survey ................................................................................................. 41

5. DISCUSSION ................................................................................................................. 43

III
5.1. Seroprevalence Survey ............................................................................................... 43

5.2. Virus Isolation and Identification ............................................................................. 48

5.3. Questionnaire Survey ................................................................................................. 48

6. CONCLUSIONS AND RECOMMENDATIONS....................................................... 50

7. REFERENCES............................................................................................................... 52

8. ANNEXES ...................................................................................................................... 57

9. CURRICULUM VITAE................................................................................................ 62

10. SIGNED DECLARATION SHEET ........................................................................... 65

IV
LIST OF TABLES

Table 1. Isolated serotypes of AHSV from different areas of Ethiopia. ................................ 3

Table 2. Measured Geo-reference database of the study areas. ........................................... 26
Table 3. Number of sampled equidae in different age category .......................................... 28
Table 4. Serum sample collected from equines of different study areas. ............................ 29
Table 5. C-ELISA result of AHSV in different study areas. ............................................... 35
Table 6. C-ELISA result of AHSV in different types of equidae. ....................................... 36
Table 7. C-ELISA result of AHSV in different age categories of equidae. ........................ 37
Table 8. C-ELISA result of AHSV in the two sexes of equidae. ........................................ 38
Table 9. C-ELISA result of AHSV in different types of equidae and sexes. ...................... 38
Table 10. C-ELISA result of AHSV in different agro-ecological zones of the study areas.
..................................................................................................................................... 39
Table 11. C-ELISA result of AHSV in different types of equidae of different agro-
ecological zones .......................................................................................................... 40
Table 12. Multivariate logistic regression estimates for assumed risk factors of AHSV in
different types of equidae ............................................................................................ 41

V
LIST OF FIGURES

Figure 1. Map of the study areas.......................................................................................... 27

Figure 2. Plate layout for spot test ....................................................................................... 31
Figure 3. C-ELISA result of AHSV in different types of equidae ...................................... 36
Figure 4. C-ELISA result of AHSV in different age categories of equidae ........................ 37
Figure 5. C-ELISA result of AHSV in the two sexes of equidae ........................................ 39
Figure 6. C-ELISA result of AHSV in different agro-ecological zones of the study areas. 40

VI
LIST OF ANNEXES

Annex 1. Expression of results of spot test .......................................................................... 57

Annex 2. C-ELISA result of AHSV in equidae of each study area ..................................... 59
Annex 3. C-ELISA result of AHSV in different ages of equidae........................................ 60
Annex 4. Questionnaire format to assess the indigenous knowledge of equine owners ..... 60

VII
LIST OF ABBREVIATIONS

AHS African Horse Sickness

AHSV African Horse Sickness Virus
Anon Anonymous
C-ELISA Competitive Enzyme Linked Immunosorbent Assay
CFT Complement Fixation Test
CSA Central Statistical Authority
CACC Central Agricultural Census Commission
CI Confidence Interval
CPE Cythopathic/Cythopathogenic Effect
CTA Technical Centre for Agricultural and Rural Co-operation
DNA Deoxyribonucleic Acid
EARO Ethiopian Agricultural Research Organization
EEV Equine Encephalosis Virus
GMEM Glasgow Modification of Eagel’s Minimum Essential Medium
I-ELISA Indirect Enzyme Linked Immunosorbent Assay
MAbs Monoclonal Antibodies
MOA Ministry of Agriculture
NS Non structural Protein
NVI National Veterinary Institute
OAU Organization of African Unity
OCG Oxalate, Carbolic acid and Glycerin
OD Optical Density
OIE Office of International Des Epizooties
OR Odds Ratio
PAbs Polyclonal Antibodies
PBS Phosphate Buffered Saline
PCR Polymerase Chain Reaction
RNA Ribonucleic Acid
RT-PCR Reverse Transcriptase PCR
TCID50 Tissue Culture Infective Dose Fifty
VNT Virus Neutralization Test
VP Viral Protein

VIII
ACKNOWLEDGEMENTS

First and foremost, I would like to express to the Almighty GOD, sustainer and lord of the
universe for his innumerable favours and benevolence to me, during the most difficult time
of my MSc studies. Special spritual respect and thanks also goes to his mother Saint Mary.

The moral support, encouragement and patience of my wife, Tenagne Molla has no
parallel, it was an opportunity for me to witness once again her determination in support of
my advancement.

My deepest gratitude goes to the School of Graduate Studies of Addis Ababa University,
Faculty of Veterinary Medicine for providing me with the opportunity to pursue my
postgraduate studies here in my homeland, for financing the research undertaking, for
paying the import tax on the C-ELISA kit, and allowing me to use all facilities.

It is a very good opportunity for me to forward my special thanks and respect to Dr. Kelay
Belihu, Associate Dean for research and graduate programmes, for his excellent leadership
quality of the postgraduate programmes.

The DONKEY SANCTUARY WORKING WORLD WIDE deserves my heartfelt special

appreciation for making available the C-ELISA kit for the detection of African horse
sickness virus antibodies serotype 9, with complete bench protocol, without whom the
realization of this study would have not been possible. The role played by Dr. ANDREW
TRAWFORD and Professor FESEHA GEBREAB in facilitating the purchase of the kit
from BDSL in Scotland is an ever-lasting print in my mind.

The allround and enthusiastic support of my advisors in guiding this work and correcting
my draft thesis to shape it to its present form is highly acknowledged. For these I am
indebted to my academic advisors: Dr. Yilkal Asfaw, Professor Feseha Gebreab and Dr.
Berhe Gebreegziabher.

IX
The very vital all rounded technical support of the National Veterinary Institute with free
access of all laboratory facilities has enabled me to conduct my work smoothly and
according to established time table. The technical support given by w/ro Almaz Sharew,
Ato Kinfe Mekuria, Ato Berhanu Beyene, Dr. Fekadu Kebede, Dr. Ahmed Issa and all
other staff members of the research and diagnostic department has been immense and very
useful and therefore sincerely appreciated.

Last but not least my sincere thanks goes to Drs. Boja Endebu, Fikre Lobago, Gelagay
Ayelet, Belay Mulate, Shihun Shimeles, staff members of the Donkey Sanctuary, the
Society for the Protection of Animals Abroad (SPANA), Ato Asefa Mekonnen, staff of the
Kombolcha Veterinary Regional Laboratory, staff of the Federal Animal Health
Department, staff of district veterinary clinics from office of Agriculture and Rural
development in the study areas, administrative bodies, friends in the study areas, staff of
the National Veterinary Institute, staff of the Faculty of Veterinary Medicine of Addis
Ababa University, Equine owners of the study areas have also provided useful inputs to
which I am grateful.

X
ABSTRACT

A study was undertaken to determine the seroprevalence of African horse sickness virus
antibodies, isolate and characterize the virus responsible as well as identify potential risk
factors in the equine population of selected study areas in Ethiopia. In total 1265 serum
samples originating from 824 donkeys, 383 horses and 58 mules were collected from
September 2005 to mid of April 2006. Competitive Enzyme Linked Immunosorbent Assay
(C-ELISA) configuration was employed to determine the presence of AHSV antibodies.
The apparent prevalence of AHSV was found to be 29.7% (95% CI = 26.8-33.0) in
donkeys, 10.4% (95% CI = 7.8-14.0) in horses and 10.3% (95% CI =4.8-22.1) in mules.
The overall apparent seroprevalence of AHSV was found to be 23% (95% CI=20.8-25.4).
There is significant variation amongst the types of equidae in seropositivity (P<0.05).
Statistically significant (P<0.05) difference in seroprevalence was observed in the different
study areas, confirming the existence of agro-ecology based variation in the occurrence of
African horse sickness. The highest seroprevalence of AHSV was documented in the
lowlands followed by midland and highland areas. This has direct correlation with the
ecological distribution of the Culicoides vectors. As for age dependent variation in
seroprevalence no statistical significant difference was found. All age groups as well as
male and female populations were equally affected. The risk of acquiring AHS is more
than two fold (OR = 2.1) with respect to the types of equidae affected. Moreover, agro-
ecology contributes nearly two fold (OR =1.5) for the occurrence of African horse
sickness. There is strong association among C-ELISA result of AHSV antibodies, types of
equidae and agro-ecology, but age is not part of the interaction. However, sex has weak
effect to precipitate the occurrence of African horse sickness. Active disease search was
conducted with the aim of virus isolation and identification. After three blind passages
were carried out on vero cell lines the sample was subjected to I-ELISA configuration.
However, the result was negative. In the presence of the disease in the field with classical
pathognomonic signs and postmortem lesions the negative result is probably due to the
improper handling of the tissues processed. The indigenous knowledge base of equine
owners about African horse sickness in the study areas was assessed through a structured
questionnaire format. The survey result indicated that, the indigenous knowledge of
owners was found to be unsatisfactory.
Key words: Equidae, AHSV, Seroprevalence, Risk factors, C-ELISA, I-ELISA, Selected
areas, Ethiopia, Virus Isolation and Identification, questionnaire.

XI
1. INTRODUCTION

Ethiopia has the largest equine population, probably with the highest density per square
kilometer in the world (Alemayehu, 2004) and it has a total of 6.9% of the world’s and
42.4% of Africa’s equine population. Moreover, it has 65% of all African mules, almost
50% of horses and 30% of donkeys. Horses, mules and donkeys represent a significant
share of the working animal population of a number of countries in Africa (Wilson, 1995;
Feseha, 1998). According to the Central Statistical Authority of Ethiopia (CSA, 2004)
there are about 3.77 million donkeys, 1.45 million horses and 321,339 mules in Ethiopia.

There are 6.1 equines per square kilometer or 1 to 64 equines per hectare. There is also one
equine per four people in the agricultural sector and one equine per five people for the total
population of Ethiopia (Feseha, 1998).

An important but often overlooked aspect is that, in most cases, donkeys and mules are
reared by landless and marginal farmers and are the means of subsistence for millions in
the least privileged parts of the world. These beasts of burden receive little care and are
subjected to intensive work throughout their lives. Their life expectancy is short, their
work output is low and the reproductive performance is poor (Feseha, 1998).

Equids play an important role in the transportation of farm produce, fodder, firewood,
agricultural inputs, construction and waste materials. Equine power is used both in a rural
and urban transport system which is cheap and viable. It provides the best alternative in
places where the road network is insufficiently developed, or the terrain is rugged and
mountainous, and in cities where narrow streets prevent easy delivery of merchandise. The
importance of equids as working animals needs to be weighed within the prevailing
context of their socio-economic value, their demography and distribution, and the
implementation of disease control programmes based on strategic system derived from a
sound epidemiological knowledge. However, many factors contribute to the poor
performance of equids, among which viral diseases characterized by high morbidity and
mortality rates are to be mentioned. Hence, it is necessary to examine the present role of
these diseases to reappraise the existing control measures with a view to improve them and
attain better control strategies (Williams and Masiga, 1998; Feseha, 1998).

1
African horse sickness (AHS), also known as pernicious fever or typhoid fever, is a highly
fatal, infectious disease of horses, mules and donkeys which is caused by African horse
sickness virus. It is an endemic disease in equatorial, eastern and southern Africa from
which it regularly spreads south and periodically north either along the Nile River or along
the west coast of Africa. It appears to be truly an African disease (NVI, 1974; Mellor et al.,
1998).

African horse sickness virus is a double stranded RNA virus, which causes a
noncontagious, but infectious arthropod borne disease of equines and occasionally dogs.
Nine distinct internationally recognized serotypes of the virus have so far been identified.
All serotypes of African horse sickness virus occur in eastern and southern Africa, this
distribution reflects that of zebra, which cycles the virus asymptomatically (Mellor et al.,
1998). At present, African horse sickness virus (AHSV) is endemic in tropical and sub-
tropical areas of Africa south of the Sahara occupying a broad band stretching from
Senegal in the west to Ethiopia and Somalia in the east, Kenya and extending as far south
as northern part of South Africa (Mellor and Hamblin, 2004).

In Egypt the disease occurred in 1928, 1943, 1953, 1958, and 1971 in which all outbreaks
originated in the areas of Aswan and Qena Provinces, and the international boundaries
between Egypt and Sudan (Coetzer et al., 1994).

The hosts in order of decreasing severity of AHS are horses, mules, donkeys and zebras.
Horses and mules have the highest mortality, donkeys have a lower mortality, and African
donkeys have a subclinical infection. The horse is an amplifier of AHSV and source of
virus for arthropods. Increased use of irrigation, water leaks, manure, urine, dung pats, tree
holes, rotting vegetation, stagnant surface water are ideal larval habitats for the
multiplication of Culicoides (Mellor et al., 1998).

African horse sickness was first recognized in Eritrea around the 16th century. In 1904,
French travelers observed the disease in the lowlands of Ethiopia particularly around Baro
and Awash valleys. At the end of 1935, there was production of vaccine in Asmara against
AHS (A and B types) by the Animal Health Protection Research and Experiment group
and the vaccine was found to be effective in mules and donkeys of Tigray and Eritrea. This
vaccine was produced using guinea pigs (NVI, 1978). Between 1962-1968, 46 outbreaks of

2
African horse sickness were declared by OIE of which 27 occurred only in the year 1967.
Among the outbreaks, 29 occurred in June, July, August and September of the years. These
outbreaks occurred particularly in the Northern part including Tigray and Eritrea. In 1967,
269 sera were collected from horses of Sidamo and 12.8% were positive to AHS virus by
CFT (NVI, 1983).

In 1974 virus isolation and serotyping of AHSV was carried out in the National Veterinary
Institute of Debre Zeit. In this work, serotype 7 and 9 were isolated on vero cells from
pieces of spleen. Isolates were typed with specific sera. Serotype 7 was isolated for the first
time and this may necessitate NVI to produce a bivalent vaccine against AHSV in the
years to come (NVI, 1978).

Table 1. Isolated serotypes of AHSV from different areas of Ethiopia.

