Rev. sci. tech. Off. int. Epiz., 2000, 19 (1), 41-54

Climatic and geographic influences on arboviral infections and vectors

P.S. Mellor ( 1 ) & C.J. Leake ( 2 )

(1) Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 ONF, United Kingdom (2) London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, United Kingdom

Summary Those components of climate that are likely to have major effects upon the geographical distribution, seasonal incidence and prevalence of vector-borne diseases are described. On the basis of existing and predicted climatic variations, examples are given of the types of changes that are to be expected, using several internationally important human and animal arboviral diseases including Rift Valley fever and African horse sickness.

Keywords Arboviruses - Climatic variables - Distribution - Public health - Temperature - Vectors -Zoonoses.

Introduction In 430 BC Hippocrates said, ' W h o e v e r w o u l d study med ic ine aright must learn of the following subjects. First, h e must consider the effect of the seasons of the year and the differences b e t w e e n them. Second, h e must study the w a r m and cold winds, b o t h those w h i c h are c o m m o n to every country and those peculiar to a particular locality'

F rom this, it can b e s e e n that the influence of climate o n infectious diseases has b e e n a topic of s o m e considerable interest for over 2 ,000 years. Recently an increase in that interest has b e e n fuelled b y the various global warming scenarios in circulation, and highlighted b y the fact that eight of the warmest years o n record have all occurred w i t h i n the last decade or so , thereby seeming to support the concept of imminent climate change. Indeed, mos t authorities n o w accept that the ba lance of ev idence suggests that h u m a n activities are warming the planet, and climate m o d e l s predict an increase in global m e a n temperature of b e t w e e n 1°C and 3.5°C during the 21st Century (33 , 34) . Further predict ions suggest that m a x i m u m warming will occur at h igh latitudes and during the winter, and that temperatures will increase more during the night than the day (40, 56 ) . Accompanying changes in rainfall, wind patterns and seasonal weather variations are also expec ted .

Effect of climatic variables on vector biology By definition, vector-borne diseases possess a vector stage, usually a crustacean, mollusc, insect o r acarid, w h i c h is

poik i lo thermic (cold b l o o d e d ) and h e n c e is especially sensitive to changes in climatic variables. Temperature and humidity are the componen t s of the macrocl imate w h i c h are l ikely to have the mos t important direct effect o n vector b io logy and eco logy (28, 4 4 , 90 , 110) . However , m a n y vectors demonstrate behavioural preferences for particular microcl imates and the impact of climate change o n microcl imate environments has only occasionally b e e n cons idered (68) . Major macroclimatic variables wi th direct effects o n vector b io logy are descr ibed b e l o w .

T e m p e r a t u r e A rise in temperature accelerates the metabol ic rate of a vector, increases biting rates and will m a k e b loodfeed ing m o r e frequent. This should lead to enhanced egg product ion and an increase in popula t ion size. However , the daily survival rate of individual vectors is l ikely to decrease as temperature rises and there will b e an upper limit b e y o n d w h i c h h igh temperature is positively detrimental. This is analagous to an extension of the ho t dry season in many parts of the tropics, w h e r e vector activity is significantly reduced prior to the cooler rainy season in w h i c h the lower temperatures have an oppos i te effect. Temperature m a y also affect the geographical range or distribution (in terms of latitude and altitude) of vectors since this tends to b e limited b y m i n i m u m and m a x i m u m temperature (and humidity) .

Humidity High relative humidity favours mos t metabol ic processes in vectors s o that at h igher temperatures, a h igh humidity (i.e. a l o w saturation deficit) will tend to prolong survival, a l though

4 2 Rev. sci. tech. Off. int. Epiz., 19 (1)

increased susceptibility to fungal and bacterial pathogens may offset this to a variable degree. L o w humidity (i.e. a h igh saturation deficit) causes a decrease in the daily survival rates of many ar thropod vectors because of dehydration, bu t i n s o m e cases it may also cause an increase in the b loodfeed ing rate, in an attempt to compensa te for the h igh levels of water loss.

Rainfall Rain is an important factor to mos t b loodfeed ing groups of insects, including the blackflies (Simuliidae), biting midges (Culicoides), horse-flies (Tabanidae) and mosqui toes (Culicidae), because the immature larval and pupal stages are aquatic or semi-aquatic. Rainfall frequently limits the presence, absence, size and persistence of breeding sites. T h e precise level of impact o n such sites will d e p e n d u p o n local evaporation rates, soil type, s lope of terrain and the proximity of large b o d i e s of water (e.g. rivers, lakes, ponds ) . However, since many insect vectors b reed in seepage of water from irrigation p ipe leaks , cattle trough overflows and residual water from previous rains, very heavy or pro longed rain may disrupt vector breeding sites and wash the immature stages away or kill t h e m directly.

Wind Since winds contribute to the passive dispersal of many species of flying insects, prevailing winds and wind s p e e d m a y have a significant effect u p o n vector distribution. S o m e insect vectors, including species of mosqui to , Simuliidae, sandflies (Phlebotominae) and Culicoides can b e dispersed for hundreds of k i lometres in this way.

In addition to these direct effects, climatic variables may also have important indirect effects u p o n vector abundance and distribution, and h e n c e u p o n disease. One vector species m a y b e displaced b y another with a different vectoral capacity in response to environmental changes s u c h as deforestation, expansion in irrigation or increase in b rack i sh water breeding sites due to a rise in sea level (56) .

