Previous PageIndexNext Page


A.A. ARATA (1), B. JOHNSON (2)


1. Vector Genetics and Bionomics, Vector Biology and Control, World Health Organization, 1211 Geneva 27, Switzerland.
2. Virology Department, London School of Hygiene and Tropical Medicine, Keppel Street (Gower Street), London WCIE 7HT, United Kingdom.

This study was carried out in Nzara, W. Equatorial Province, Sudan, (in January-February, 1977) following the protocol agreed upon at the meeting at the London School of Hygiene and Tropical Medicine, 4-5 January 1977.

Mammals having the highest possible contact with humans in the Nzara cotton factory and agricultural environs where primary cases were noted were studied (Table 1). Although primates have been suggested as reservoirs, these were not collected because of the difficulties involved in collecting adequate samples in such a short period.

Organ and tissue samples were collected, frozen in liquid N2 and sent to Porton Down, U.K., for virological and serological studies on 13 February, 1977 (Table 2). Voucher specimens of the animals collected, preserved in formalin, were sent at the same time to the British Museum (Natural History), London, for confirmation of field identifications.

Data on sex, age and habitats (including human contacts) of all potential hosts were recorded to be correlated with virological and serological results. As the virology and serology are not yet available, the paper emphasizes the rationale used in establishing a survey for potential reservoirs of a disease of unknown aetiology.


As we have yet to receive information on material delivered to the Microbiological Research Establishment concerning the isolation and/or serological identification of Ebola virus from mammals within the Southern Sudan area, we nave tried to organize this paper as a general protocol for research for unknown reservoirs in disease situations such as occurred in Southern Sudan.

The first consideration should be given to the apparent ecology of the disease. If an insect vector, such as a mosquito, was involved we might have expected a large number of primary cases. In this instance, there were few primary cases which resembled the situations observed in the case of Lassa virus and other known rodent-borne viral haemorrhagic diseases, or even for that matter bacterial diseases such as plague and tularemia, so that an arthropod such as a tick or a mite is not excluded.

Because the initial concern suspected Marburg virus, it was suggested that primates would be logical reservoirs. However, in the case of Marburg, inoculation with the virus in primates was highly, if not inevitably, fatal suggesting that they were not a prospective reservoir. On the other hand, rodents had tolerated initial Marburg inoculations. Initial studies on other unknown virus reservoirs were, in a sense, rather easy to conceive as in the case of Lassa, once recognized as an arenavirus, it was expected to be associated with rodents because of its similarities to Argentinian and Bolivian haemorrhagic fevers (1,2,3,4,5,6).

It was decided therefore that the most common mammalian species, with the highest contact rates to man in the Nzara area, would be collected. Bats and rodents heavily infested the area of the cotton factory (Figures 1 & 2) in which the primary cases were noted, and field rodents were collected in the areas surrounding where the primary cases lived in and about dwelling compounds (Figure 3). (7,8, 9).

As bats live a long time, we feel that we collected animals that would have been alive during the period of the epidemic. Life expectancy of rodents is much shorter, about 10-12 months, (varying by size, species, habitat, population levels and from year to year. Our collections were made in mid-January to mid February (1977), 8-9 months after the initial outbreak so the adults should have been young during the outbreak. Considering potential antibody levels in the animals we collected, we aged all mammals as age structure is important in antibody and isolation studies during post-epizootic analysis. Too often this is not considered when working with non-human mammalian species. In general, therefore, we feel that we collected a high percentage of mammals assumed to be alive in the early and acute stages of the epidemic.



Total Number Collected

British Museum Insectivora (Shrews) (Natural History) Registr. Numbers of Voucher Specimens

Insectivora (Shrews)


Crociduva flavescans


B.M.(N.H.) 1977 661-667

Chiroptera (Bats)


Epomophevus anupus


B.M.(N.H.) 1977 595-596

Micropterus pusillus


B.M.(N.H.) 1977 597-598

Taphozous mauvitanius


B.M.(N.H.) 1977 599-608

Sootoecus hirundo


B.M.(N.H.) 1977 609-618

Scotoecus hindei


B.M.(N.H.) 1977 619-620

Tadarida (Mops) nanula


B.M.(N.H.) 1977 621-625

Tadarida (Mops) trevori


B.M.(N.H.) 1977 626-660

Rodentia (Rodents)


Thamnomys (Gramomys) avidulus


B.M.(N.H.) 1977 3240-3242

Lermiscomys striatus


B.M.(N.H.) 1977 3243-3251

Aethomys sp.