Host Technique
Month Place Serotyping Serology
January Awassa Horse 9 -
February Dire Dawa Horse 9 -
June Addis Ababa Horse 9 -
June Debre Zeit Horse 9 -
June Jimma Horse 7 -
August Jimma Horse 9 -
September Debre Zeit Horse 9 -
November Mekele Mule 9 -
November Mekele Mule 9 -
November Mekele Mule 9 -
November Mekele Mule 9 -
December Mekele Mule 9 -
Summary Debre Zeit Horse 9 All
serotypes
except 2 and
8
Jimma Horse 7 and 9 All
serotypes
except 8

Source: National Veterinary Institute, Debre Zeit (1974)

A report issued by NVI in 1978, indicated that AHS is prevalent in eight out of the
previous fourteen provinces of Ethiopia. Moreover, a paper released by the National
Veterinary Institute asserts that out of the 9 serological types recognized in the world, type
9 has been isolated from samples received from Awassa, Dire Dawa and Mekele

3
suggesting that AHS is spread in lowlands and midlands of Ethiopia. In addition to this,
some cases of the disease have been encountered in the highlands of the country including
Addis Ababa and Arsi. This information may indicate the existence of AHS in almost all
agro-ecological zones of Ethiopia (NVI, 1983).

An epidemiological study of AHS was conducted from 1977 to 1981 by the National
Veterinary Institute and the study came out with isolation of virus using live animals
including intracerebral inoculation of guinea pigs, mice, and vero cell culture (monkey
kidney) (NVI, 1983).

The study also involved in an assessment of the seasonal occurrence of the disease and the
results of the study indicated that most cases of the disease occurred in the rainy season
including June, July, August and September, but also some cases of the disease occurred at
the end of the rainy season including November (NVI, 1983).

Two research works were conducted in Wollo administrative region on the

pathomorphological and hematological changes in affected equidae due to African horse
sickness. The findings of the two research works indicated that AHS occurs every other
year and it is very difficult to rear horses in this area. Furthermore, they have confirmed
that both cardiac and pulmonary forms of the disease exist in the area (NVI, 1983).

The Virus Neutralization Test (VNT) indicated that two serotypes of AHSV were involved
in the outbreak occurred in 2002-2003 in Southern Ethiopia (Awassa, Hossana,
Wondogenet, and Hagereselam); Western Ethiopia (Jimma, Bedelle, Nekemte,
Horroguduru, and Chaliya) and Central Ethiopia (Debre Zeit, Meki, Zeway, Filtimo, and
Bekejo). Serotypes 9 and 6 were isolated from blood, spleen, and lymph nodes collected
from 12 sick and dead animals. The identification of serotype 6 represents the first report
in Ethiopia (Aschalew et al., 2005). The serotypes of AHSV are distinguished by
differences in virulence, immunogenicity, and production of different disease patterns. Of
the nine serotypes identified, type 9 is predominantly found throughout the African
continent, and it is the only serotype previously identified in Ethiopia (Aschalew et al.,
2005).

4
The study conducted by Aschalew and others (2005) at the National Veterinary Institute of
Ethiopia has shown the widespread occurrence of AHS across various agro-ecological
zones of Ethiopia. In countries such as Ethiopia, where large population of equines are
raised, the presence of multiple serotypes of such a devastating virus poses a serious
hindrance to the national development was furtherly pointed out by Aschalew and others
(2005).

According to the Ministry of Agriculture and Rural development, AHS is reported to occur
in most regional states of Ethiopia (personal communication, 2005). Therefore, the present
study is initiated to complement the previous findings in some of the study areas and
pondered over the following objectives:
► to determine the seroprevalence of African horse sickness virus in different agro-
ecological zones of the study areas,
►to undertake AHSV isolation and identification through active disease search in the
study areas, and
► to identify the potential risk factors associated with African horse sickness.

5
2. LITERATURE REVIEW

2. 1. The Epidemiology of AHS

2.1.1. The disease AHS

African horse sickness is a peracute, acute, subacute or mild infectious but noncontagious
disease of equines caused by an orbivirus, of which there are nine serotypes, all transmitted
by Culicoide midges (Gallowally, 1974; Sewell and Brocklesby, 1990; Coetzer et al.,
1994). It is a list A disease according to the OIE classification of diseases. (Mellor et al.,
1998).

AHS is endemic in Africa including Ethiopia affecting equidae, which are of paramount
importance to the livelihood of millions of farmers (Mellor and Hamblin, 2004). It is quite
a serious disease in Ethiopia and has reappeared in Senegal after being unrecorded since
1965 (Hall, 1985).

Among the equidae family, horses are the most susceptible to AHS with a mortality rate of
50-95%, followed by mules with mortality around 50%. In enzootic regions of Africa,
donkeys are very resistant to AHS and experience only subclinical infections. In Europeans
and Asian countries, however, donkeys are moderately susceptible and have a mortality
rate of 10%. Zebras are also markedly resistant with no clinical signs, and may have
extended viraemia, up to 40 days (Galloway, 1985; Hunter, 1994; OIE, 2004).

The disease has both a seasonal (late summer/autumn) and cyclical incidence with major
epizootics in Southern Africa during warm phase events. But in other areas it may occur
regularly if conditions are suitable for midges to be active all the year round. Mortality due
to AHS is related to the types of equidae affected and to the strain or serotype of the virus
involved (Galloway, 1985; Hunter, 1994; OIE, 2004). Where the disease is enzootic,
outbreaks often occur at intervals of 10 to 20 years, and this is considered to be due to the
periodic fluctuation/elevation in immunity, the yearly seasonal variation in vector
population and periodic redistribution of new or mutated strains of the virus (Hall, 1995;
Radostits et al., 2000). From time to time, AHS has spread beyond its usual distribution,
causing devastating epizootics in many countries, including the Middle East, Pakistan,
Afghanistan, India, North Africa, and the Iberian Peninsula, due to importation of infected

6
zebras into Spain in 1987 (Galloway, 1974; Anon, 1976; Della-Porta, 1985; Knipe and
Howley, 2001).

African horse sickness epidemics do not occur in the condition of heavy rain followed by
drought. However, the reversed pattern, drought followed by heavy rain favours epidemics.
The reasons why this pattern of drought followed by heavy rain is conducive to AHS are
unclear though these authors suggested that it might have some thing to do with the
congregation of zebra near the remaining waterholes during the drought period where they
come into contact with and infect more vectors which then disperse rapidly once rain
provides more breeding sites (Mellor and Hamblin, 2004).

2.1.2. Agent

African horse sickness virus is considered enzootic only in sub-saharan Africa, but on
several occasions it has caused devastating outbreaks in horses outside Africa. Since world
war-II, major epizootics have occurred in the Middle East and Indian sub-continent in
1959-1961 and in North Africa and Southern Spain in 1965-1966 (Fenner et al., 1987;
Sewell and Brocklesby, 1990).

African horse sickness is caused by a double stranded RNA virus. The virions are 70-80
nm in diameter and have an icosahedral symmetry. They contain 10 double-stranded RNA
genome segments encapsulated within a double layered capsid made up of 32 large ring
shaped capsomeres, each comprising 7 structural proteins and 4 nonstructural proteins
(Joklik, 1988; Seifert, 1996; Anon, 1998; OIE, 2004).

African horse sickness (peste equina africana, peste equine, pestis equorum) virus has
similar morphological and biochemical properties with the other members of orbiviruses
but with distinctive pathological and antigenic properties as well as host ranges. The
genome of African horse sickness virus encodes several structural proteins (VP1-7), most
of which have been completely sequenced for AHSV serotypes 4, 6 and 9 and four
nonstructural proteins (NS1, NS2, NS3, NS3A). Proteins VP2 and VP5 form the outer
capsid of the virion, and proteins VP3 and VP7 are the major inner capsid proteins.
Proteins VP1, VP4 and VP6 constitute minor inner capsid proteins. Recently, it was
indicated that NS3 proteins are the second most variable AHSV proteins, the most variable

7
being the major outer capsid protein, VP2. This protein, VP2, is also mainly responsible
for AHSV serotypes and, together with VP5, takes part in virus neutralization activity.
Nine antigenically distinct serotypes of AHSV have been identified by virus neutralization,
but no cross-reactions with other known orbiviruses have been observed (Hungerford,
1975; Della-Porta, 1985; OIE, 2004).

AHSV serotypes numbering up to 42 are suspected, of which the last was isolated in 1960.
Eventhough nine AHS viruses were recognized in South Africa, additional viruses are
suspected to occur elsewhere in Africa. Antisera which neutralize homologous virus may
cross-neutralize heterologous serotypes. There is cross-neutralization between serotypes 1
and 2, 3 and 7, 5 and 8, and 6 and 9. However, field evidence indicates that no significant
intratypic variation occurs (Galloway, 1974; Radostits et al., 2000; Seifert, 1996; OIE,
2004). There is no significant serologic or genetic relationship between AHSV and other
orbiviruses of veterinary importance (Fenner et al., 1999; Murphy et al., 1999).
.
The virus is relatively impervious to heat; it does not become inactivated after exposure to
temperatures of 55-75OC for 10 minutes and persists for 3 months at 4OC, but at -25OC a
marked reduction in virus titre occurs if the virus is not diluted in a stabilizer. Minimal loss
of titre occurs in lyophilized vaccine kept at 4OC. Putrid blood remains infective for more
than 2 years. The optimal pH for virus survival is 7.0-8.5 (Galloway, 1985; Seifert, 1996).
The virus is moderately resistant to external environmental influences such as drying and
heating (Radostits et al., 2000; Anon, 1998).

A high percentage of sera (79.8%) were positive, confirming the continued prevalence of
AHSV antibodies in Nigerian horses and donkeys (Adeyefa and Hamblin, 1995). In the
same country 75% of donkeys, 68% of horses, 19% of camels and 9% of dogs were found
to have antibodies against AHSV (Baba et al., 1992).

AHSV can be inactivated by repeated freezing and thawing, and by treatment with acetic
acid (inactivated at a pH of 6.3 or lower), remaining for 2 weeks at 370C, and being placed
for 5 minutes at 700C (Mellor et al., 1998). In OCG (5g of potassium oxalate, 5g of
carbolic acid (phenol), 500ml of water, and 500ml of glycerin), AHSV is stable for more
than 20 years at 40C (Mellor et al., 1998). Acetic acid (2%), iodophore disinfectants, and
chlorine dioxide disinfectants are all killers of AHSV (Mellor et al., 1998).

8
Serotypes 1-3 and 5-8 are confined mainly to Africa but serotypes 4 and 9 had occurred
outside Africa (Mellor et al., 1998). The virus is present in the body from the advent of
fever to the time of complete recovery (Hall, 1985).

2.1.3. Vector

Species of Culicoides have been conclusively identified as the major and biological vectors
because AHSV can replicate within them. In the United States, Culicoides that can
transmit blue tongue virus can most likely transmit AHSV. Culicoides are most active at
sunset and at about sunrise. Arthropods other than the Culicoides (e.g., biting flies and
mosquitoes) may spread the virus as mechanical vectors (Mellor et al., 1998, Knowles,
1991). At least two field vectors are involved: Culicoides imicola and Culicoides bolitonos
(Galloway, 1985; Hunter, 1994; OIE, 2004).

The spread of the disease is influenced by climatic conditions which favour the spread of
carrier insects (vectors) including warm, moist weather and high rainfall, as well as spread
by wind dispersal. It is likely that the virus persists (overwinters) in other unknown species
in Africa when the insect is not active. This explains why the disease does not persist in
other countries following an outbreak (Mellor et al., 1998).

The disease is spread by the passive transfer of very small quantities of blood by biting
insects, and spread does not occur between animals in direct contact unless the prerequisite
insect vectors are present. Climatic conditions which govern the breeding status and
movement of these insects also govern the spread and morbidity rate of horse sickness.
These insects have almost worldwide distribution so that spread of the disease could be
universal. Many other biting insects have been named as vectors but there is lack of
satisfactory evidence in many reports (Radostits et al., 2000; Anon, 1998). AHS occurs
during rainy warm seasons, which favour propagation of the vectors (Radostits et al., 2000;
Farouk, 1997).

An entomological survey was conducted with the objective of identifying the types of
Culicoides species prevailing in Ethiopia. Samples were collected from water meadows
near riverine woodland in all months of May to November. The efficiency of catching
Culicoides was poor, due partly to their smallness and darkness and also because the

9
weight of their numbers was sometimes overwhelming. C. grahamii Austen is a day biting
type with peaks shortly after dawn and before dusk; biting intensely during and after
showers at any time of the day. It is widely spread around the Rift Valley lakes;
particularly prolific at sites of geothermal springs and marshes beside Lake Shala and Lake
Abaya in the months of June to August (White, 1997).

C. milnei Austen is a night biting type and frequently abundant in animal shelters, resting
by day and biting at night. It is greatly abundant from May to November throughout Shoa,
parts of Arsi, Kaffa, Wollo, Gojjam and Gondar areas where main roads permit perennial
access (altitudes of 1,550-2,100m.). If it were genetically susceptible to infection with
relevant arboviruses, however, milnei would obviously rate as an outstanding vector of
veterinary disease throughout much of the Ethiopian highlands, where its abundance in
association with sheep, cattle and horses far exceeds the numerical densities encountered in
the Kenyan highlands (White, 1997).

2.1.4. Host

African horse sickness differentially affects members of the equidae family; horses are the
most susceptible, and donkeys and mules show mild symptoms (Galloway, 1974; Anon,
1976; Della-Porta, 1985; Knipe and Howley, 2001). African horse sickness causes diseases
in horses, mules and donkeys, with up to 95% mortality among susceptible horse
population (Anon, 1976; Robinson, 1987). Dogs fed infected raw horse meat have shown
signs similar to those observed in horses. Dogs may also become transient virus carriers
(Anon, 1976; Knowles, 1991). The host range of AHSV includes zebra, but clinical disease
in this species is unusual. The virus persists in zebra longer than in horses, suggesting that
it may be the original reservoir host (Fenner et al., 1987; Sewell Brocklesby, 1990).

The virus was isolated from blood of clinically healthy street dogs. Insectborne infection
probably never occurs in dogs, because insects are less attracted. Dogs, goats, mice, guinea
pigs and rats can be infected experimentally. Elephants demonstrate susceptibility by
seroconverting when exposed to infection unlike those of rabbits, camels, sheep and goats
(Radostits et al., 2000; Farouk, 1997).

10
In enzootic areas the morbidity rate of susceptible equines varies with the number of insect
vectors present. Mules and donkeys do not suffer from high mortality, but the disease is a
crippling one in these equines because of gross debility (Radostits et al., 2000; Anon,
1998).

After natural infection or vaccination by a homologous strain, immunity is solid but can be
overcome by strong challenge by another strain. The development of immunity is slow and
may require 3 weeks to be appreciable; titers may continue to rise for 6 months after
vaccination. Foals from vaccinated dams appear to derive passive immunity from the
colostrum and are immuned until 5-6 months of age (Hall, 1985; Radostits et al., 2000;
Anon, 1998).