Effect of climate and other factors on the distribution of insect vectors Palaeoclimatic records s h o w that mos t shifts in the distribution of insect taxa have b e e n associated with temperature change and that these shifts have occurred m u c h more rapidly than changes in the distribution of vegetation and the higher animals (23 , 24 ) . Considering the h igh level of mobility of insects, especially winged insects, this should c o m e as n o surprise and strongly suggests that the current distribution of many insect vector spec ies wou ld change rapidly following any future cl imate change . In this context, a 1°C rise in temperature h a s b e e n est imated to correspond to 90 k m of latitude and 150 m of altitude (79) .

Human activity The introduction of Aedes albopictus, an efficient vector of yel low fever and dengue viruses, into Italy in 1989, demonstrates that even winged insects d o not always have to

extend into n e w areas of suitable climate under their o w n power . This mosqui to was probably introduced as drought-resistant eggs laid in second-hand car tyres that h a d b e e n impor ted from the United States of America (USA). Aedes albopictus is n o w found in m u c h of north and central Italy (38) and at several locations in Albania (3). Since the distribution of Ae. albopictus is b o u n d , conservatively, b y the m e a n 0°C isotherm in the coldest m o n t h of the year (77) , this suggests that in Europe, where this i sotherm encompasses mos t of the continent south of Scandinavia, Ae. albopictus m a y already b e able to extend into mos t or all of these areas. In the USA, w h e r e the mosqu i to was in t roduced f rom J a p a n in 1985, Ae. albopictus n o w occupies 678 counties in 25 south-eastern States ranging as far nor th as Nebraska and Iowa (73). T h e species is also progressing southwards through Florida at the rate of 65 k m per year, displacing Ae. aegypti. Should the pro jec ted temperature increases occur, then the range of this mosqui to is l ikely to extend to include mos t of the popula ted areas of eastern USA, in addit ion to southern Canada (77).

Aspects of insect physiology Diapause It is of interest to no te that b o t h temperate and tropical geographical strains of Ae. albopictus are k n o w n . T h e temperate strains are physiologically competen t to enter diapause (hibernation) during per iods of co ld weather b y switching feeding behaviour f rom b l o o d to plant juices , thereby producing large quantities of lipids from the ingested plant sugars. Tropical strains cannot m a k e this switch and therefore have to remain within temperature ranges where activity is possible year-round. Thus, the southward spread of temperate Ae. albopictus in North America is already likely to b e close to the uppe r thermal limit. As temperature increases still further, the current distribution of the species in Florida m a y b e c o m e m o r e limited.

Transovarial transmission In temperate areas, many arboviruses are l ikely to b e maintained over the co ld winter per iods b y transovarial transmission through the virus vectors (41) .

Effect of temperature on vector competence W h e n attempting to predict the effect of climate or climate-change o n the distribution of vector-borne diseases, temperature-related vector and pa thogen deve lopment rates must b e considered simultaneously. In many cases, the successful comple t ion of the developmenta l cycle of a pa thogen within the vector m a y occur only within a clearly defined temperature band . Therefore, for e ach vector-pathogen combinat ion, a limiting temperature range is likely to exist, outside w h i c h the pa thogen will invariably fail to b e transmitted, al though the vector itself m a y b e able to survive reasonably or very well . This b a n d of 'permissive' temperatures may b e unique to each pa thogen. Within this b a n d is l ikely to b e a narrower set of temperatures within w h i c h the pa thogen is transmitted mos t efficiently, usually

Rev. sci. tech. Off. int. Epiz., 19(1) 4 3

because pa thogen deve lopment is rapid and vector survival rates are h igh. Beyond this narrow band , transmission will b e c o m e progressively less efficient due to a lower vector survival rate, a s lower or defective pa thogen developmenta l cycle or a combinat ion of these factors.

The a im of this paper is to consider the above climatic effects u p o n vector-pathogen interactions and disease distribution, using appropriate examples of human , animal and zoonot ic vector-borne diseases.

Mosquito-borne encephalitides Japanese encephalitis This mosqu i to -borne flavivirus is currently the mos t important cause of vector-borne encephalitis in humans , causing s o m e 45 ,000 clinical cases and 11,000 deaths annually (14) . Japanese encephalit is virus (JEV), was probably first identified clinically in J a p a n in the late 1800s with the first major ep idemic r ecorded in that country in 1924. T h e virus was first isolated from humans (from the brain of a fatal case) in 1935, and from the principal mosqu i to vector, Culex tritaeniorhynchus, in 1938. Currently, the k n o w n distribution of the viras covers all areas extending westwards from J a p a n to the bo rde r areas of western Pakistan and India and f rom the Himalayan barrier in the north to a largely equatorial distribution including Sri Lanka , Malaysia, Indonesia and the Philippines in the south. Virus activity has also b e e n repor ted from islands in the Torres Strait and JEV is therefore considered a threat to the mainland of Australia.

In general, recent ep idemic activity has b e e n repor ted across the northern part of the extensive range of JEV, with major epidemics occurring in Japan , the People 's Republic of China, Vietnam, Cambodia , nor th Thailand, nor th India and Nepal (42). Comparatively little activity has b e e n repor ted f rom Myanmar (formerly Burma), Bangladesh and Laos, a l though this may b e due to p o o r or under-reporting (42) . In J a p a n and the Republic of Korea , extensive use of a commercia l ly available vaccine has resulted in a dramatic decl ine in cases reported since the 1960s. However , the zoonot ic transmission cycle of the virus is continuously maintained and represents an ever-present publ ic hea l th threat, requiring a continuous public heal th vaccination strategy.