B.M.(N.H.) 1977 3252-3255

Rattus rattus


B.M.(N.H.) 1977 3256-3267

(1)Mastomys (Praomys) natalensis


B.M.(N.H.) 1977 3268-3295

Lophuromys sikapusi


B.M.(N.H.) 1977 3296-3300

Cricetomys gambianus


B.M.(N.H.) 1977 3301-3302

Cryptomys ochraceocinereus


B.M.(N.H.) 1977 3303











137 voucher specimens were deposited in the B.M.(N.H.) - 66 bats 64 rodents 7 shrews and 5 batraciens.

(1) In other parts of Africa this nominate species is certainly a species complex based on cytotaxonomic studies which could not be carried out in the Sudan because of time and facilities.







Liver spleen and kidneys


(pooled by individuals)




(Collected based on recorded retention of Ebola virus in this tissue in an accidental human case)


Salivary glands (bats only)


(Bats only; as these tissues are good sites of bat-borne viruses)







(From 501 vertebrates)


All positive traps were soaked in hypochloride solution each day. Trap sacs (as used for plague studies where fleas are involved) were not used because of possible urine contamination. Baits were made of local grains, mostly maize, boiled with vegetable oils or tinned fish, the latter of which proved more suitable.

Fig. 1. Exterior of Nzara cotton factory where primary cases of African haemorrhagic fever were working. Bat roost (Tadarida (Mops) trevori) was in attic of last unit on the right.

Fig, 2. Interior of store room in Nzara cotton factory where primary cases worked. Droppings and urine can be seen at the corners of the walls.

Fig. 3. Vegetational characteristics of the dwelling compound now abandoned according to tribal tradition of the primary case who worked in the cotton factory store room. His tomb is in the foreground.

Fig. 4. Outside working area in Nzara where mammals were bled and organs collected.

Only the two of us (Arata and Johnson) handled animals during removal from traps, dissection, taking of blood, sera and organs, and disposal. Others assisted in setting and collecting traps. Gowns, visors, masks, gloves, etc., were used whenever possible and dissection and organ collection were done outside in open air with breeze (sometimes too much) (Figure 4). Only primitive sterilization was available (boiling of instruments in pressure cookers on Primus stoves, boiling clothes and gowns daily in large local kettles over open fires, burning of all disposable items especially syringes, gloves, masks, caps, etc.: all animal carcasses were incinerated with petrol in oil drums).

Voucher mammalian specimens were preserved and sent in formalin (10% + 1-2% glycerine) to the British Museum (Natural History), London. Each was numbered individually in the field after dissection to cross reference ecological condition of collection sites and the estimated age (juvenile, sub-adult or adult) of individuals. Also, as the fauna of this area of Africa on the forest savannah belt is poorly known, accurate determination by species, and occasionally by genus was difficult in many instances under such field conditions (e.g. rodents and bats are specifically identified by tooth form and skull structures, meristics, colour variants, etc., which must be compared to museum collections). British Museum (Natural History) numbers are provided for representatives of each of the mammalian taxa studied (Table 1).

Good local facilities, as were available, were provided: all equipment arrived, albeit late. Khartoum provided a good supply of liquid N2 during the study and the local authorities and population were excellent in collaboration.


The area about Nzara, Western Equatorial Province, where the epidemic is presumed to have begun is on the northern level of the Congo forest in a transitional zone of derived Guinea savannah - the vegetation and fauna demonstrate this. The bats collected (personal communication - J. Hill, British Museum (Natural History) indicate that this faunal component represents a southern sylvatic element with few savannah species involved. In fact, our collections provide northern records for several mammalian species, and Mr.Hill noted that the known range of Tadarida (Mops) trevori, expanded by our collection, closely paralleled the distribution in Sudan and Zaire, the area of the epidemic. This bat species was collected from a large roof colony in the cotton factory directly over the storeroom where the primary cases worked (Figure 2) and which was laden with faeces and urine. Several bats were collected in the room on the desk of the unfortunate primary case. Although conjecture only, it is interesting.