2.1.5. Reservoir host

There were extensive epizootics of AHS in North Africa, the Middle East, and the Indian
subcontinent, the disease has always died out there. Possible explanations include the
absence of a suitable reservoir mammalian host such as the zebra or lack of persistence of a
vector temporarily introduced into these areas (Fenner et al., 1987). In tropical countries
where there is no cold season it is believed that a reservoir host may not be necessary
because the disease is not so strictly seasonal, the virus is being transmitted via the vector
from horse to horse continuously (Della-Porta, 1985; Hall, 1985). Nevertheless, Culicoides
is not considered to be the reservoir which retains the virus between the seasonal
appearances of the disease in horses, as the disease can disappear in areas where Culicoides
abounds. If there is a separate inter-season host reservoir, its density has not yet been
established (Hall, 1985). Although clinically affected equidae are the major source of virus
during an outbreak, the current view is that in enzootic areas there must be a silent, non-
equine reservoir host, which perpetuates the virus between seasons when no insects are
present. Dogs, elephants and zebra have all been proposed as reservoir hosts. In some
countries the disease has been introduced but has died out in the succeeding winter,
presumably because no reservoir hosts were available (Radostits et al., 2000; Anon, 1998).
In enzootic areas of Africa, there is strong evidence that a non-equine host is a reservoir for
survival of the virus between seasonal attacks. (Knowles, 1991; Radostits et al., 2000).

11
There is no confirmed information about a reservoir of infection; but animals recovered
from the disease do not remain carriers of the virus. There may, however, be a continuous
transmission cycle of AHSV in tropical areas of Africa between Culicoide midges and wild
or domestic equine species or other wild reservoir hosts (Seifert, 1996; Anon, 1998).

2.1.6. Environment

The ecology of AHS is not properly understood (Radostits et al., 2000). Culicoides
involved in transmission is shown by the fact that, in countries where cold winters occur no
new outbreaks have been recorded after 10 days of frost because the insect vectors are
killed in the cool conditions. Also, extensive outbreaks are always preceded by heavy rain
(Della-Porta, 1985; Hall, 1985).

Consideration must also be given to the wind in carrying the insect vector to new areas.
Thus, the disease is carried from one country to another even over long distances. English
workers have reported that it has been transmitted from Morocco to Spain and from
Senegal to Cape Verde by this means (Hall, 1985; Hunter, 1994). Because of the nature of
Culicoides, enzootic areas are more likely to be in low-lying, warm, marshy regions (Hall,
1995; Radostits et al., 2000).

AHS infection rates of vector Culicoides and rates of virogenesis within them are
temperature dependent. As temperature increases infection rates also tend to increase.
Virogenesis is faster and transmission can occur sooner. However, midge survival rates
decrease. Conversely, as temperature is reduced the reverse is true for each of these
variables. The likelihood of transmission is therefore a function of the interaction of these
two opposing sets of trends (Mellor and Hamblin, 2004).

2. 2. Experimental production

Experimental and natural transmission of AHS to dogs has been reported through ingestion
of infected raw horse meat. However, there is only limited and unsubstantiated evidence
that dogs can be infected by insect bites (Anon, 1998; OIE, 2004). Artificially, the disease
is readily transmitted by the intravenous injection of very small amounts of blood.
Transmission can also be effected by subcutaneous injection or oral dosing, but larger
amounts of blood are required, particularly with oral dosing (Anon, 1976; Radostits et al.,

12
2000). The disease was transmitted experimentally to a horse by the bites of Aedes aegypti
mosquitoes which had been fed a virus suspension (Galloway, 1985; Anon, 1998).

2. 3. Economic importance

In a fully susceptible horse population, the effect of African horse sickness can be
devastating, because up to 95% mortality can be expected. For instance, the cost of the
Newzealand horse industries of an outbreak of AHS would be expected to exceed
substantially the cost of an outbreak of equine influenza which has been estimated at nearly
173 million USD. The presence of AHS in a country would cause a major disruption to the
export of horses (MAF, 1991).

The serious nature of the disease for equines is compounded by the tremendous problem of
eradication. Vaccination reduces the ravages of horse sickness, but even when practiced on
a wide scale cannot eradicate the disease because the infection is insectborne, and
uncontrolled hosts provide a reservoir of infection (Radostits et al., 2000).

2. 4. Human disease

Very rarely, African horse sickness can be zoonotic. The first evidence of this came when
laboratory workers, exposed to the virus during vaccine manufacture, developed
encephalitis, chorioretinits, and disseminated intravascular coagulation (Anon, 1998;
Murphy et al., 1999). There is no evidence that humans can become infected with any field
strain of viscerotropic AHSV, either through contact with naturally or experimentally
infected animals or by virus manipulation in laboratories. However, certain neurotropic
vaccine strains that may cause encephalitis and retinitis in humans following transnasal
infections have been described (Anon, 1998; OIE, 2004).

13
2. 5. Pathogenesis

The outcome of infection in horses, including the incubation period and severity of disease,
depends largely on the virulence of the virus and susceptibility of the animal. The
incubation period varies between 3-10 days (Coetzer et al., 1994; Seifert, 1996).

The factors determining the course and severity of infection are not fully understood
(Seifert, 1996). After infection, initial multiplication of virus occurs in the regional lymph
nodes and is followed by primary viraemia with subsequent infection of target organs,
namely the lungs and lymphoid tissues throughout the body. Virus multiplication at these
sites gives rise to a secondary viraema of variable duration; in horses it is generally not
higher than 105 TCID50/ml and lasts 4 to 8 days but does not exceed 21 days, whereas in
donkeys and zebras the levels of viraema are lower but they may last as long as 4 weeks. In
zebras, viraemia has been reported to occur in the presence of circulating antibodies. The
virus is closely associated with the erythrocytes in the blood (Coetzer et al., 1994; Hunter,
1994; Seifert, 1996).

In experimentally infected horses, high concentration of virus is found in the spleen, lungs,
caecum, pharynx, choroid plexus and most lymph nodes by the second day after
inoculation. This precedes the onset of fever or detectable viraemia. By the third day after
virus inoculation, it is present in most organs. Even in the “dikkop” or “cardiac” form of
AHS, virus levels in the myocardium is no longer than in the blood, indicating that the
myocardium is not a primary site of virus replication. High concentration of virus in
lymphoid tissues may possibly be responsible for the depletion of lymphocytes which
occurs during the course of the disease. However, transmission electron-microscopic
studies have not revealed the presence of virus in lymphocytes (Anon, 1976; Coetzer et al.,
1994).

Effusions into body cavities and oedematous changes of various tissues (particularly of the
lungs), as well as serosal and visceral haemorrhages, are often evident in fatal cases of
AHS, and indicate endothelial cell damage. Although no significant ultrastructural changes
or evidence of viral replication could be detected in endothelial cells in the lungs in one
study, the presence of virus and ultrastructural changes in, and separation of, endothelial
cells in the lungs were found in another (Pagot, 1992; Coetzer et al., 1994).

14
The development of the various forms of the disease depends on the envelope chemistry of
the individual strain of virus concerned and the chemistry dictates the tissue to which the
serotype will be directed (Radostits et al., 2000; Anon, 1998).

The virus is present in the blood stream from the first day of clinical illness and persists for
about 30 days and even up to 90 days. It can be recovered from defibrinated blood by
intracerebral inoculation into infant mice (Radostits et al., 2000; Anon, 1998).

2. 6. Clinical signs

Small plaque variants of AHSV produce severe clinical reactions, while large plaque
variants produce no reactions or only mild ones. Subsequent observations indicated that the
rigid distinction between different forms of AHS is not fully justifiable and that most cases
of AHS are of the ‘’mixed’’ form. Although more than one serotype may be active during
an outbreak, isolation of more than one serotype from a naturally infected animal has never
been recorded (Coetzer et al., 1994; Anon, 1998).

According to Coetzer et al. (1994) and Radostits et al. (2000) there are four forms of AHS
in horses and these are still useful in categorizing the different effects AHSV may have on
equine animals. The four forms are listed below:
1. the peracute, “pulmonary” or “dunkop” form= “thin head”, i.e. subcutaneous
swelling of the head is absent;
2. the acute or “mixed” form;
3. the subacute, oedematous, “cardiac” or “dikkop” form= “thick head” or swollen head
and
4. the horse sickness fever (abortive) form.

2.6.1. “Dunkop” form

This form of the disease occurs most commonly when AHSV infects fully susceptible
horses, notably foals that have lost their passively acquired maternal immunity; and when
equines are infected with virulent strains of virus. It is also the usual form in dogs (Coetzer
et al., 1994; OIE, 2004).

15
This form of the disease has a short incubation period (3-5 days), is characterized by very
marked severe dyspnoea, and progressive respiratory involvement. An acute febrile
reaction, lasting 1-2 days and reaching a maximum of approximately 40-41OC may be the
only sign. This is followed by various degrees of respiratory distress. The respiratory rate
may increase to 60 or even to 75 breaths/minute. The animal may be observed to stand
with its fore legs spread apart, its head extended and its nostrils fully dilated. Profuse
sweating is common and spasmodic coughing may be observed terminally, with frothy
serofibrinous fluid exuding from the nostrils. Death usually occurs with in a few hours
after the first clinical signs are observed, the animal having literally drowned in its own
serous fluid. Recovery from this form is very rare, occurring in less than 5% of cases
(Radostits et al., 2000; Hunter, 1996; OIE, 2004).

2.6.2. “Mixed” form

This form represents a mixture of the pulmonary and cardiac forms. Although seldom
diagnosed clinically, it is the most common form and is seen at postmortem examination in
most fatal cases of AHS in horses and mules. The incubation period varies from 5-7 days,
and the disease may manifest itself in the following ways (Galloway, 1974; Hunter, 1996;
OIE, 2004):
 Initial pulmonary signs of a relatively mild degree are followed by marked
oedemaotous swellings of the head and neck, with death resulting from heart
failure,
 Oedematous swelling, typical of the subacute form, is followed by sudden onset
of dyspnoea and other clinical signs typical of the peracute pulmonary form.
The mortality rate in the mixed form is greater than 80% and death usually follows within
3-6 days after the onset of febrile reaction (OIE, 2004).

2.6.3. “Dikkop” form

The incubation period of this form varies from about 7-14 days, and the onset of clinical
disease is marked by a febrile reaction (39-41OC) that lasts for 3-6 days. Shortly before the
decline of fever, characteristic oedematous swellings may appear. These initially involve
the temporal or supraorbital fossae and the eyelids, and later extend to the lips, cheeks,
tongue, intermandibular space and laryngeal region. Subcutaneous oedema sometimes

16
extends a variable distance down the neck towards the chest and, in severe cases, may
involve the chest and shoulders, but generally not the lower limbs. Terminally, petechial
haemorrhages may be observed in the conjunctivae and under the ventral surface of the
tongue. The animal finally becomes restless and may show signs of colic before death from
cardiac failure. Difficulty in swallowing due to paralysis of the oesophagus is also seen.
The mortality rate is about 50% and death usually occurs within 4-8 days after the onset of
febrile reaction. In recovering cases, swelling gradually subsides within a period of 3-8
days. This clinical form of AHS is usually associated with infection by virus strains of low
virulence or is encountered in vaccinated animals infected by heterologous virus strains, or
may be a function of biological variation in the infected animal (Coetzer et al., 1994;
Hunter, 1996; OIE, 2004).

2.6.4. Horse sickness fever (abortive form)

Horse sickness fever is the mildest form and is frequently overlooked in natural outbreaks.
The incubation period varies from 5-14 days, and is followed by a febrile reaction (39-
40OC) of the remittent type, with morning remissions and afternoon exacerbations, lasting
for 5-8 days. Apart from the febrile reaction, other clinical signs are rare. The conjunctivae
may be slightly congested, the pulse rate may be increased, and a certain degree of
anorexia and depression may be present. This form of the disease is usually observed in
partially immuned animals or in resistant species, such as the donkey and zebra (Sewell
and Brocklesby, 1990; Hunter, 1996; OIE, 2004).

2.7. Pathology

The pathological features are most prominent in the pulmonary and cardiac forms. The
most striking features of the pulmonary form are severe oedema of the lungs and
hydrothorax. Several liters of pale yellow fluid which may coagulate on exposure to air are
found in the thoracic cavity. Epicardial and endocardial haemorrhages of the endocardium
may be evident. Congestion of the mucosa of the stomach and patchy congestion and
petechiation of the serosa and sometimes of the mucosa of the intestine appear. The liver
may be slightly enlarged and congested; there is usually some degree of ascites (Radostits
et al., 2000; Seifert, 1996).

17
The most characteristic signs of the cardiac form are the distinctly yellowish gelatinous
oedema of the subcutaneous and intramuscular connective tissues of the head and neck,
which in severe cases extends to the back, shoulders and chest. The eyelids, supraorbital
fossae, lips, cheeks, tongue and intermandibular space are commonly involved. The tongue
may be severely swollen, cyanotic and have mucosal petechiae on its ventral surface.
Lesions in the heart are more severe than described, for the pulmonary form and
accompanied by severe hydropericardium. Similarly, the lesions in the gastrointestinal tract
are usually more severe than in the pulmonary form (Seifert, 1996; Anon, 1998).

There are few published reports on the histopathology of AHS. However, exudative
pneumonia, congestion of alveolar capillaries, arterioles and venules as well as
perivasculitis are evident. Degeneration and necrosis of myocytes, oedema of the
myocardium with infiltration of mononuclear cells, plasma cells, siderocytes and
polymorphonuclear leukocytes as well as lysis of necrotic myocytes appear (Coetzer et al.,
1994; Seifert, 1996).

2. 8. Diagnosis

Wherever AHS is endemic the epidemiology, clinical signs and macroscopic lesions are
sufficiently specific to allow its diagnosis. The chronic form is more difficult to diagnose
(OIE, 2004).

2.8.1. Identification of the agent

Several techniques are already available for AHSV identification ranging from the rapid
Enzyme Linked Immunosorbent Assay (ELISA), using either polyclonal antibodies (PAbs)
or monoclonal antibodies (MAbs), to the polymerase chain reaction (PCR) test, including a
new reverse-transcription (RT) PCR for discrimination of the nine AHSV serotypes, or cell
culture and inoculation of newborn mice. If possible more than one test should be
performed to diagnose an outbreak of AHS, especially the index case. The initial test can
be a quick test such as ELISA or PCR, followed by virus isolation in tissue culture. Virus
neutralization (VN) for serotype identification should be performed as early in the outbreak
as possible so that the correct vaccine can be selected. Subsequently, the ELISA may be
very useful in laboratory diagnosis (OIE, 2004).