Epidemic transmission of JEV is strongly influenced b y climate. T h e p e a k transmission pe r iod is focused in a few w e e k s at the end of the tropical h o t season. During this period, temperatures are typically 32°C or higher, allowing rapid virus replication in the vector, and the beginning of the rains provides numerous breeding sites leading to h igh vector productivity. As temperatures coo l , ep idemic activity invariably ceases . Closer to the equatorial regions, average p e a k temperatures are typically several degrees centigrade lower, and case reporting b e c o m e s m u c h m o r e sporadic.

Japanese encephalit is virus has a c o m p l e x ep idemiology , be ing maintained in a continuous mosqui to-animal zoonot ic cycle associated with rice-field ecosystems. T h e animals w h i c h act as virus-ampfifying hosts are various species of ardeiid water birds and domest ic swine. T h e mos t important maintenance and ep izoot ic /ep idemic vectors over m u c h of the range are thought to b e a number of mosqu i to species in the Sitiens group of the subgenus Culex. T h e principal vector is Cx. tritaeniorhynchus, a species w h i c h is subjected to a wide range of climatic variation across an extensive geographic range. In marit ime Siberia and north Japan , areas at the ext reme edge of the range of JEV, the winters are very cold . This p o s e s severe p rob l ems for the maintenance of b o t h the virus and the vector. T h e popula t ion dynamics of Cx. tritaeniorhynchus are strongly l inked to temperature in these areas (84) , and individuals m a y hibernate over the winter as unfed adults. In such locations, therefore, species of floodwater Aedes mosqu i toes , such as Ae. japonicus, are. thought to maintain the virus over the winter b y transovarial transmission in drought resistant eggs (86) . In tropical climates, popula t ion patterns of Cx. tritaeniorhynchus are m o r e closely associated with available moisture, either in the form of annual rainfall or rice-field irrigation, and adult activity usually continues throughout the year.

Experimental studies wi th vectors of JEV, under controlled climatic conditions, have demonstrated that mosqui toes maintained under autumn or winter condit ions carry only l o w concentrations of virus w h i c h are restricted to the posterior mid-gut cells. In contrast, h igh titres and disseminated infections occur under summer condit ions (88) . At 28°C, 60% of Cx. tritaeniorhynchus transmit virus nine days after infection, and 100% after fourteen days, whereas n o transmission was de tec ted at 20°C for u p to twenty days post infection (100) . A similar effect of temperature was also s h o w n b y L e a k e and J o h n s o n w h o demonstra ted positive salivary gland infections in inoculated Cx. tritaeniorhynchus after five days incubation at 32°C, bu t only after twenty-one days at 26°C (43). In a summary of vector c o m p e t e n c e studies, Burke and L e a k e demonstra ted that the predominant ly tropical mosqui to species , Cx. tritaeniorhynchus, Cx. gelidus and Cx. fuscocephalus, are efficient vectors of JEV, whereas m o r e temperate Culex species , such as Cx. tarsalis (a vector of equine encephali t ides in North America) are significantly less efficient (14) . T h e findings suggested that increases in environmental temperature could convert less efficient species into important vectors, thereby significantly expanding the potential range of JEV.

Transmission of St Louis and western equine encephalitis virus in North America O n the basis of field and laboratory studies per formed in California, Reeves et al. forecast that a rise in environmental temperature of 3°C to 5°C would cause a significant northward shift of b o t h St Louis encephalitis virus (SLEV) and western equine encephalitis virus (WEEV) (83). T h e latter is

4 4 Rev. sci. tech. Off. int. Epiz, 19 (1)

moving northwards and is predicted to disappear from many regions in North America where it is currently endemic , while the range of SLEV will extend m u c h further northwards into Canada, where sporadic outbreaks of disease already occur. These predicted changes are based partly o n the fact that the lowest temperature at wh ich WEEV transmission can occur via the vector Cx. tarsalis is 11°C while for SLEV the m i n i m u m temperature is 15°C (56). At temperatures above 30°C, Cx. tarsalis modulates WEEV infection so that the propor t ion of infected vectors decreases dramatically, thus reducing the overall transmission rates. However, SLEV infection is not modula ted b y the vector and so the transmission of the virus will not b e similarly adversely affected at the higher temperatures.

African horse sickness virus Introduction African horse sickness virus (AHSV) is a double-stranded ribonucleic acid (RNA) virus that causes an infectious, non-contagious, ar thropod-borne disease of equids and occasionally dogs (African horse sickness [AHS]). T h e disease is particularly severe in horses , killing 80% to 9 5 % of naive animals, but infection is virtually sub-clinical in zebra and donkeys . Although at least four h u m a n cases of severe disease have b e e n documented , AHS is not a zoonosis (25 , 101). Nevertheless, the principles involved in transmission of AHS b y insect vectors are identical to those of many ar thropod-borne zoonot ic diseases and AHS can therefore serve as an instructive m o d e l to illustrate the effects of climatic variables u p o n the ep idemio logy and distribution of vector-borne viral diseases in general.

Distribution and epidemiology Nine distinct serotypes of AHSV exist, and these are transmitted b e t w e e n vertebrate hosts almost entirely b y species of Culicoides biting midge . As the n a m e implies, AHSV originated in Africa and is enzootic in tropical and subtropical areas south of the Sahara in a b a n d stretching from Senegal in the west to Ethiopia and Somalia in the east, and extending as far south as northern South Africa (31 , 65) . T h e Sahara desert seems to provide a geographical barrier w h i c h has prevented any permanent spread of the virus to North Africa, or beyond .