On the other hand, we collected few true savannah rodents (possibly due to the short time). However, Arvicanthis, a grassland form, was absent in our collections but has been reported as being common north of Nzara. Mastomys, which reaches the northern limits of its distribution in this area, was the most common rodent encountered, but our collections were not in sylvatic areas but only where human cases were noted, all of which were highly "cut-over" areas. Mastomys will live with Rattus rattus, and both were taken in traps set in several areas side by side (10,11).

As in many areas from east to west Africa, there is increasingly derived transition between the Guinea savannah and the original forest zones to the south providing adequate and enlarging habitats for mammals adapting to human environmental changes and making new links between sylvatic and domestic cycles.

It should also be noted that human modification of environment, especially for agricultural purposes which require the greatest land usage, while eliminating certain "wild species" selects for others, which, partially freed from natural predators, may achieve very high populations. This results in potentially greater contact with man thereby enhancing the transmission of adaptable pathogens through "favoured" reservoirs and vectors to man. Numerous small mammalborne diseases show these characteristics. Agricultural practices related to increasing human populations clearly show this phenomenon in both insect vectors, and vertebrate reservoirs of numerous diseases. Occasionally, the reverse might be true. However, we would suggest that as this phenomenon increases, and as surveillance and diagnostic techniques improve and become more widespread we may expect to discover "new" diseases which may show epidemic rather than endemic characteristics.


Until our samples for serological and viral isolation studies have been examined, we have nothing further on the virological side to report regarding possible Ebola reservoirs, other than what we did in Nzara and how and why we did it. However, this exercise has made us reconsider the design of reservoir studies. Above all, there is the need for taxonomic and ecological relationships, the latter relating to man and his use of the environment vis-a-vis the suspect reservoirs. Suspect reservoirs, for control purposes, are those most directly in contact with man, even if they are secondary reservoirs or amplifying hosts from an unknown sylvatic focus. Often medical zoologists working in poorly known areas have ascribed to a "catch all and analyze all'' philosophy. However, there are places in the world (like Nzara) where time and resources are very limited and one could not afford (in three weeks) to do a general faunistic survey. Although experience and a certain amount of intuition are required, some better approaches should be made:

(a) Is the reservoir survey to be serological ? Do we do the survey at the time of convenience of the investigator (university or governmental) or based upon the known or presumed ecological structure of the rodent population ? For example, the percentage of adults and juveniles may fluctuate widely throughout the year in temperate zones and in most dry tropics. The wet tropics differ considerably characterized by more modest fluctuations with fewer young susceptibles in any one period of the year due to more balanced annual reproduction.

(b) In temperate zones there are fewer potential species of reservoirs (and vectors) but most are linked with high seasonal reproductive cycles that tend toward epizootic patterns as they are species/density dependent. Dry tropics show similar patterns, but in the wet tropics, with a wealth of species, it is different. 5 Species are abundant but the number of individuals of each species are generally lower in number with greater seasonal stability producing a "buffering" effect, thus providing an enzoonotic expression of disease in the animal population. Of course, as mentioned above, this balance can be easily shifted if the ecological conditions are changed, especially by rapid land use schemes.


Although this work is still incomplete, and others will have to be recognized we would like to thank especially Dr. Abdel Nuur and his technicians Seif el Din and Al Taraife who worked in the field with us, Dr. Babiker el Tohir who arranged for liquid nitrogen shipments, and Dr. S. Singh (WHO, Juba), who organized communications and supplies most effectively. W.N. Jonathanson, Senior Controller, Pub]. Administration, Agr. Production Corporation, Nzara, assisted us greatly in Nzara and Mr. Culcliffe (FAO) was most helpful with communications in the Nzara area and provision of transport until ours arrived. Milghani, the local physician in Nzara, was very considerate in assisting with local arrangements.