18
At present, there are no international standards for viruses or diagnostic reagents, and there
is no standard methodology for the determination of AHSV. However, a panel of viruses
has been evaluated, and comparative studies between different ELISAs for AHSV antigen
determination have been carried out in different laboratories. The results have
demonstrated a high level of correlation for antigen detection using the indirect sandwich
ELISAs for antigen studies (Seifert, 1996; OIE, 2004). A very important aspect of the
diagnosis is the selection of samples and their transportation to the laboratory (OIE, 2004).

2.8.1.1. Samples for virus isolation

Unclotted whole blood collected during the early febrile stage of the disease from sick
animals, as well as small pieces (2-4g) of spleen, lung and lymph nodes from animals that
have recently died, are the samples of choice for diagnosis. Samples should be kept at 4 OC
during transportation and storage (OIE, 2004).

2.8.1.2. Cell culture

According to OIE (2004) direct isolation of AHSV on baby hamster kidney (BHK-21),
monkey stable (MS) and African green monkey kidney (Vero) cell lines have been used
successfully. Blood samples collected in heparin can be used undiluted. After 60 minutes
of adsorption, the cell cultures are washed and maintenance medium is added.
Alternatively and more usually, the blood is washed, lysed and diluted 1/10. This
procedure removes unwanted antibody, which could neutralize free virus, and promotes
release of virus associated with the red blood cell membranes. When tissue samples, such
as spleen, lung etc., are used, a 10% tissue suspension is prepared in phosphate buffered
saline (PBS) or cell culture medium, containing antibiotics. A cytopathic effect (CPE) may
appear between 2 and 8 days post infection. Three blind passages should be performed
before considering the samples to be negative.

19
2.8 1.3. Newborn mice inoculation

This method of isolation of AHSV involves the intracerebral inoculation of two families of
1-3 day-old mice. In positive cases, animals develop nervous signs between 3 and 15 days
post-inoculation. The brains from sick animals must be collected, homogenised and re-
inoculated intracerebrally into at least six 1-3 day-old mice. This second passage should
present a shortened incubation period (2-5 days) and 100% infectivity (Coetzer et al.,
1994; OIE, 2004).

2.8.1.4. Polymerase chain reaction (PCR)

A number of PCR-based assays for the specific detection of AHSV RNA have been
developed. Primers correspond to the 5’ end (nucleotides 1-21) and 3’ end (nucleotides
1160-1179) of RNA segment 8 have been developed (Coetzer et al., 1994; OIE 2004).

Extraction of nucleic acids from spleen samples is carried out as follows: 1g of tissue
sample is homogenize in 1ml of denaturing solution (4M guanidium thiocyanate, 25mM
sodium citrate, 0.1M 2-mercaptoethanol, 0.5% sarcosyl). After centrifugation, 1g of yeast
RNA, 0.1ml of 2M sodium acetate pH 4, 1ml of phenol and 0.2ml of chloroform/isoamyl
alcohol mixture (49/1) are added to the supernatant. The suspension is vigorously shaken
and cooled on ice for 15 minutes. After centrifugation, the RNA present in the acqueous
phase is phenol extracted, ethanol precipitated and resuspended in sterile water. The
methods for cDNA synthesis and PCR amplification are performed using, in all cases,
370C as renaturing temperature. The sequences of the PCR primers used are 5’-GTT –
AAA –ATT –CGG –TTA –GGA –TG –3’, which corresponds to the messenger RNA
polarity and 5’-GTA-AGT-GTA-TTC-GGT-ATT-G-3’, which is complementary to the
messenger RNA polarity. The PCR procedure itself involves 40 cycles (94 0C for 1 minute,
550C for 1.5 minutes, 720C for 2.5 minutes and 700C for 7 minutes) and then the PCR
tubes are kept at 40C. Analysis of the PCR products is carried out by electrophoresis in
1.2% (w/v) agarose gels containing ethidium bromide. AHS positive samples will resolve
in an 1179 base-pair band (OIE, 2004).

A new RT-PCR for discrimination of the nine AHSV serotypes has been described. Nine
pairs of primers were designed for each specific serotype. The results obtained show a

20
perfect agreement between the RT-PCR and the VNT. Typing of the nine AHSV serotypes
has also been carried out with probes developed from a set of cloned full-length VP2 genes
and can be an alternative to amplification of genome segment (OIE, 2004).

2.8.1.5. New approaches

Genomic probes can also be developed and applied for in-situ hybridization in tissues.
Immunohistochemical staining methods have also been used successfully to determine the
localization of AHS antigen within various tissues. The advantages of these new
approaches are that, they have the potential to be rapid, sensitive and versatile, and they
may supplement or replace some of the older conventional methods. Furthermore, they can
be applied to specimens from clinical cases that do not contain live virus (Coetzer et al.,
1994).

2.8.2. Serological test

There is an international reference sera developed to standardize the ELISA, which is an

OIE prescribed test. In addition, a panel of reference antisera has been evaluated and
comparative studies between different ELISAs using MAbs and PAbs, and involving
several laboratories have been carried out. The results have demonstrated a high level of
correlation using indirect or competitive ELISAs for antibody detection. The competitive
ELISA (C-ELISA) can also be used for testing wild life, as species-specific anti-globulin is
not required with this method (OIE, 2004). VNT is the method of choice for serotyping
(OIE, 2004; Mellor and Hamblin, 2004).

The indirect sandwich ELISA is an extremely useful method for the rapid identification of
AHSV antigen in solid tissues taken from animals that have died following an acute
infection (Mellor and Hamblin, 2004; OIE, 2004). Sandwich ELISA’s that detect viral
antigen in mammalian and insect tissue homogenates as well as in cell culture supernatant
fluid have been reported to be used in the diagnosis of AHS (Coetzer et al., 1994).

21
2. 9. Differential diagnosis

The clinical signs and lesions reported for AHS can be confused with those caused by
closely related Orbivirus, equine encephalosis virus (EEV). Fortunately, rapid, sensitive
and specific ELISA’s are available to enable the detection of the antigen and antibody of
both AHSV and EEV. The haemorrhages and oedema reported in cases of purpura
haemorrhagica and equine viral artritis may be similar to those seen in the pulmonary form
of AHS, although with AHS the oedema tends to be less extensive and the haemorrhages
are less numerous and widespread. The early stages of babesiosis (B. equi and B.caballi)
can be confused with AHS, particularly when the parasites are difficult to demonstrate in
blood smears (Mellor and Hamblin, 2004).

2. 10. Treatment

Apart from supportive treatment, there is no specific therapy for AHS. Affected animals
should be carefully nursed, well fed and given rest as the slightest exertion may result in
death. During the recovery process they should be rested for at least 4 weeks before being
returned to light work. As babesiosis may be a complication of AHS, blood smears as well
as the body temperature (to detect a secondary febrile reaction) should be taken regularly
and, if found positive, animals should be appropriately treated. Careful massaging of the
oesophagus in those cases in which it is paralyzed may result in a gradual improvement of
the condition (Hall, 1985; Coetzer et al., 1994; Hunter, 1994).

2. 11. Prevention and control

Polyvalent and, to a limited extent, monovalent vaccines are used for immunization. A
polyvalent, formalin-inactivated spleen tissue (viscerotropic) vaccine gives immunity for 1
year. A second vaccine is a mouse-brain attenuated live virus (neurotropic) vaccine that
has been passaged in mice by intracerebral inoculation, which is also polyvalent. Annual
vaccination is advocated to obtain true polyvalent immunity. It has been observed that the
use of polyvalent vaccine rarely results in polyvalent immunity after the 3 rd or 4th annual
vaccination. Prophylactic immunization against AHS is a very efficient method of
preventing serious losses (Coetzer et al., 1994).

22
2.11.1. Endemic Areas

2.11.1.1. Vaccination

Annual vaccination in late winter or early summer (September to November) which is

sometime before the peak of AHS season with a live polyvalent attenuated vaccine is
mandatory and allows immunized animals to respond adequately to the vaccine before
possibly being challenged by natural exposure. Vaccinated animals should not be exercised
for three weeks following vaccination. Foals should be vaccinated at six months of age and
there after annually (Galloway, 1974; Della-Porta, 1985; Pagot, 1992; Hunter, 1994;
Coetzer et al., 1994; Seifert, 1996).

Horses that have received three or more courses of immunization are usually well protected
against the disease. Generally immunization has no, or only limited side effects. A slight
temperature response may ensue between 5 and 13 days after inoculation as a result of low
level virus replication in immunized animals. Cases of fatal encephalitis, characterized by
blindness and neurological disorders, in fully susceptible horses and donkeys 6 to 8 weeks
after initial vaccination with neurotropic mouse strains of virus has occasionally been
reported (Coetzer et al., 1994; Seifert, 1996).

Cross-immunity between serotypes may be enhanced by repeated inoculation of the same

virus. There is evidence that animals recovered from infection with a field virus require a
broader cross-immunity to the other serotypes than that which is obtained from
immunization. Animals should be vaccinated twice in their first and second years of life
and thereafter only once annually to avoid unresponsiveness (Coetzer et al., 1994).
Through application of recombinant DNA technology, subunit vaccines are expected to
become available for AHSV within the foreseable future (Seifert, 1996).

2.11.1.2. Vector Control

Infection of susceptible horses can be prevented to a large degree by stabling them some
hours before sunset and letting them out a few hours after sunrise as Culicoides species are
nocturnal and are not inclined to enter buildings (Coetzer et al., 1994; Seifert, 1996).

23
2.11.2. Non-endemic Areas

Following an outbreak of AHS in a country that has been free of it, much more stringent
control measures must be taken which involve quarantine, slaughtering of viraemic
animals, vaccination, stabling and controlling Culicoides (Coetzer et al, 1994).

24
3. MATERIALS AND METHODS

3.1. Description and physiography of the study areas

The study was conducted in selected sites of Addis Ababa city administration and its
surroundings as well as Amhara and Oromia regional states of Ethiopia. In total 4 zonal
administrations, 13 districts and 19 peasant associations were included in the study.
Specific study sites encompassed Addis Ketema from Addis Ababa region; Dessie Zuria,
Kalu and Kombolcha of South wollo; Debrebrihan and Basona Werana of North shoa from
Amhara region; Adama, Dugda Bora, Akaki, Ada’a Liben, Berehna Aleltu and Boset of
East shoa from Oromia region. These study areas are classified into three agro-ecological
zones based on temperature and length of plant growing period (LGP) (EARO, 1998).
Keith Powell’s (2005) adopted classification that categorizes areas with elevations of 1500
meters above sea level (m.a.s.l.) or less was endorsed as lowland, between 1500-2500
m.a.s.l. as midland and over 2500 m.a.s.l. as highland.

The altitudinal range of the study areas is from 1413 (Boset) to 2812 (Addis Ababa) meters
above sea level. The rainfall in the study areas is bimodal in disribution and falls in the
range of 900 to 1000 mm per annum. There are long and short rainy seasons extending
from June to September and February to March, respectively. The minimum, maximum
and average air temperature of the study areas are 160C, 27.50C and 21.750C respectively.
The mean relative humidity of the study areas is in the range of 35-68 %. Study areas such
as Ada’a Liben, Adama, Boset and Dugda Borra are near to watering bodies such as
Awash river, Koka dam, irrigation canals, Zeway lake and different lakes of Ada’a. Ada’a
is near but others are within the belt of Rift Valley (see figure 1).

25
3.2. Geo-reference (GPS database)

Table 2. Measured Geo-reference database of the study areas.

Elevation
No. Study sites (m.a.s.l.) Latitude (N) Longitude (E)
1 D/B kebele 8 2678 090 01’ 232’’ 0380 48’ 177’’
2 D/B kebele 4 2641 090 01’ 232’’ 0380 48’ 177’’
3 Tulamba 2694 090 01’ 232’’ 0380 48’ 177’’
4 Chollie 2501 090 01’ 232’’ 0380 48’ 177’’
5 S.chollie 2495 090 01’ 232’’ 0380 48’ 177’’
6 Legebollo 2465 090 01’ 232’’ 0380 48 ’177’’
7 Addis Ababa 2812 090 01’ 232’’ 0380 48’ 177’’
8 Akaki 2025 090 01’ 232’’ 0380 48’ 177’’
9 Sidamawash 2064 090 01’ 232’’ 0380 48’ 177’’
10 Gogiecha 1956 090 01’ 232’’ 0380 48’ 177’’
11 Ada’a 1856 090 01’ 232’’ 0380 48’ 177’’
12 Gerado 2321 110 09’ 810’’ 0390 61’ 740’’
13 Kombolcha 1818 110 07’ 090’’ 0390 73’ 851’’
14 Kalu 1837 110 08’ 475’’ 0390 73’ 741’’
15 Adama 1566 090 01’ 232’’ 0380 48’ 177’’
16 Boset 1413 090 01’ 232’’ 0380 48’ 177’’
17 Godino 1922 080 45’ 681’’ 0390 00’ 143’’
18 Guaworko 2025 080 51’ 245’’ 0390 01’ 329’’
19 Gende gorba 1963 080 51’ 856’’ 0390 00’ 574’’
20 Meki 1600 0801 2’ 887’’ 0380 52’ 488’’
21 Alemtena 1672 080 09’ 472’’ 0380 05’ 248’’

D/B = Debrebrihan

26
 South Wollo, Amhara East Shoa, Oromia
North Shoa, Amhara and Oromia

Addis Ababa

Figure 1. Map of the study areas

3.3. Target Population

The target population of the study areas were 16,480 donkeys, 4,791 horses and 58 mules.
Equidae above six months of age, both sexes and with no previous history of vaccination
against African horse sickness were sampled. The age category of sampled equidae is
shown by table 3.

27
Table 3. Number of sampled equidae in different age category.

Category Age range Sampled equidae

(years) Donkey Horse Mule Total
1 1-7 364 112 25 501
2 8-14 455 239 32 726
3 15-20 5 32 1 38
Total - 824 383 58 1265

3.4. Sampling strategy

Purposive sampling method was employed up to district level. Stratification of equines

into donkeys, horses and mules was made at peasant association level. Equines were taken
as the primary units of the stratum and the stratum was determined to be one (equines).
Discussion was made with equine owners of different peasant associations to know the
equine resource base of their respective areas. The number of equines to be sampled from
different study areas was determined by the method of proportional allocation based on
5%, 8% and 100% sampling fractions for donkeys, horses and mules, respectively.
Random sampling was made by employing a lottery system to select the primary units.
Since owners were not willing their mules to be sampled, their sampling was determined to
be 100% (see table 4).