Until relatively recently, the virus was be l ieved to b e confined to continental Africa, except for occasional excursions into Yemen (70, 81) . However , from 1959 to 1961 , AHSV serotype 9 (AHSV-9) m a d e a dramatic extension out of Africa and spread along a b r o a d front, across the Arabian peninsula and the Middle East, reaching as far as Pakistan and India (31). Nevertheless, b y the end of 1961 , as the result of a massive vaccination campaign and the deaths of over 300,000 equids, the disease was eradicated in Asia (7). T h e failure of the virus to persist in Asia was p robab ly d u e to a combinat ion of factors, including vaccination, vector control and m o r e

particularly, adverse climatic condit ions that r educed or prevented adult vector activity during the winter per iods .

During 1965, AHSV-9 again spread b e y o n d the enzoot ic zones in sub-Saharan Africa, appearing first in Morocco and then in Algeria and Tunisia, be fo re crossing the Straits of Gibraltar into southern Spain (62) . T h e virus persisted in North Africa until 1966, be fo re finally disappearing from the area, but the extension into Spain lasted less than three w e e k s (19) .

After 1966, reports of AHSV were l imited to sub-Saharan Africa for over 20 years. However , in July 1987, an ou tb reak of AHS due to serotype 4 (AHSV-4) was confirmed in central Spain (50). T h e mos t l ikely source was a number of zebra wh ich were impor ted f rom Namibia into a safari park where horses were present, approximately 45 k m south-west of Madrid (50) . Zebra are susceptible to infection wi th AHSV, but unl ike horses , usually s h o w n o clinical signs. This safari p a r k subsequently b e c a m e the site of the first twenty-seven cases of AHS in Spain in 1987. T h e epizoot ic continued for three to four mon ths in central Spain, apparently ending in October (20).

T h e 1987 ou tbreak in Spain caused alarm because AHS is a notifiable disease and Madrid was the furthest nor th that the virus h a d ever b e e n recorded . However , concern was mode ra t ed because all previous information suggested that the virus is incapable of overwintering in Europe. Consequently, the apparent e n d of the epizootic in October 1987 was in line with expectations and n o further outbreaks w e r e expec ted . Unfortunately, this was not the case, additional severe outbreaks fol lowed in Spain from 1988 to

1990, in Portugal in 1989 and in Morocco from 1989 to 1991 , before vaccination and zoosanitary measures eliminated the virus from the area. All of these outbreaks were due to AHSV-4, a serotype that h a d never previously b e e n recorded outside southern Africa. During the course of the outbreaks n o other evidence was found of any serotype of AHSV within 3,000 k m of Spain and northern Morocco . Consequently, it s eems certain that there was only one introduction of the virus into the area and that was via the impor ted zebra in 1987. Subsequent to this introduction, the virus h a d persisted in the area for at least five years, overwintering four times in the process .

This situation was unprecedented; nowhere else and at n o other time h a d any serotype of AHSV succeeded in overwintering m o r e than twice outside sub-Saharan Africa. Much speculation ensued as to h o w and w h y the situation h a d changed.

African horse s ickness virus causes a viraemia in horses and zebra that m a y extend for 18 and 40 days, respectively, though in b o t h species the duration of viraemia is usually m u c h shorter. T h e virus is transmitted b e t w e e n equ id hos ts

Rev. sci. tech. Off. int Epiz., 19 (1) 4 5

virtually entirely b y Culicoides biting midges , w h i c h are biological vectors. These insects are only able to transmit the virus b y biting; n o evidence of transovarial transmission exists. In m u c h of sub-Saharan Africa, the climate allows the adult vector Culicoides to b e active throughout the year. Consequently, AHSV survives in these enzoot ic areas, through continuous and uninterrupted cycles of transmission be tween the vertebrate and invertebrate hosts . Outside sub-Saharan Africa, the failure of AHSV to persist h a d b e e n attributed mainly to an absence of efficient vector spec ies of Culicoides or, if such vectors were present, to their seasonal incidence due to the relatively harsh winters exper ienced in such areas. However , for the virus to have b e e n able to survive in the western Mediterranean bas in for five consecutive years, efficient vector species of Culicoides must b e present in Spain, Portugal and Morocco , and in at least s o m e areas of these countries, the climate must b e sufficiently mi ld for adult vectors to b e active throughout the year. In these areas, any vector-free per iods that d o occur, must b e shorter than the max imum duration of viraemia in the local susceptible vertebrate popula t ion (i.e. 18 days in this case) , otherwise the last infected equ id will have d ied or recovered before n e w vectors arrive to continue the cycle.

Culicoides imicola: the only proven field vector of African horse sickness virus The only p roven field vector of AHSV is Culicoides imicola (61). This is an Afro-Asiatic species w h i c h is c o m m o n throughout Africa and m u c h of South-East Asia. Until recently, the species was thought to b e absent f rom Europe but it is n o w k n o w n to b e widely distributed throughout south-west Spain and mos t of Portugal, ranging as far nor th as 41°5'N and including all of the areas in w h i c h AHS has b e e n recorded (82). Culicoides imicola also has a wide distribution in northern Morocco (9). Furthermore, in central Spain, in the area of the 1987 AHS outbreak, C. imicola has a seasonal incidence; the adults disappear during autumn, not to reappear until the following April, three to four mon ths later (64). Given that the m a x i m u m duration of viraemia in horses is 18 days, this is far too long for the virus to survive in vertebrates alone. Therefore , under existing climatic conditions, AHSV wou ld not b e expec ted to b e able to overwinter in central Iberia, and indeed it d id not persist there. The virus arrived in the area in July 1987 and disappeared in October the same year, never to reappear (64) . However, further south in Spain, and also in southern Portugal and northern Morocco , the situation is different, particularly in those areas that are b o u n d e d b y the 12.5°C isotherm for the average daily m a x i m u m temperature in the coldest m o n t h of the year (January). Adult C. imicola are present and active throughout the year and as a direct result, these areas are potential enzoot ic zones for AHSV (9, 64) .