1. I.Fabiyi, A. (1976) Lassa fever (Arenaviruses) as a public health problem, PAHO Bull., Vol. X, N'4, 335-337.
2. 2.Keane, E., Gilles, H.M. (1977) Lassa fever in Panguma Hospital, Sierra Leone, 1973-76, British Medical Journal, 1, 1399-1402.
3. 3.Mackenzie, R.B. (1972) Public health importance of rodents in South America, Bull. Wld. Hlth. Org., 47, 161-169.
4. 4.Mercado, R.R. (1975) Rodent control programmes in areas affected by Bolivian haemorrhagic fever, Bull. Wld. Hlth. Org., 52, 691-696.
5. 5.Webb, P.A., Justines, G., Johnson, K.M. (1975) Infection of wild and laboratory animals with Machupo and Latino viruses, Bull. Wld. Hlth. Org., 52, 493-499.
6. 6.Dowdle, W.R. (1976) Marburg Virus. Current Developments in Virology, PAHO Bulletin, Vol. X, No4q p. 333.
7. 7.Center for Disease Control (1977) Morbidity and Mortality Weekly Report "Viral Hemorrhagic Fever - Sudan and Zaire", Vol. 26, Nº26 (July), 209-216.
8. 8.WHO (1977) Weekly Epidemiological Record - Viral Haemorrhagic Fever, Wkly. Epidem. Rec., 52, 185-192.
9. 9.WHO (1977) Viral Haemorrhagic Fever (Corrigendum, WER 1977, No21, pp.177-180) Wkly. Epidem. Rec., 52, 229236.
10. 10.Gratz, N.G., Arata, A.A. (1975) Problems associated with the control of rodents in tropical Africa, Bull. Wld. Hlth. Org., 52, 697-706.
11. 11.WHO (1975) Biological Hazards Associated with mastomys, Wkly. Epidem. Rec. N-33, 289-296.
S.R. Pattyn : Are there any results on the virological examination of these specimens ?
E.T.W. Bowen : We have tested to date approximately one hundred of these small mammals. Our problem is that it would be bad practice to carry out these primary isolation studies while working at the same time with the virus itself. We have to fit it in with our own experimental work. We are in fact mainly a diagnostic laboratory and we get requests for Lassa fever studies, Marburg studies, Ebola studies, so we have to fit all this in. I do apologize for the slowness in processing but we should get it through within the next six months. We tested about a hundred I think, mainly rats, and a few bats and so far all are negative.
J.G. Breman : Testing 1.200 laboratory specimens for virus seems to be an enormous job. Wouldn't it be reasonable to start with serum surveys? Perhaps this is difficult because of the need for specific species conjugates.
A.A. Arata : I think there is something that many laboratory oriented people do not recognize very well about animal populations. Collecting animals in a given place at a given time, can produce about 75% juveniles and about 25% adults resulting from a very high reproduction rate during that period of time Collecting the same population in the same location at another moment could produce 90% adults and only 10% juveniles. Therefore, the understanding of an animal population is extremely important in the design of the collection, and whether one is going to start with isolation or looking for antibodies. In certain populations in Central Asia, all the reproduction is concentrated at one period of the year. There is estivation in the summer, hibernation during the winter and all animals ape born within a nice little one or two months' period. In the dry tropics, there is a tendency for the same situation. In the wet tropics, rodent populations tend to have a very equitable reproductive Pate, so that there is always a small number of susceptibles in the population and never a massive group. One system tends toward an epidemic, epizootic situation, the other will tend toward an enzootic or endemic situation. It is extraordinarily important that those who collect and consider this type field studies try to obtain some idea of the ecology of the species that they are dealing with.
K.M. Johnson : I agree with everything you said. But, by and Large, it is probably true that when Looking for something in nature, it is more easy, on a numerical basis, at any time to find evidence of antibodies against infection than it is to isolate the agent itself in the individual animal. Although there are exceptions : some of the arenavirus infections in their specific rodent reservoirs are a good example.
A.A. Arata : I would agree and this is why from these 500 animals, there were sera, and the Laboratory would start with the serology before the isolation attempts. But again, if you deal with older animals you look for the serology, if you happen to be at an epizootic, then you should concentrate on the younger animals and do isolation because there would not be enough time for antibody formation.

Previous PageTop Of PageNext Page