3.5. Sample size determination

The sample size was determined according to Thrusfield (1995) simple random sampling
method for an infinite population with 95% confidence level, 5% desired absolute
precision and 50% expected prevalence, since there was no previous information on the
prevalence of AHSV antibodies in the study areas. The sample size was calculated by
making use of the following formula after Thrusfield (1995).

n= 1.962 pexp (1-pexp )

___________________
d2

Where,
n = required sample size
pexp = expected prevalence
d = desired absolute precision
1.96 = z-value for the 95% confidence level

28
In total 1265 serum sample comprising 824 donkeys, 383 horses and 58 mules were
collected.

3.6. Study design

A cross-sectional observational epidemiological study was conducted using competitive

Enzyme Linked Immunosorbent Assay (C-ELISA) configuration. Data on the potential
risk factors associated with the occurrence of African horse sickness were collected during
sampling through recording and questionnaire administration. Active disease search was
carried out in those areas with a history of AHS outbreak and manifestation of typical
clinical signs by sick equidae. Field sera and data collection was undertaken from
September 2005 to mid of April 2006.

3.7. Study Methodologies

3.7.1. Blood Collection and Assay Methodology

3.7.1.1. Seroprevalence survey

Equine serum collected from different study areas are shown by table 4.

Table 4. Serum sample collected from equines of different study areas.

Number of equines Number sampled

Study areas Donkeys Horses Mules Donkeys Horses Mules
Gerado 1900 1000 43 95 80 43
Kombolcha - 500 - - 40 -
Kalu - 700 - - 56 -
Debrebrihan 8 780 613 - 39 49 -
Debrebrihan 4 800 675 - 40 54 -
Tulamba 840 - - 42 - -
Fagi 800 - - 40 - -
Senkolliechollie 1020 163 - 51 13 -
Legebollo 820 138 - 41 11 -
Sidamawash 1200 - 1 60 - 1
Gogiecha 1120 - - 56 - -
Godino 2280 - 1 114 - 1
Ada’a town 580 438 13 29 35 13
Abeno one 140 - - 7 - -
Alemtena 740 13 - 37 1 -
Soridolliesa 900 - - 45 - -
Adama town - 413 - - 33 -
Treesengota 1560 138 - 78 11 -
Addis ketema 1000 - - 50 - -
Total 16,480 4,791 58 824 383 58

29
Whole blood of 10ml was collected by vein puncture using sterile venoject needles and
plain vacutainer tubes including needle holder under aseptic precautions. Each sample was
labelled with identification number. The blood was allowed to clot over night at room
temperature. The recovered serum was decanted into another vacutainer tube and labelled
with similar identity. The samples were kept at –200C until evaluated with C-ELISA. Area
of sampling, age, sex, type of equidae (donkey, horse, mule), date of collection, agro-
ecology, management practices, and vacutainer ID number were recorded at the time of
sampling. The test was done in 2-4 months time after serum collection. The kit was AHSV
serotype 9 specific. Sensitivity and specificity of the test kit was not given in the bench
protocol. Standardization and validation of the kit was conducted before the test sera are
subjected to C-ELISA test procedures. Then similar results were obtained with the
interpretation given in the test protocol except the conjugate. The information given in the
protocol to prepare chromogen/substrate complex was also insufficient. Through a request
forwarded to get clarification on minor issues to the manufacturer, problems of some
parameters and documentations were pointed out. Finally the overall test procedure was
evaluated and confirmed to run safely. The C-ELISA test was conducted at NVI
immunology laboratory, Ethiopia according to the Standard Operating Procedure (SOP)
developed by the joint venture of Biological Diagnostic Supplied Limited (BDSL)
Company and the Institute of Animal Health (IAH) Pirbright, Scotland, United Kingdom.
The SOP procedure is described as follows:

AHSV antigen was thawed and mixed by vortexing to ensure homogeneity. It was diluted
at 1:150 in PBS. 50μl of the reconstituted antigen was coated to each microwell of the
ELISA plates (Nunc Maxisorb) and sealed with plate sealer so as to be incubated for 1hour
at 370C on an orbital shaker. The stock antigen was returned to the freezer (-200C).

Plates were washed 4 times by flooding and emptying the wells with PBS and blot dried on
an absorbent towel. Blocking buffer of 40μl was added to all the test wells and to each of
the negative control wells. 50μl of blocking buffer was added to each of the guinea-pig
control wells. Moreover, 100μl of blocking buffer was added to the conjugate control
wells. Finally 50μl of blocking buffer was added for the positive control of wells G up to A
and 80μl in well H of the first column. Test samples of 10μl were added (in duplicate) to

30
their respective wells and 10μl of the negative control sera was added to the negative
control wells after it was reconstituted with 1.0ml bidistilled water.

After it was reconstituted with 1.0ml bidistilled water, positive control of 20μl was added
to well H in the first column and mixed by careful pipetting up and down a number of
times. A 1:5 dilution of 100μl of the positive control was obtained in well H. A two-fold
serial dilution was carried out up the remaining wells of this column by taking 50μl from
well H and dispensing into well G, to give a 1:10 dilution. After mixing and dispensing,
50μl was taken from well G and dispensed into well F, to give a 1:20 dilution. This
procedure was continued until a 1: 640 dilution in row A was obtained. The last 50μl of the
positive control from well A was discarded so that all wells in column one were found to
have 50μl in total.

The immuned guinea-pig antiserum (monoclonal antibody) was thawed and gently mixed
to ensure homogeneity. It was reconstituted with 1.0ml bidistilled water. The guinea-pig
antiserum was diluted to 1:100 in blocking buffer and 50μl of this was added to each well
except the duplicate conjugate control wells. The ELISA plates were covered with plate
sealer and incubated for 1hour at 370C on an orbital shaker. The stock guinea-pig
antiserum was returned to the freezer (-200C).

After washing 4 times the rabbit anti-guinea-pig conjugate was diluted to 1:1000 in
blocking buffer and 50μl of this was added to all wells. The plates were sealed and
incubated at 370C for 1hour in an orbital shaker. The other microwells were filled with test
serum samples, which were duplicated as indicated by figure 2.

1 2 3 4 5 6 7 8 9 10 11 12
A -Ve cont. 8 8
B -Ve cont. 7 7
C Conj.Cont. 6 6
D Conj. Cont. 5 5 Etc Etc
E Mab. Cont. 4 4 Etc Etc
F Mab. Cont. 3 3 11 11
G Mab. Cont. 2 2 10 10
H +Ve cont. Mab. Cont. 1 1 9 9

Figure 2. Plate layout for spot test

31
One plate can accommodate a maximum of 40 samples.

After the plates were washed 4 times, the chromogen (OPD = Orthophenylene diamine
Dihydrochloride) solution was prepared at a rate of 0.4mg/ml in sterile bidistilled water.
Then a substrate (hydrogen peroxide) was added to the OPD solution. 50μl of the OPD
solution was added to the wells of each micro plate, sealed and left in dark place for
10minutes at an ambient temperature. The reaction was stopped by the addition of 50μl of
1M H2SO4 to all wells including the wells acting as the blank. Finally the reaction was read
spectrophotometrically at 492 nm filters (wavelength) and interpreted according to the
interpretation given in the SOP (see annex 1).

3.7.1.2. Virus isolation and identification

Whole blood was collected (and kept at -200C until processed) in heparinized vacutainer
tubes by venoject sterile needles, with needle holder under aseptic precautions from the
jugular vein of horses and a mule manifesting the cardiac form of African horse sickness
characterized by supraorbital oedema and febrile reaction (400C). An equal volume of
blood and bidistilled water were mixed together. After three times washing of the RBCs
with bidistilled water centrifugation was carried out at 1500 rpm for 15 minutes so that the
RBCs are haemolyzed to let the virus free. The supernatant was inoculated onto vero cells
originated from African green monkey kidney (Pan African Veterinary Vaccine
Center/FAO origin) at the National Veterinary Institute in Debre Zeit, Ethiopia. Vero cells
were grown in GMEM at an initial concentration of 380,000 cells/ml, which were
supplemented with 10% calf serum, antibiotics (Penicillin and Streptomycin) and
mycostatin. The cells were prepared 2-3 days prior to use. After the cells reached 100%
confluency 0.2ml of the inoculum was seeded onto early confluent monolayers of vero
cells in tissue culture flasks. The cultures were incubated at 37oC for 30 minutes. After 30
minutes of adsorption time, the inoculum was washed with PBS and refilled with growth
medium containing antibiotics (Penicillin and Streptomycin). After 4 days of post
inoculation maintenance medium containing 5% calf serum is added and incubated at 37oC
for 7 days. The development of cytopathic effect was daily followed through an inverted
microscope.

32
A cytopathic effect (CPE) was expected to appear between 2 and 8 days post inoculation.
Three blind passages were performed to recover AHSV. AHSV serotype specific serum
titration by 10 fold serial dilution was supposed to be carried out to determine TCID50/ml
of solution. The 50% end-point titre of the serum was supposed to be calculated by the
Spearman-Karber method and expressed as the negative log10. Then virus neutralization
test with the objective of serotyping by serotype specific serum was planned to be
conducted by taking 25μl from either of them in order to seed on confluent monolayer of
vero cells. However, serotype specific serum was not available at NVI and the procedure
was discontinued. Then the sample was recommended to be tested by I-ELISA protocol by
making use of the monoclonal antibody developed for AHSV serotype 9 of the C-ELISA
configuration. This procedure was carried out to rule out the possible existence of African
horse sickness serotype 9.

3.7.1.2.1. Indirect-ELISA

The control antigen was thawed and diluted in PBS at a rate of 1:150. Sixteen microwells
of polystyrene microplate were precoated with 50μl of the reconstituted antigen as a
control. The supernatant of the suspected nine samples was coated in nine microwells at
different rates of 1:1, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64, and 1:128 dilutions in the rows of A3-
A11 and in columns of A-H. Totally 8*9=72 microwells were filled with test samples.
Then 50μl from each dilution was added to the nine different rows of microwells until 72
microwells were filled. Both control antigen and the samples were incubated for 1hour in
an orbital shaker. After washing four times with PBS and drying, 50μl of monoclonal
antibody diluted at a rate of 1:100 with blocking buffer was added to the control and test
microwells and incubated for 1hour in an orbital shaker. After washing four times and
drying, 50μl of the conjugate diluted at a rate of 1:1000 with blocking buffer was added to
the controls as well as to the test microwells and incubated for 1hour in an orbital shaker.
Chromogen/substrate complex was prepared, of which 50μl was added to all of the
controls and test microwells. After 10 minutes 50μl of 1M H2S04 was added to each of the
controls and test microwells in order to stop the reaction and facilitate the reading. Then
reading was made spectrophotometrically at 492nm filter (wavelength). The test is said to
be positive if there is colour development and negative if there is no colour development.

33
3.7.2. Questionnaire survey

A structured questionnaire format was prepared to interview individual owners of donkeys,

horses and mules. The format was used to assess the potential risk factors such as
management practices (stabling in enclosed, semi-opened and fenced but with no cover of
roof type of housing during the night), presence of equine biting insects and availability of
water bodies near the study areas (see annex 4) and the knowledge base of equine owners
about AHS. The format was initially being pretested so as to incorporate relevant points in
the final questionnaire format. The interviewees were initially being pretested for their
knowledge of AHS and attention was paid not to miss key informants. A total of 57 equine
owners, three from each of the 19 study areas were interviewed and the result was
summarized.

3.8. Data Management

Data recorded during sampling, laboratory findings and questionnaire survey were entered
and stored in separate MS-access database and MS-Excel spread sheet. The data were
thoroughly screened for errors and proper coding before subjected to statistical analysis.

3.9. Data Analysis

The data were imported from the Microsoft Excel spread sheet and analyzed using
Intercooled Stata 7.0 software to establish association (χ2 test), to measure the strength of
associations between serological test results and the identified potential risk factors
(logistic regression). Win Episcope 2.0 software was used to calculate confidence intervals
for seroprevalence of AHSV antibodies. Descriptive statistics was used to calculate
seroprevalence of AHSV antibodies.

34
4. RESULTS

4.1. Seroprevalence survey

The C-ELISA result of AHSV antibodies in equidae of different study areas is shown by
Table 5.

Table 5. C-ELISA result of AHSV in different study areas.

C-ELISA result
Study areas Total Positive Seroprevalence 95% CI
tested (%)
Abeno One 7 6 85.7 63.3-116.0
Ada’a 54 16 29.6 19.6-44.7
Adama town 42 10 23.8 13.9-40.9
Addis Ketema 50 7 14 7.0-27.8
Senkolliechollie 64 11 17.2 12.0-34.5
Debrebrihan 8 and 4 264 20 7.6 5.0-11.5
Sidamawash 61 24 39.4 28.8-53.7
Gerado 218 0 0 0
Godino 55 25 45.5 34.0-60.7
Guaworko 74 32 43.2 33.3-56.1
Kalu 56 17 30.4 20.4-45.1
Legebollo 52 21 40.4 29.0-56.2
Kombolcha 40 0 0 0
Kombollie 38 19 50 36.4-68.7
Gogiecha 56 25 44.6 33.3-59.8
Alemtena 45 22 48.9 36.3-65.9
Boset 89 36 40.4 31.4-52.0
Total 1265 291 23 20.8-25.4

(Pearson χ2 (16) = 248.2619; P = 0.000)

The variation in seroprevalence of AHSV in different areas was found to be significant

(P<0.05).

Of the total 1265 sampled equidae a mean seropositivity of 23% was obtained.
Seroprevalence increases as one goes from highland and midland to lowland areas.
Seroprevalence was found to be high areas near permanent water bodies such as lakes,
rivers, irrigation canals and water reservoirs. The seroprevalence of African horse sickness
virus in different types of equidae is presented in Table 6 and Figure 3.