The expansion of C. imicola f rom North Africa into Spain and Portugal, a l though apparently a recent occurrence, appears to b e permanent , and Rawlings et al. have suggested that a northwards expans ion m a y still b e continuing (82) . It is not

k n o w n whether this northwards extension is the result of s o m e general climatic change or whe the r the recent series of exceptionally w a r m years is mere ly providing a preview of the poss ib le effects of such a change.

'New' vectors of African horse sickness virus Although C. imicola is the only confi rmed field vector of AHSV, during the outbreaks in Spain, isolations of this virus were also m a d e from m i x e d p o o l s consisting almost entirely of C. obsoletus and C. pulicaris, but excluding C. imicola (63) . Neither of these two species h a d previously b e e n connec ted with AHSV but given that the distribution of b o t h spec ies is m o r e northerly than the usual range of AHSV, this is not surprising, since presumably they have h a d little opportunity to b e c o m e infected. Therefore , the northerly extension in range of AHSV into southern Europe, facilitated b y the earlier expansion in the distribution of its traditional vector, C. imicola, cou ld have brought the virus into contact with new, unsuspected vectors. Had the virus not b e e n eliminated from the area b y the implementa t ion of vaccination and zoosanitary measures , this could have resulted in spread of the virus even further northwards, rather l ike passing o n a b a t o n in a relay race. T h e involvement of novel vectors is always a possibility and is to b e suspected whenever the distribution of a vector-borne pa thogen alters significantly.

T h e significance of the 1988 AHSV isolations from C. obsoletus and C. pulicaris in Spain is emphas i sed further through the implications of recent research o n a northern European midge , C. nubeculosus. This species has an oral susceptibility rate for AHSV of less than 1% w h e n the immature stages of the midge are reared at 25°C and w h e n the adults are fed u p o n a standard titre of virus. Furthermore, this rate of oral susceptibility cannot b e enhanced b y selective breeding from susceptible parents (66) . However , if the rearing temperature of the immature stages of the midge is raised to b e t w e e n 30°C and 35°C, no t only d o e s the developmenta l t ime from egg to adult decrease dramatically to p r o d u c e significantly smaller adults, but the oral susceptibility rate increases to over 10%, and virus replicates to levels suggesting that transmission is possible (67). Similar results have b e e n obta ined using C. nubeculosus and b luetongue virus, a major pa thogen of ruminants (111) . Since selective breeding h a d n o effect u p o n the C. nubeculosus oral susceptibility rate, this is clearly not a genetically controlled, heritable trait, as is usually exhibi ted b y biological vectors. Instead, it m a y b e that the increase in deve lopment temperature not only gives rise to smaller adults but m o r e 'fragile' adults with an increased incidence of the so-cal led l e a k y gut' p h e n o m e n o n (11) . In such a situation, virus is able to pass directly f rom the ingested b l o o d mea l in the gut lumen, into the h a e m o c o e l without having to first infect and replicate in the gut wall cells. Once in the h a e m o c o e l , it is wel l documen ted that mos t arboviruses will replicate and then may b e transmitted, e v e n b y insects that d o not normally act as vectors. Such a s equence of events m a y b e envisaged as a hybrid m e c h a n i s m w h e r e b y infection is initiated b y a

4 6 Rev. sci. tech. Off. int. Epiz., 19 (1)

mechanical event (passage of virus from the gut l u m e n into the h a e m o c o e l without replication), but transmission is the result of a series of subsequent biological events (infection of and replication in the salivary glands, and release of progeny virus particles into the salivary ducts). This research with C. nubeculosus suggests that under the appropriate environmental conditions (i.e. elevated temperature), species of Culicoides not normally considered to b e vectors, may b e ab le to transmit AHSV, a premise that presumably can b e extended to other insect groups and other viruses. Indeed, the isolations of AHSV from mixed poo l s of C. obsoletus and C. pulicaris in Spain during 1988 m a y b e an example of the leaky gut p h e n o m e n o n in operation, potentiated b y the unusually warm conditions prevailing in the area at the time. Increasing temperature due to climate change or even to m o r e ephemera l climatic variations will tend to increase the l ike l ihood of such 'unusual' isolations.

Effect of temperature on African horse sickness virus infection in vector species of Culicoides The infection rates and rates of virogenesis of AHSV within vector species of Culicoides are temperature dependent (109) . At elevated temperatures, infection rates tend to b e higher, rates of virogenesis faster and transmission earlier. However , individual midges survive for a shorter per iod of time. As ambient temperature is reduced, the reverse is true for e ach of these four variables. Furthermore, AHSV replication d o e s not occur b e l o w 15°C and at temperatures less than or equal to 15°C, the apparent infection rate rapidly falls to zero (109). However, w h e n midges are maintained for ex tended per iods at temperatures of 10°C to 15°C, and then transferred to temperatures within the virus permissive range (>20°C), 'latent' virus that has persisted in s o m e individuals, c o m m e n c e s replication and rapidly reaches a h igh titre.