35
Table 6. C-ELISA result of AHSV in different types of equidae.

Types of equidae C-ELISA result Seroprevalence (%)

Total tested Positive 95% CI
Donkeys 824 245 29.7 26.8 - 33.0
Horses 383 40 10.4 7.8 - 14.0
Mules 58 6 10.3 4.8 - 22.1
Total 1265 291 23 20.8 - 25.4

( Pearson χ2 (2) = 60.4253; P = 0.000).

In this study a significant variation of seropositivity was observed among different types of
equidae (P<0.05). Figure 3 shows the highest seropositivity result in donkeys, horses and
mules showed almost similar seropositivity.

29.70%

30.00%

25.00%
Seroprevalence

20.00%

15.00% 10.40% 10.30%

10.00%

5.00%

0.00%
Donkey Horse Mule
Types of equidae

Figure 3. C-ELISA result of AHSV in different types of equidae.

36
Table 7. C-ELISA result of AHSV in different age categories of equidae.

C-ELISA Result Seroprevalence

Age category (years) Total tested Positive (%) 95% CI
1 (1 - 7) 501 125 25 21.4 - 29.0
2 (8 -14) 726 161 22.2 19.4 - 25.4
3 (15 - 20) 38 5 13.2 5.8 - 29.8
Total 1265 291 23 20.8 - 25.4

(Pearson χ2 (2) = 3.4320; P= 0.180)

There was no significant variation in seroprevalence among the different age groups of
equidae affected by African horse sickness virus (P>0.05).

25
25 23
22

20
Seroprevalence( % )

15 13

10

0
1(1-7) 2(8-14) 3(15-20) Total
Age category

Figure 4. C-ELISA result of AHSV in different age categories of equidae.

37
Table 8. C-ELISA result of AHSV in the two sexes of equidae.

C-ELISA Result Seroprevalence

Sex Total tested Positive (%) 95% CI
Female 447 129 28.9 24.9 - 33.4
Male 818 162 19.8 17.3 - 22.7
Total 974 291 29.9 20.8 - 25.4

(Pearsons χ2 (1) = 13.3796; P = 0.000)

The seroprevalence in the two sexes of equidae was found to be statistically significant
(P<0.05).

Table 9.C-ELISA result of AHSV in different types of equidae and sexes.

Types of Sex Seroprevalence (%)

equidae Result Female Male Total Female Male
Donkey Negative 282 297 579 69 71
Positive 124 121 245 31 29
Total 406 418 824 - -
Horse Negative 14 329 343 74 90
Positive 5 35 40 26 10
Total 19 364 383 - -
Mule Negative 22 30 52 100 83
Positive - 6 6 - 17
Total 22 36 58 - -

38
 Female
Total 29%
30%

Female
Male
Total

Male
20%

Figure 5. C-ELISA result of AHSV in the two sexes of equidae.

Table 10. C-ELISA result of AHSV in different agro-ecological zones of the study areas.

C-ELISA Result Seroprevalence

Agro-ecology Total tested Positive (%) 95% CI
Highland 378 38 10 7.4 - 13.6
Lowland 212 92 43 37.2 - 50.6
Midland 675 161 24 20.8 - 27.3
Total 1265 291 23 20.8 - 25.4

(Pearson χ2 (2) = 85.8429; P=0.000)

There is significant variation in seroprevalence of AHSV antibodies in different agro-

ecological zones of the study areas (P<0.05). The highest seropositivity was obtained in the
lowland followed by midland and highland.

39
Table 11. C-ELISA result of AHSV in different types of equidae in different agro-
ecological zones.

Agro-ecology
Types of equidae Result Lowland Midland Highland Total
Donkey Negative 86 265 228 579
Positive 81 130 34 245
Horse Negative 34 197 112 343
Positive 11 25 4 40
Mule Negative - 52 - 52
Positive - 6 - 6
Total 212 675 378 1265

The seroprevalence of AHSV in donkeys is 13% in the highlands, 33% in the midlands and
49% in the lowlands. Seropositivity values of 3% in the highlands, 11% in the midlands
and 24% in the lowlands were obtained in horses. Since mules were sampled from only
midlands a seroprevalence of 10.3% is obtained.

50
45
40
Seroprevalence(%)

35
30
25
20 43
15
24 23
10
5 10
0
High land Low land Mid land Total
Agro-ecology

Figure 6. C-ELISA result of AHSV in different agro-ecological zones of the study areas.

40
Table 12. Multivariate logistic regression estimates for assumed risk factors of AHSV in
different types of equidae.

C-ELISA Result Odds Ratio Std. Err. P 95%CI

Types of equidae 2.1 0. 22 0. 000 1.73 - 2.60
Agro-ecology 1.5 0. 13 0. 000 1.27 - 1.77
Age 1.0 0. 02 0. 115 0.99 - 1.06
Sex 0.9 0. 15 0. 937 0.74 - 1.32

The risk of acquiring African horse sickness is more than two fold with respect to the types
of equidae affected. Agro-ecology contributes nearly two fold for the occurrence of
African horse sickness. There is association among C-ELISA result, type of equidae and
agro-ecology. Age is not part of the interaction. Sex is less likely to be part of the
interaction.

4.2. Virus Isolation and Identification

Virus isolation and identification was carried out on blood collected from clinical cases of
AHS. Three blind passages were undertaken on vero cell lines. Then serotyping could have
been carried out. However, due to lack of specific serum in order to run the test the
procedure was discontinued. So the samples were tested with indirect ELISA by making
use of the monoclonal antibodies developed for C-ELISA kit of AHSV serotype 9. The
result was negative.

4.3. Questionnaire Survey

All of the fifty seven owners interviewed complied to the minimum requirement of equine
possession i.e. owning at least either a donkey, horse or mule. 26% (15/57) of the
respondents replied of having encountered the disease once or more in their life time.
While 74% (42/57) of the respondents proved to know nothing regarding the disease. Their
experience of other equine health problems such as lymphangitis (Bichie), rabies, anuria,
impaction, gastrointestinal parasitism, cough (furro), lameness, colic and car accidents was
considerable.

Respondents reported horses to be more affected, followed by mules. The interviewees

were not able to describe the age groups of equines affected by the disease. However, they
said that both sexes are equally affected. In addition owners with the indigenous

41
knowledge of the disease said that the disease is quite common in the lowlands and
midlands, and that it is seen occasionally in the highlands. None of the respondents were
able to explain the mode of AHS transmission, except few, who said that it is through
contact with equines with wounds and bites by Tabanus.

All of the owners replied that, there are equine biting insects in their localities called by
local names such as “ kerchassa, ewir zinb’’ implying Tabanus. These flies are abundant in
the months of September to November as well as from mid April to June. Respondents
were however, unaware of Culicoides and their role as vectors.

As to the maintenance and shelter of equines, owners replied that stabling their equines to
be a common practice during the night. The type of stables however, vary from place to
place. Equine owners of South Wollo stable their equines in houses made of intact walls
and complete roof cover. On the other hand equines of North Shoa are stabled in houses
without walls but with well-constructed roof. Equines of Akaki, Ada’a, Adama and
Dugdaborra areas are stabled in round fenced compound without roof, traditionally called
“Beret”. While some equines in Adama area are stabled in a fence made of wood and
plastic covered roof.

Environment geared enquiry led to the establishment of the following scenario: The study
areas in South Wollo zone are near to rivers such as Borkena in Kombolcha. In North
Shoa zone the existence of rivers such as Beriesa was reported. Equine owners from
Senkolliechollie and Legebollo responded that, their localities are a bit far from the rivers.
However, daily watering of equines in rivers is a routine activity. Owners of Akaki area
reported the existence of marshy areas near their localities. Similarly interviewees from
Ada’a replied that, their localites are surrounded by small lakes and rivers. Equine owners
of the Rift Valley areas described that, there are irrigation canals, marshy areas, water
reservoirs, water wells, large lakes and rivers near their localities.

In view of the importance of animal movement in the epidemiology of AHS, attempt was
made to probe into the degree and pattern of equine migration in the study area. In the
response obtained the informants said generally the movement of equines is from
highlands and midlands to the lowlands. Cart horses are being purchased from the
midlands and highlands.

42
5. DISCUSSION

5.1. Seroprevalence Survey

This study has once again confirmed the existence of agro-ecology based variation in the
occurrence of AHS. The difference in seroprevalence of the various study areas is
statistically significant (P<0.05). Seroprevalence rates of 0 at Gerardo, 0 at Kombolcha,
30.4% at Kalu, 7.6% in Debrebrihan, 17.2% at Senkolliechollie and 40.4% at Legebollo
were obtained in this study. The altitude of these study areas ranges from 1818
(Kombolcha) to 2694 (Debrebrihan) meters above sea level with mean annual rainfall
ranging between 900-1000 mm with bimodal pattern; a long rainy season extends from
June to September and a short rainy season from February to May. The average monthly
minimum air temperature is in between 2.40C in November to 8.50C in August. The
average monthly maximum air temperature ranges from 180C to 240C in June. The mean
relative humidity of these study areas is 68%. Frost usually occurs in the months of
October to January. The length of growing period (LGP) of plants in these study areas
ranges from 121-180 days.

Seroprevalence rates of 14% in Addis Ababa, 39.4% at Sidamawash, 44.6% at Gogiecha,

45.5% at Godino and 29.6% in Ada’a were the findings of this study. The altitudinal range
of these areas is between 1820 (Ada’a) to 2812 (Addis Ababa) meters above sea level. The
mean annual rainfall of these study areas is also in the range of 900-1000 mm with similar
bimodal pattern of the study areas mentioned previously. The average monthly air
temperature is in between 18-240C. The mean relative humidity of these study areas is
similar with the one mentioned above. The length of plant growing period also ranges from
121-180 days.

Seroprevalence rates of 23.8% in Adama, 40.4% at Boset, 48.9% at Alemtena, 85.7% at

Abeno one and 50% at Kombollie were the findings of this study in these study areas. The
altitude of these study areas ranges from 1413 (Boset) to 1672 (Alemtena) meters above
sea level. The mean annual rainfall of these areas is in the range of 900-1000 mm with
bimodal distribution pattern mentioned above. The minimum and maximum temperature
ranges from 30-320C. The daily temperature is said to be 100C. The relative humidity of

43
these areas is between 35-40%. The length of plant growing period ranges from 61-120
days.

According to the seroprevalence study of AHSV conducted by Keith (2005) in different

agro-ecological zones of Amhara and Tigray regions seropositivity values of 48% in
donkeys and 70.6% in mules were obtained. The study further described that 44% donkeys
and 71.7% mules of Amhara region as well as 50% donkeys and 62.5% mules of Tigray
region were found to be seropositive of AHSV. In Keith’s study the main areas of
sampling were lowlands and midlands (endemic areas) as well as areas near to permanent
water bodies such as lake tana and the likes.

The seroprevalence findings in donkeys in this study is in agreement with that of Keith
(2005) except minor variations. The reasons for variations are the sample size, study
design and agro-ecology. The sample size of mules in the present study is not
representative to compare with the findings of others.

The findings in seroprevalence of AHSV in this study in donkeys of different study areas
were 86% at Abeno one, 38% in Ada’a, 14% in Addis Ababa, 44% in Boset, 16% at
Senkolliechollie, 12% in Debrebirhan, 40% at Sidamawash, 46% at Godino, 48% at
Guaworko, 39% at Legebollo, 51% at Kombollie, 45% at Gogiecha and 49% at Alemtena.
Moreover, the seroprevalence obtained in this study in horses of different study areas were
9% in Ada’a, 27% in Adama, 18% at Boset, 23% at Senkolliechollie, 1% in Debrebrihan,
45% at Legebollo and 18% at Kalu. 46% of the mules sampled from Ada’a were found to
be seropositive in this study.

In this study 29.7% of donkeys, 10.4% of horses and 10.3% of mules were found to be
seropositive. Seropositivty increases as one goes from highland and midland to lowland
areas. Hence higher seropositivity was indicated in the lowland followed by midland and
highland areas. Study areas in the lowlands are near permanent water bodies such as lakes,
rivers, irrigation canals, watering points and water reservoirs that support insect breeding
and multiplication. The annual report of NVI (1974), Hall (1995), and Radostits and others
(2000) also suggested that endemic areas are more likely to be in low-lying, warm and
marshy regions that creat favourable environment for multiplication of Culicoides and
mechanical vectors.

44
These days and in every day life the movement pattern of equidae, other livestock and
people is from the highlands and midlands to the lowlands, because of natural resource
scarcities. Some of the highland areas are known to be free from AHS. So in the naïve
population that moves from the highlands, there is maximum mortality at the time of first
exposure to the disease. Horses are not normally kept in the lowlands due to failure of
physiological adaptation mechanisms; only donkeys and mules that are adapted to semi-
arid conditions are let to the lowlands for transportation and packing purposes.

The seroprevalence of African horse sickness virus antibodies in different types of equidae
was determined (P<0.05). In this study seroprevalence was higher in donkeys (29.7%) and
almost similar in horses (10.4%) and mules (10.3%) eventhough the sample size of mules
is not representative like the case encountered by Keith (2005) in Amhara and Tigray
regions. According to Robinson (1987) and OIE (2004) AHS causes disease in horses,
mules and donkeys, with up to 95% mortality in susceptible horse population. Robinson
further described that mules have got high morbidity but low mortality. Moreover,
Robinson explained that donkeys are the least sensitive showing only a mild febrile
response without mortality. Sewell and Brocklesby (1990), Hunter (1996) and OIE (2005)
described that donkeys are affected by horse sickness fever, the mildest form, which is
frequently overlooked in natural outbreaks. Furthermore, the cohort study conducted
among the donkey and mule population of Amhara and Tigray regions (Keith, 2005)
confirmed that African horse sickness was not a significant cause of mortality in donkeys,
but it was a significant cause of mortality in mules. In this study the sample size of mules
is not representative because owners were not willing for their mules to be sampled.
According to Galloway (1974), Hunter (1994) and OIE (2004) among equidae, horses are
the most susceptible to AHS with a mortality rate of 50-95%, followed by mules with
mortality around 50%. They further pointed out that, in enzootic regions of Africa
including Ethiopia donkeys are very resistant to AHS and experience only subclinical
infections. Moreover, they described that, in Europeans and Asian countries however,
donkeys are moderately susceptible and have a mortality rate of 10%. From this fact of life
the author can infer that, there is 90-95% chance of recovery in donkeys from infection due
to AHS unlike horses and mules. Contrary to this, exposed horses and mules often die of
AHS viral infection. Field experiences of the author during active disease search and
serosurvey indicated that, getting cases of seropositive recovered horses and mules after
being exposed to the natural challenge is rare in the study areas. The higher seroprevalence

45
observed in donkeys in this study is in agreement with what had been said by the previous
authors.