Concurring with these studies, other researchers have demonstrated that adults of C. imicola, the major vector of AHSV, are active in the field at temperatures u p to 3°C lower than the min imum required for AHSV replication (92). In practical terms, this suggests that AHSV transmission m a y occur only over the warmer parts of the range of C. imicola. In the cooler areas of C. imicola distribution (e.g. further north or at higher altitudes), even if the virus were introduced, transmission wou ld b e imposs ib le or restricted to certain times of the year (e.g. h igh summer) or climatically favourable localities. These findings suggest that even if C. imicola were to continue expansion in range northwards, the ability to transmit AHSV will progressively decrease with increasing latitude, unless accompanied b y climate moderat ion. Nevertheless, an additional aspect must b e considered, namely: ex tended midge survival at l o w temperatures, up to 90 days in s o m e cases (12) , and replication of latent virus in s o m e of these individuals o n c e the temperature rises to permissive levels (e.g. in spring). These p h e n o m e n a could constitute a viral overwintering mechan i sm in the absence of vertebrate involvement, and also in the apparent absence of infected vectors (Table I). Although the precise 'trigger'

Table I African horse sickness virus: overwintering mechanism

No. Sequence of events

1

2

3

4

5

6

AHSV: African horse sickness virus

temperatures m a y vary, this is a general m e c h a n i s m that is

l ikely to b e applicable to a wide range of arboviruses and

vectors.

African horse sickness and El Nino/Southern oscillation in South Africa In South Africa, major epizootics of AHS occur every ten to fifteen years, al though the reason for this pattern has b e e n unclear. However , a strong l ink b e t w e e n the timing of these epizootics and the w a r m (El Niño) phase of the El Niño/Southern oscillation (ENSO) has recently b e e n discovered. Baylis et al. suggest that the association is media ted b y the combinat ion of heavy rainfall and drought brought to South Africa b y ENSO (10) . T h e l ink b e t w e e n this climatic p h e n o m e n o n and AHS is via the vector, C. imicola,

which b reeds in saturated soil. In years of heavy rain, populat ions of this insect can increase b y over 200-fold. Baylis and his co-workers s h o w e d that since 1803 , thirteen of the fourteen major epizootics of AHS in South Africa have coincided with a warm-phase ENSO. However , since 1803, forty-two ENSOs have also occurred w h i c h w e r e not m a r k e d b y epizootics of AHS. Baylis et al. explain this apparent anomaly b y showing that in South Africa, wa rm phase ENSOs typically bring heavy rain fol lowed b y drought; n o AHS epizootics occurred during these years. However , a sub-set of ENSO years exists where the pattern is reversed, i.e., p ronounced drought is fo l lowed b y heavy rain. These were the years w h e n the thirteen AHS epizootics occurred. T h e reasons w h y this pattern of drought fol lowed b y heavy rain is conducive to AHS, remain unclear, but recognition of the pattern means that predict ion of future AHS epizoot ics in South Africa will b e m a d e significantly easier. Similar l inks b e t w e e n the excessive rainfall brought to certain areas of the g lobe b y El Niño or other climatic variation, and vector-borne disease prevalence, have b e e n reported for many diseases including the zoonosis , Rift Valley fever (RVF) (112) .

The virus seems to require temperatures >= 15°C in order to replicate in, and be transmitted by, the vector Culicoides.

Adults of the AHSV vector, Culicoides imicola, are active at temperatures as low as 12°C and can survive even cooler temperatures in an inactive state.

Since the virus requires a higher minimum temperature than the vector, this suggests that replication and transmission may be possible over only part of the range of the vector.

However, during cold periods, virus can survive for extended periods at 'undetectable' levels in adult vectors, the lifespan of which is extended at these temperatures, to up to 90 days in some cases.

When temperatures rise to permissive levels (>15°C), virus replication in the vector commences and transmission becomes possible.

This situation provides an overwintering mechanism for AHSV in the absence of vertebrate involvement and also in the absence of detectable virus in the vectors.

Rev. sci. tech. Off. int. Epiz, 19 (1) 47

Rift Valley fever virus Introduction An acute and highly fatal disease of lambs , associated with heavy rains and accompan ied b y reports of illness in humans , was first recognised in the Rift Valley of Kenya at the beginning of the 20th Century, al though the cause was not identified until 1930 (16, 72 , 95) . T h e agent, Rift Valley fever virus (RVFV) is an arbovirus and is n o w k n o w n to b e a typical m e m b e r of the genus Phlebovirus of the family Bunyaviridae (75).

A wide variety of animals can b e infected b y RVFV, ranging from rodents to the h ippopo tamus . However , disease s eems to b e limited to domest ic ruminants and humans (60). Clinical signs can range f rom inapparent to peracute, depending u p o n the strain of virus and the species of hos t infected. Disease is m o s t severe in s h e e p , cattle and goats, producing h igh mortality rates in n e w - b o m animals and abortion in pregnant females. Infection in humans is usually associated with a mi ld to modera te ly severe influenza-like illness, al though severe ocular sequelae , encephalitis and haemorrhagic disease occur in a small p ropor t ion of patients, occasionally leading to death (99) .

Major outbreaks and distribution of the virus Major outbreaks of RVF, affecting s h e e p and cattle, have b e e n recorded in Kenya from 1930 to 1931 , in 1968 and f rom 1978 to 1979, with minor outbreaks at. irregular intervals during the intervening years (18, 60) . T h e disease was first recognised in southern Africa w h e n a major epizoot ic occurred in sheep-rearing areas in north-eastern South Africa, in 1950. In the affected regions, an estimated 100,000 s h e e p died and 500,000 s h e e p aborted, with smaller losses in cattle and some cases of h u m a n disease (74, 89) . Severe outbreaks also occurred in Namibia in 1955, and, following heavy rains, in South Africa from 1974 to 1976 (99). Additionally, sporadic outbreaks of RVF were s e e n in. South Africa in 1952-1953, 1955-1959, 1969-1971 , 1981 and 1999 (8, 5 1 , 53 , 55, 98 , 104, 108) . Further extensive outbreaks of the disease occurred in southern Africa, involving mainly cattle, in Zimbabwe in 1955, 1957, 1969-1970 and 1978 (15, 94, 96, 97) , in Mozambique in 1969 (51 , 52 , 80, 103), and in Zambia in 1973-1974, 1978 and 1985 (4, 32) . During the 1978 epizootic in Z imbabwe, an estimated 60 ,000 abortions occurred and almost 10,000 cattle d ied. Similar losses were thought to have occurred in the earlier epizootic from 1969 to 1970 (97). Serological ev idence of RVFV or isolations of the virus have also b e e n repor ted in at least twenty other countries of Africa, including Madagascar. However , the virus has, as yet, not b e e n r ecorded outside the African continent (60, 99) .