In this study the seroprevalence of AHSV in different age groups of equidae was assessed..
In the findings there was no significant variation in seropositivity among the different age
groups of equidae (P>0.05). Seroprevalence in the ranges of 20-45% in age category one
(1-7 years), 19-38% in age category 2 (8-14 years) and 19-50% in age category 3 (15-20)
were the findings of this study in donkeys. By the same talken seroprevalence in the ranges
of 5-19% in age category 1 (3-7 years), 6-15% in age category 2 (8-14 years) and 6-24% in
age category 3 (15-20 years) were obtained in this study in horses. Similarly seropositivity
values of 14% in 3 years of age, 17% in 6 years of age, 20% in 8 years of age, 17% in 10
years of age, 33% in 14 and 16 years of age were documented in this study in mules. This
finding is supported by Keith (2005) in that, all foals that have lost their maternal antibody
by six months of age would be protected by vaccination. Keith further described that
different age groups of equidae that are above six months of age had equally seroconverted
and protected after they were being vaccinated. From the findings of this study and Keith’s
survey results the author can infer that all age groups of equidae seem likely to be equally
affected by AHS provided that the equines were not previously exposed and recovered as
well as vaccinated.

In the present study the seroprevalence of AHSV was determined in the two sex groups of
equidae. A Seroprevalence of 31% in female donkeys and 29% in male donkeys; 26% in
female horses and 10% in male horses as well as 17% in male mules were indicated in this
study. In the result statistically significant difference of seropositivity was obtained in the
two sexes (P<0.05). However, the author did not come across previous work/s done
pertaining to seroprevalence and sex of equidae in the present study areas to compare with
the finding of this study. From field experience the author can infer that both sex groups
are equally likely to be affected by AHS.

In this study the seroprevalence of AHSV was assessed in different agro-ecological zones
of the study areas. In the findings statistically significant difference of seropositivity of
AHSV was obtained in the three agro-ecological zones of the study areas. Seroprevalence
rates of 13% in highland donkeys, 3% in highland horses; 49% in lowland donkeys, 24%
in lowland horses; 33% in midland donkeys, 11% in midland horses as well as 10% in

46
midland mules. Mules were sampled only from the midlands. According to NVI (1983) the
distribution of the disease seems to have positive correlation with the ecology of its
vectors. Furthermore, the disease is considered to be endemic to the lowlands and midlands
of Ethiopia. However, some cases of the disease are known to appear in surveyed
highlands of Ethiopia. According to Anon (1998) and Radostits et al. (2000) the breeding
status and movement of vectors is governed by climatic conditions. They further described
that Culicoides have almost worldwide distribution so that spread of AHS is universal.
Mellor et al. (1998) described that increased use of irrigation, water leaks, manure, urine,
dung pats, tree holes, rotting vegetation, stagnant surface water are ideal larval habitats for
the multiplication of Culicoides. Galloway (1974), Anon (1976), Della-Porta (1985) and
Knipe and Howley (2001) explained that, from time to time AHS has spread beyond its
usual distribution, causing devastating epidemics in many countries including Ethiopia.
The report issued by NVI (1983) documented that AHS is prevalent in eight out of the
previous fourteen provinces of Ethiopia. Keith (2005) explained that, it is environment
rather than species or husbandry that is relevant to sero-conversion. Therefore, from the
findings of this study and contextual comparison of the present findings with previous
survey results of other authors in different parts of Ethiopia, it is likely to infer that African
horse sickness exists in all agro-ecological zones of the present study areas.

In this study the association of seropositivity with different potential risk factors that
facilitate the precipitation of AHS as well as the production of antibodies against AHSV
was assessed. In the result the risk of acquiring African horse sickness is more than two
fold (OR=2.1) with respect to the type of equidae affected. Agro-ecology contributes
nearly two fold (OR=1.5) for the occurrence of African horse sickness. The strength of
association among seropositivity and putative risk factors indicated that types of equidae
and agro-ecology contribute for important arrays of interaction to facilitate the occurrence
of AHS. Ages of the animals in question are not part of the interaction (OR=1). On the
other hand sex of equidae has got insignificant effect to favour the occurrence of AHS
(OR=0.9). Hence all age groups and both sexes of equidae are equally affected. However,
the author was unable to get research work/s done pertaining to the degree of association
among seroprevalence rates and assumed risk factors in the present study areas to compare
with the findings of this study.

47
To put in a net shell this epidemiological study is the first of its kind in some of the study
areas and the second of its kind in areas where the National Veterinary Institute undertook
its last epidemiological study in the year 1978 (28 years ago) pertaining to the
epidemiological status of AHS in some of the present study areas.

5.2. Virus Isolation and Identification

Virus isolation and identification was carried out on whole blood collected during active
disease search from clinical cases of horses and a mule. However, the test result is
negative. In the presence of the disease in the field with classical pathognomonic signs and
history of postmortem lesions, the negative result is probably due to the improper handling
of the tissues processed.

5.3. Questionnaire Survey

Knowledge of equine owners on African horse sickness was assessed. The majority of the
owners were found to have no experience of the disease. However, those owners whose
livelihoods depend on equines (cart horse taxi owners) were able to recognize the disease.
They described horses to be more affected than mules and donkeys. Mules were the second
most susceptible and donkeys were seen rarely with the disease.

The majority of owners did not know the age of their equines, because of lack of know-
how on age determination. It was determined that both sexes of equines were equally
affected by AHS. The disease was quite frequently seen in the lowlands and midlands and.
rarely seen in the highlands. Like in all other livestock diseases the respondents were
unable to describe the mode of transmission of AHS.

Equine biting flies are very common mainly in the lowlands followed by midlands. It is not
uncommon to find equine biting flies in the highlands too. Equine biting flies are called by
different names in different regions: Kerchassa in Oromia, Ewir zinb in Amhara which all
mean Tabanus. In the assessment made to know the knowledge base of equine owners
about Culicoides vectors, none of the respondents were be able to know these vectors and
the role they play as well. Rather, they are totally unaware of them. It was conclusively
described that flies are abundant immediately after the rainy season, because the climate is

48
conducive for their multiplication. All equine owners stabled their equines in different
types of houses. Those stables with intact walls and complete cover of roof are insect
proof. Stables with open walls but with complete cover of roof and those fenced ones
without roof can equally predispose equines for insect bite.

Almost all the study areas are located close to permanent water bodies such as lakes,
rivers, watering points, water reservoirs and irrigation canals which support insect breeding
and multiplication. The movement pattern of equines from the highlands and midlands to
the lowlands is a predisposing factor to AHS.

49
6. CONCLUSIONS AND RECOMMENDATIONS

This epidemiological study is the first of its kind in the present study areas since the 1981
comprehensive epidemiological study undertaken by NVI. The significant variation in
seropositivity of African horse sickness virus, serotype nine, obtained in donkeys, horses
and mules in the present study areas of this study indicated that the three equine types have
got differences in the degree of their response to AHSV. Moreover, the disease is equally
likely to affect all age groups of equidae; hence there is no significant variation in
seropositivity among the different age categories of equines. Moreover, both sexes are
equally affected by African horse sickness.

In this study the highest seroprevalence was observed in the lowland followed by midland
and highland areas. Furthermore, types of equidae and agro-ecology contributed more than
two and nearly two folds for the occurrence of AHS, in the given order. Age was not part
of the association and sex has got weak effect to precipitate the disease. Generally AHSV
exists in all agro-ecological zones and affects all types of equidae in the study areas.

The result of virus isolation and identification on vero cell lines was negative which may
be due to improper handling of the specimen during processing. The questionnaire survey
result indicated that the knowledge base of equine owners about AHS is unsatisfactory.
They are conclusively unaware of Culicoides vectors and the mode of transmission of
AHS.
Based on the above concluding remarks the following points are recommended:

 There is a need of deeper understanding of the epidemiology of African horse

sickness in the present study areas through an integrated approach of serotyping
and identification of the Culicoides complex as well as other potential vectors.
 Appropriate equine enclosure system to avoid insect bite by stabling them some
hours before sunset and letting them out a few hours after sunrise, as Culicoides
species are nocturnal in nature and are not inclined to enter equine stables, should
be adopted.
 There is a need to classify the present study areas into endemic and epidemic zones
in relation to AHS. There should be a need of serotyping in out break cases. Based
on the epidemiological situation of the areas and representativeness of the study,

50
 annual vaccination programme in endemic areas by making use of the vaccine that
incorporates the local serotype/s should be practiced. There should be vaccination
strategy that indicates the degree of coverage and protective value of the vaccine.
This is because there have been repeated outbreaks of AHS now and then in the
present study areas that might probably be related to insufficient vaccination
coverage and lack of feed back from field veterinary personnels. Since donkeys are
affected by the disease subclinically, they are potential source of infection for
horses and mules through insect bite. So all types of equines should be vaccinated
against AHS.
 The mechanism of AHSV survival between interepidemics should be studied in
detail and the precise reservoir host be properly identified.
 The seasonal occurrence, feeding behaviour, vector ecology and veterinary
importance of Culicoides complex should be studied through an in depth approach
and be well documented in order to institute economically feasible and practically
applied AHS control interventions.
 Immediate outbreak reporting system to the relevant bodies should be in place.
 The epidemiological status of the disease should be studied in domestic as well as
wild animal species.
 The present knowledge base of the disease among equine workers is not
satisfactory. So there should be awareness creaction among the people through an
organized extension package.

51
7. REFERENCES

Adeyefa, C. A., and Hamblin, C. (1995): Continuing prevalence of African horse sickness in
Nigeria. Rev. Elev. Med. Vet. Pays Trop. 48 (1), 31-33.

Alemayehu, L. (2004): Case studies on reproductive activity of equines in relation to

environmental factors in central Ethiopia. Humboldt University of Berlin: Books on
Demand, GmbH, PhD thesis, p. 1.

Anon (1976): Handbook on Animal Diseases in the Tropics. 3rded. London: British Veterinary
Association, pp. 1-3.

Anon (1998): Foreign Animal Diseases, United States Animal Health Association. Virginia:
Carter Printing Company, pp.41-49.

Aschalew, Z., Teshale, S., Keith, P., Feseha, G., Bojia, E. (2005): Isolation and identification of
circulating serotypes of African horse sickness virus in Ethiopia. In: USA, Veterinary
Solutions LLC. Intern. J. of Appl. Res. Vet. Med, 3 (1), 40-42.

Baba, S.S., Akinyele, H.A., and Olaleye, O.D. (1992): Prevalence of complement-fixing antibody
to the African horse sickness virus in domestic animals in Nigeria. Beitr. Trop.
Landwirtsch Veterinarmed. 30 (4), 471-477.

CACC (2003): Statistical report on livestock and farm implements. Ethiopian Agricultural Sample
Enumeration, 2001/2,Addis Ababa.

CACC (2003): Statistical report on livestock and farm implements. Ethiopian Agricultural sample
enumeration, 2001/2, Oromia Region.

CACC (2003): Statistical report on livestock and farm implements. Ethiopian Agricultural sample
enumeration, 2001/2. Amhara Region.

52
Coetzer, J.A.W., Thomson, G.R., and Tustin, R.C. (1994): Infectious Diseases of Livestock with
Special Reference to South Africa. Vol.1, Cape Town: Oxford University Press, pp.460-
472.

CSA (2004): Federal Democratic Republic of Ethiopia. Central Statistical Authority (CSA),
Agricultural Sample Survey 2003/2004, Volume II, Report on Livestock and Livestock
Characteristics (Private Peasant Holdings) Statistical Bulletin 302, Addis Ababa, June
2004.

EARO (1998): Animal Science Research Strategy and Short Term Plan. Executive Summary,
August 1998, Addis Ababa, p.9.

Farouk, M. (1997): African horse sickness. In: The Merck Veterinary Manual (1997), 8th ed.
Whitehouse station: Merck & Co. Inc., pp. 496-498.

Fenner, F., Bachmann, P.A., Gibbs, E.P.J., Murphy, F. A., Studdert, M. J., White, D. O. (1987):
Veterinary Virology. United Kingdom Edition, London: Academic Press, Inc., pp.587-590.

Fenner, F.J., Gibbs, E.P.J., Murphy, F.A., Rott, R., Studdert, M. J., White, D.O. (1993): Veterinary
Virology. 2nd ed., California: Academic Press, Inc., pp.547-549.

Feseha, G. (1998): Proceeding on Equine Infectious Diseases, 8th International conference on

Equine Infectious Diseases. Dubai, April, 1998 pp. 318-319.

Galloway, J.H. (1974): Farm Animal Health and Disease Control. Philadelphia: Lea & Febiger,
pp. 215-219.

Hall, H.T.B. (1985): Diseases and Parasites of Livestock in the Tropics. Intermediate Tropical Ag.
Series, 2nd ed. Singapore: Print House, pp. 57-61.

Hanson, R.P. and Hanson, M.G. (1983): Animal Disease Control, Regional Programs, 1st ed.
Ames, Iowa, Australia: the Iowa State University Press, p. 12.

53
Hungerford, T.G. (1975): Diseases of Livestock. 8th ed. University of Sydney, Auckland,
Australia: McGraw-Hill Book Company, pp. 833-834.

Hunter, A. (1994): The Tropical Agriculturalist, Animal Health Vol.2, Specific Diseases. London:
The Macmillan/ CTA Press Ltd. pp. 111-113.

Hunter, A. (1996): The Tropical Agriculturalist, Animal Heath Vol.1, General Principles. London:
Macmillan/ CTA Education Ltd. pp. 49, 71.

Joklik, W.K. (1988): Virology 3rd ed. USA: Prentice-Hall International Editions, p. 32.

Keith, R.P. (2005): Survey into the factors affecting the health and welfare of donkeys and mules
in Amhara and Tigray regions of northern Ethiopia. Report on the Health and Welfare of
Donkeys and Mules to the Joint Venture of Addis Ababa University Faculty of Veterinary
Medicine, Debre Zeit, Ethiopia and the Donkey Sanctuary, Sidmouth, Devon, England,
UK, pp. 68-81.

Knipe, D. M. and Howley, P.M. (2001): Fields Virology, Vol.2, 4th ed. Philadelphia: Lippincott
Williams & Wilkins, p. 1861.

Knowles, R. C. (1991): African horse sickness. In: Colahan, P. T., Mayhew, I. G., Merritt, A. M.,
Moore, J. N. (1991): Equine Medicine and Surgery, Vol.1, 4thed. California: American
Veterinary Publications, Inc. pp. 378-380.

Mellor, P.S. and Hamblin, C. (2004): African horse sickness. http: // www. edpsciences. Org.

Mellor, P.S.; Baylis, M.; Hamblin, C.; Calisher, C. H.; Mertens, P.P.C. (ed) (1998): African horse
sickness. Vol. 8, http: // www. aphis. usda. gov/ us/ep/ fad.

Ministry of Agriculture and Fisheries (1991): Surveillance. Exotic Disease Issue, MAF,
Newzealand, 18 (3), 3-4.

Murphy, F.A., Gibbs, E.P.J., Horzinek, M.C., Studdert, M.J. (1999): Veterinary Virology, 3rd ed.
California: Academic Press, pp. 400-402.

54
NVI (1974): National Veterinary Research Institute. Rev. Elev. Med. Vet. Paystrop., 36(2), 117-
129.

NVI (1983): The haematological changes of African horse sickness. Annual report, Debre Zeit,
pp.1-11.

NVI (1983): The pathomorphological changes of African horse sickness. Annual report, Debre
Zeit, pp.1-2.

OIE (2004): African horse sickness. In: Manual of Diagnostic Tests and Vaccine for Terrestrial
Animals, 5th ed., Paris: Office International des Epizooties, pp. 1-21.

Pagot, J. (1992): Animal Production in the Tropics and Subtropics, 1st ed. The Netherlands: The
Macmillan Press, pp. 233, 266.

Palya, V., Prandota, J., Rweyemamu, M.M., Sylla, D. (1994): Quality control testing of rinderpest
cell culture vaccine. FAO Animal Production and Health Paper, Standard Operating
Procedure (SOP), Pan African Veterinary Vaccine Centre (PANVAC), Addis Ababa, pp.1-
78.

Personal communication (2005): Dr. Zelalem Tadesse, Ministry of Agriculture, Epidemiology

Unit, Addis Ababa.

Proceeding of an International Seminar on Virus Diseases of Veterinary Importance in South-East

Asia, held at Australian National Animal Health Laboratory, Geelong, Australia, 27-30
August 1984, Academic Press, pp.89-97. In: Della-Porta, A.J. (1985): Veterinary Viral
Diseases, Their Significance in South-East Asia and the Western Pacific.

Radostits, O.M., Gay, C.C., Blood, D. C., Hinchcliff K.W. (2000): Veterinary Medicine, A
textbook of the Diseases of Cattle, Sheep, Pigs, Goats, and Horses, 9th ed. London: W.B.
Saunders Company Ltd. pp. 1038-1040.

55
Robinson, N. E. (1987): Current Therapy in Equine Medicine-2. Philadelphia: W.B. Saunders, pp.
582-583.

Seifert, H.S.H. (1996): Tropical Animal Health, 2nd ed. Dordrecht: CTA-Kluwer Academic
Publishers, pp. 234-240.

Sewell, M.M.H., Brocklesby, D.W. (1990): Handbook on Animal Diseases in the Tropics, 4th ed.
London: Bailliere Tindall, pp. 257-259.

Stata Corp. (2001): Stata statistical software, Release 7.0. Lakeway Drive, College station, Texas.

Thrusfield, M. (1995): Veterinary Epidemiology, 2nd ed. Edinburgh: Blackwell, p. 339.

Thrusfield, M., Ortega, C., Blas, I. DE., Noordhuizen, J.P., Frankena, K. (2001): Win Episcope
2.0. Improved epidemiological software for Veterinary Medicine. Veterinary record,
148:567-572.

White, G. B. (1977): Man-biting species of Chrysops Meigen, Culicoides Latreille and Simulium
Latreille in Ethiopia, with discussion of their vector potentialities. Department of
Entomolgy, British Museum (Natural History), Transactions of the Royal Society of
Tropical Medicine and Hygien, London, 71(2), pp.164 and 171.

Williams, A.O. and Masiga, W.N. (1988): Viral Diseases of Animals in Africa. Lagos:
OAU/STRC Scientific Publication, p. 174.

Wilson, R.T. (1995): The Tropical Agriculturalist. Livestock Production Systems, Distribution and
Importance of Livestock in the Tropics, 1st ed. London: CTA-Macmillan, p. 5.

56
8. ANNEXES

Annex 1. Expression of results of spot test

The principle of the test is the interruption of the reaction between AHSV antigen
and an immuned anti-AHSV guinea-pig serum by a test serum sample. AHSV
antibodies in the test serum sample will compete with those in the immuned
guinea-pig antiserum resulting in a reduction in the level of expected colour
(following the addition of enzyme labelled anti guinea-pig antibody and substrate).

The optical density (OD) values should be converted to percentage inhibition (PI) values
by using the following formula after the mean of the 4 guinea-pig control is determined:

Percentage inhibition (PI) = 100 - OD of each test or control value x 100

Mean OD of Guinea-Pig Controls

The data expressed as OD values and PI values are used to determine whether or not the
test has been performed within acceptable limits of variability and therefore, whether or
not the test sera data may be accepted for any given microplate. The ranges which are
acceptable for each of the controls are listed below:

OD value range

Guinea-Pig Antibody control >0.4 but <1.4

PI value range

Conjugate control >95

Guinea-Pig control -20 to +20
Positive control serum >50
Negative serum -25 to +25

First level of acceptance:

57
Both intermediate OD values (disregard the highest and the lowest) for the guinea-pig
control must be within the lower and upper limits. If not the plate must be rejected.

Second level of acceptance:

At least three of the quadruplicate wells for the guniea-pig control (PI) values must be
within the limits set above. If more than one is incorrect the plate must be rejected.

Both of the duplicate Conjugate and Negative control (PI) values should be within the
value range. Although failure to fall within the acceptable limits does not provide grounds
for rejecting the plate, this provides a warning that background levels are increasing and
fresh reagents should be used.

The Positive control, which has been titrated up column 1 of the ELISA plate, should
record a 50% end-point of PI, that falls somewhere in the dilution range of 1:120 to 1:480.
In some circumstances, the 50% end-point may fall outside of this range (>1:480),
however a test can still be accepted provided the negative controls are negative and the test
sera are also negative.

Acceptance of individual test sera data

The diagnostic threshold is set at 50% inhibition (50% PI). This is a reduction of
50% of the OD value shown in the guinea-pig control. Any PI value greater than or
equal to 50% is considered positive.

To accept individual test sera results, both of the replicate PI values must fall either
Above (i.e. positive) or below (i.e. negative) the 50% threshold. Test sera should be
re-tested if their replicate PI values lie either side of 50% PI.

If both values are acceptable, the mean PI value for the two replicates is then calculated by
dividing the sum of both values by 2. If the resulting value is greater than 50%, the sample
is considered seropositive for AHS. In some circumstances, you may wish to repeat such
samples on a titration test to determine the level of positivity, by calculating the 50% end-
point.

58
Annex 2. C-ELISA result of AHSV in equidae of each study area.

Sampled equidae
Study areas Result
Total
Donkey Horse Mules
1
Abenogebriel Negative 1 - -
6
Positive 6 - -
57
Ada’a Negative 18 32 7
20
Positive 11 3 6
24
Adama Negative - 24 -
9
Positive - 9 -
43
Addis Ababa Negative 43 - -
7
Positive 7 - -
53
Boset Negative 44 9 -
36
Positive 34 2 -
53
Chollie Negative 43 10 -
11
Positive 8 3 -
244
Debrebirhan Negative 142 102 -
20
Positive 19 1 -
37
Gellan Negative 36 - 1
24
Positive 24 - -
218
Gerado Negative 95 80 43
-
Positive - - -
30
Godino Negative 29 - 1
25
Positive 25 - -
31
Guaworko Negative 31 - -
29
Positive 29 - -
31
Kijeri Negative 25 6 -
21
Positive 16 5 -
79
Kombolcha Negative - 79 -
17
Positive - 17 -
19
Kombollie Negative 18 1 -
19
Positive 19 - -
31
Mendella Negative 31 - -
25
Positive 25 - -
Negative 23 - - 23
Tejitu 22
Positive 22 - -
1265
Total 824 383 58

59
Annex 3. C-ELISA result of AHSV in different ages of equidae.

Types of equidae
Age (yrs) Donkey Horse Mules
Nega. Post. % Neg. Post. % Nega. Post. %
1 20 5 20 1 - - 2 - -
2 34 11 24 7 - - 1 - -
3 31 14 45 15 3 17 6 1 14
4 49 11 18 21 1 5 1 - -
5 29 8 22 11 2 15 3 - -
6 49 35 42 22 5 19 5 1 17
7 43 25 37 21 3 13 5 - -
8 93 22 19 23 3 12 4 1 20
9 45 24 35 15 1 6 2 - -
10 52 25 32 33 - - 5 1 17
11 25 11 31 8 1 11 2 - -
12 36 22 38 32 1 3 1 - -
13 13 6 32 11 2 15 3 - -
14 22 8 27 24 3 11 2 1 33
15 17 4 19 17 1 6 3 - -
16 8 7 47 30 5 14 2 1 33
17 5 3 38 7 1 13 1 - -
18 2 2 50 22 1 4 1 - -
19 1 - - 7 2 22 - - -
20 5 2 29 16 5 24 3 - -

Annex 4. Questionnaire format to assess the indigenous knowledge of equine owners

1.Owner’s address: Name _______________________________

Region________Zone___________District______________

Peasant association_________________

2. Equine owning status: Donkey____Horse______Mule_______

3. Do you know the disease of equidae characterized by supraorbital oedema and dullness?
A / yes B / no
4. Which type of equines is more affected? A / Donkeys B / Horses C / Mules
5. Which age groups of equines are more affected? A / young age B / Middle age
C / Old age D / Difficult to define
6. Which sex of equines is more affected? A / Female B / Male
7. Can you tell the agro-ecology in which the disease is frequently seen?

60
8. Can you tell the mode of its transmission?
9. Are there equine biting insects in your locality? A / yes B / no
10.Can you mention their local names?
11. Which months of the year that flies are abundant?
12. Do you stable your equine/s during the night? A / yes B / no
13. What is the type of your stable in use? A / intact wall with cover of roof B / semi-
intact wall with cover of roof C/ round fence but no cover of roof (Beret)
14. Are there water bodies near to your locality? A / yes, there are lakes, rivers etc. B /
there are no water bodies
15. Is there equine movement? If yes, from where to where?

61
9. CURRICULUM VITAE

1. Personal data
Name: Kassa Demissie Abdi
Date of birth: April 12, 1964
Nationality: Ethiopian
Marital status: Married
Children: None
Religion: Orthodox Christian
Position: Field veterinarian, Team leader, District
Livestock Project Coordinator
Membership: Ethiopian Veterinary Association
Office contact: E-mail: [email protected]
Mobile 0911 99 32 42

2. Educational background

September 1972 – July 19976 E.C Elementary and Junior education

at Deneba Junior Secondary School

September 1977 – July 1980 E.C High school education at Hailemariam

Mamo Senior Comprehensive Secondary
School,
Achievement: ESLCE Certificate
September 1981 – July 1981 E.C Addis Ababa University, Science Faculty
September 1982 – July 1987 E.C Addis Ababa University, Faculty of
Veterinary Medicine
Achievement: Degree of Veterinary
Medicine (DVM)

62
3. Work experience

19 May 1988 – 2 March 1992 E.C Field veterinarian at Metema

District Off. Of Agriculture

25 March 1990 -24 February 1992 E.C Regulatory and protection acting
team leader at Metema District
Office of Agriculture

3 March 1992 – 1 September 1997 E.C Field veterinarian at Chilga district

Office of Agriculture
21 March 1992 – 1 September 1997 E .C District coordinator of Integrated
Livestock Development Project

1 September 1995 – 1 September 1997E.C Team leader of livestock,

Veterinary and fisheries resource
Development

1 September 1995 – 1 September 1997E.C District coordinator of National

Livestock Development Project
And UK Overseas bee keeping
Development Project

4. Research Papers

Camel-Cephalopsis: The Epidemiology and Biodaynamics in dromedary of Eastern

Ethiopia, DVM thesis (1987), FVM, AAU.
Feed resource inventory at Chilga district, ILDP booklet number one, Gondar.

African horse sickness: Seroprevalence and Identification of risk factors in

Equidae at selected sites in Ethiopia, MSc.thesis
(2006), FVM, AAU.

63
5. Additional Training:

1. Continuing Professional Development Training course organized by the Donkey

Sanctuary Working Worldwide and Donkey Health and Welfare Project of Ethiopia
(from 19 to 23 March, 2001).

2. Applied sampling methods, socio-economic and agricultural research data

analysis and interpretation using SPSS and MSTATC at ILDP head quarter
office (9-13 March 2002).

3. Forage and seed production, and inventory held at ILDP head quarter office,
Gondar (15-20 April 2002).

4. Participatory Rural Appraisal organized by ILDP head quarter office and the
Ethiopian Management Institute, held at gorgora port, Gondar (23 Dec.2002 to 1
Jan 2003).

5. Educational tour to National Dairy Development Board of India (NDDB), Anand

, Gujirat state, organized by ILDP and AEDC (1 July – 26 August 2003).

5. Reference Persons

1. Professor Feseha Gebreab: Addis Ababa University Faculty of Veterinary Medicine,

Consultant of the Donkey Sanctuary Working World Wide, Ethiopia, P.o.Box 34,
Tel. 0114 33 85 33, 0114 33 03 04, 0114 33 93 26.
2. Dr. Kelay Belihu, Addis Ababa University Faculty of Veterinary Medicine,
Associate Dean for research and graduate programmes, P.O. Box 34,
Tel. 0114 33 85 33.
3. Dr. Yilkal Asfaw, Addis Ababa University Faculty of Veterinary Medicine,
P.O.Box 34, Tel. 0114 33 85 33.

64
10. SIGNED DECLARATION SHEET

I, the undersigned, declare that the thesis is my original work and has not been presented
for a degree in any other university, and that all sources of material used for the thesis have
been duly acknowledged.

Name Kassa Demissie Abdi

Signature _____________________

Date of submission _______________

This thesis has been submitted for examination with our approval as university advisors.

Dr. Yilkal Asfaw ____________________________

Prof. Feseha Gebreab _________________________

Dr. Berhe Gebreegziabher _____________________

65