The furthest nor th that RVF has b e e n reported to date, is Egypt, where from 1977 to 1978, a major ep idemic occurred in the Nile delta and valley, causing an unprecedented number of h u m a n infections and deaths, in addit ion to

numerous deaths and abortions in s h e e p , cattle, goats, water buffalo and camel (99) . Estimates of the number of h u m a n cases ranged from 18,000 to over 200 ,000 , with almost 600 deaths occurring f rom encephalitis and/or haemorrhagic fever ( 5 7 , 5 8 , 6 0 , 78 , 93) . F r o m 1987 to 1988, a severe ep idemic of RVF occurred in Wes t Africa, in the Senegal River basin in southern Mauritania and northern Senegal. In Mauritania alone, an estimated 224 p e o p l e d ied and a h igh rate of abort ion in s h e e p and goats was also reported (35, 36, 39) . More recently, following rainfall of 60 to 100 times the seasonal average, an exceptionally severe outbreak of RVF in humans (and domest ic l ivestock) c o m m e n c e d in Kenya in m i d - D e c e m b e r 1997, and rapidly spread into Somalia and Tanzania. T h e number of h u m a n cases was estimated to b e in the region of 89 ,000 , w h i c h suggests that this m a y b e one of the largest outbreaks of RVF on record (5, 87) . However, the ep idemic was short-lived and b y February/March 1998, the f loods h a d r eceded , rainfall h a d returned to normal and the ou tbreak was cons idered to b e over (5). Nevertheless, the v ims has remained active e lsewhere in Africa, and h u m a n deaths and animal abortions f rom RVF were again repor ted f rom Mauritania in Sep tember 1998 (6).

Transmission and epidemiology T h e enzoot ic /epizoot ic transmission cycle of RVFV is exceedingly complex . T h e v ims has b e e n isolated in Africa from twenty-three species of mosqu i to from five genera, a p o o l of m i x e d Culicoides spp . , a Simulium sp . and a Rhipicephalus tick (60). This wide diversity of ar thropod vectors is in direct contrast to mos t other arboviruses w h i c h tend to have very limited numbers of vector species , and serves to emphas ise the c o m p l e x nature of the transmission cycle of RVFV. Nevertheless, the vector findings demonstrate that mosqu i toes are b y far the mos t important ar thropod vectors of the v ims and that different species of mosqu i to are the major vectors to l ivestock in the various areas of Africa w h e r e the v ims is enzoot ic or epizootic .

Many researchers have obse rved that epizoot ics of RVF appear to fol low per iods of excessive rain (2, 60, 99) . However , the disease d o e s not have a regular yearly or seasonal distribution. For instance, in South Africa, rainfall appears to fol low an 18 to 20 year cycle, with 9 to 10 years of above average precipitation alternating with a similar per iod of l o w rainfall (60, 99) . T h e epizootics of RVF tend to cluster during years of h igh rainfall. In northern Africa, RVF outbreaks occur in ecologically drier areas than further south and appear to b e centred near irrigated areas around the River Nile. T h e rationale is that wetter condit ions are usually those that favour the increased breeding of vector species of mosqu i to and h e n c e m o r e intense v ims transmission. An increase in numbers of mosqu i toes has b e e n no ted during times of RVF ep idemics and epizoot ics (1 , 2 , 58) . However , the central enigma in the ep idemio logy of RVF concerned the fate of the v ims during inter-epizootic per iods . T h e v ims was thought to b e enzoo t i c in indigenous forests w h e r e it circulated in mosqu i toes and u n k n o w n vertebrates, extending into

4 8 Rev. sci. tech. Off. int. Epiz., 19 (1)

l ivestock rearing areas w h e n heavy rains favoured the breed ing of the vector. However , an u n k n o w n vertebrate reservoir has never b e e n identified and it is unclear h o w virus could spread f rom such cryptic foci, to simultaneously infect l ivestock over w i d e geographic areas separated b y hundreds of ki lometres (60). A breakthrough occurred in 1985 w h e n Linthicum et al. isolated RVFV during an inter-epizootic per iod from b o t h male and female floodwater-breeding aedine mosqui toes that h a d b e e n col lected as larvae in the field and reared to adults in the laboratory (48). These workers thus demonstrated that transovarial transmission is the mos t likely mechan i sm for maintenance of RVFV during times w h e n environmental conditions are unsuitable for active transmission b e t w e e n vector and vertebrate hosts. Floodwater aedine mosqui toes lay eggs o n vegetation at the edge of an expanse of water, frequently in pans or ' dambos ' in Africa (low-lying grassy areas that are inundated after heavy rains). Since these d a m b o s often dry out rapidly, the aedine eggs are adapted to undergo an obligatory per iod of drying before being able to hatch o n rewetting (during subsequent flooding). Thus, floodwater aedine mosqui toes can 'overwinter' as eggs, and the eggs can survive for long per iods , possibly several seasons, if the d a m b o s remain dry (99).

T h e suggestion is, therefore, that during inter-epizootic periods, RVFV is maintained principally via transovarial transmission of the virus in aedine mosqui toes , possibly a ided b y venereal transmission from males to females, and with a l o w level of transmission to l ivestock. Epizootics are precipitated b y abnormally heavy rains wh ich flood d a m b o s or pans and lead to an explosive increase not only in the numbers of aedine mosqui toes but also of non-aedine mosqui toes wh ich colonise such breeding sites and then serve as epizootic vectors (48) . During such epizootics , a success ion of vector species is likely to occur subsequent to flooding of potential breeding sites (45, 4 6 , 47 ) . Once infected, domest ic l ivestock circulate very h igh levels of RVFV (up to 1 0 1 0 . 1 med ian infective dose /ml) and therefore, mechanical transmission of virus b y mosqui toes , biting midges , ph lebo tomids , s tomoxids , simuliids and other biting flies is also thought to play an important role in epizootics (17, 30 , 37, 105). Ticks have b e e n s h o w n to transmit RVFV b y bi te following inoculation of the virus, but are not thought to b e significant in the ep idemio logy of the disease (49, 54 ) . Similarly, direct transmission b e t w e e n animals has b e e n demonstrated under artificial condit ions, but is not considered to b e important in l ivestock, in contrast to humans (106, 107, 108, 113) .

A major factor contributing to the e n d of epizootics of RVF is the onset of co ld weather w h i c h suppresses vector activity. In southern Africa for example , outbreaks tend to terminate abruptly after the first frosts of winter (108) . Similarly, virus transmission may persist throughout the year in those parts of Africa that exper ience warmer winters, such as Z imbabwe (99).

Rift Valley fever in North and West Africa T h e outbreaks of RVF that occurred in North and Wes t Africa during the 1970s and 1980s, differed in several respects f rom the m o r e usual patterns of disease s e e n in southern and East Africa. In particular, the outbreaks occurred in arid countries, independent ly of rainfall and in association with mosqu i to species that b r e e d in large rivers and dams, rather than floodwater aedines. Indeed , floodwater mosqu i to spec ies are not represented at all in Egypt where the evidence suggests that the major vector was likely to have b e e n Cx. pipiens, a species that b r e e d s in pol luted water and has b e e n s h o w n to b e an efficient laboratory vector of RVFV (26, 59, 102). However, neither this species , nor any other mosqui to present in Egypt, has b e e n s h o w n to b e capable of transovarially transmitting RVFV. Thus, the survival of the virus in Egypt f rom 1977 to 1978, and in isolated foci until 1980, m a y b e due to the hibernat ion of infected adults over the winter per iods (99) .

T h e appearance of RVFV in Egypt apparently represented the extension of the virus into a n e w area, possibly f rom the Sudan, w h i c h in 1976 was the nearest source of infection (22). T h e route of entry is uncertain but it has b e e n suggested that the virus could have b e e n introduced via infected humans travelling b y air. However , the h u m a n infection rate in the Sudan is reported as having b e e n extremely l o w at that time (99) . Alternatively, the virus could have b e e n introduced through the shipment of infected s h e e p and/or cattle, via the Nile or overland from northern Sudan, or b y sea from other countries of Africa. This s e e m s to b e a strong l ike l ihood (99) , a l though a third route has b e e n p roposed . In general, flying insects including mosqui toes , tend to have rather l imited flight ranges. However , strong evidence exists to suggest that under certain a tmospher ic conditions, flying insects are capable of be ing carried o n the wind, as 'aerial plankton' , over very long distances, possibly hundreds of ki lometres (13, 29 , 69, 76 , 85) . Such condit ions were apparently present just prior to the start of the RVF ep idemic in Egypt and could have facilitated the translocation of infected mosqui toes from disease foci in northern Sudan to southern Egypt, in association with northerly movemen t s of the inter-tropical convergence z o n e (ITCZ) (90, 9 1 , 93) . T h e seasonal incidence of b luetongue virus in north-central Sudan, has similarly b e e n attributed to carriage of the vector o n the ITCZ, in this case a biting midge, C. imicola (71).

In Wes t Africa, presence of RVF h a d long b e e n repor ted without the occurrence of serious disease. However , the massive ep idemic of 1977, in Egypt, w h i c h was probably associated with an increase in mosqui to breeding sites, brought about b y deve lopments such as the Aswan Dam, p r o m p t e d an increased monitoring of RVF activity b e t w e e n 1983 and 1987 in those countries in Wes t Africa affected b y similar constructions (21 , 27) . On the basis of this w o r k , the threat of an ep idemic of RVF was predic ted in southern Mauritania and northern Senegal (21) . Consequently, the RVF

Rev. sci. tech. Off. int. Epiz., 19 (1) 4 9

epidemic that c o m m e n c e d in these countries in October 1987

was the first to have b e e n predic ted in an area w h e r e n o

similar outbreak had previously b e e n recorded (99).

The recent series of ep idemics of RVF in Senegal, Mauritania and Egypt s h o w that the virus can not only extend rapidly b e y o n d its usual range, bu t also suggest that the virus has the potential to occur outside Africa (39, 58 , 93) .

Rift Valley fever in humans Humans b e c o m e infected with RVFV either b y mosqui to bite

or through contact with virus-infected animal tissues. In

Egypt, the major mosqu i to vector is l ikely to have b e e n Cx. pipiens. This is a per i -domest ic species w h i c h readily bi tes

man, so many h u m a n cases p robab ly arose from mosqu i to

bite. However, in southern Africa, the major mosqui to vectors

of RVFV are zoophi l ic and sylvatic, so the majority of h u m a n

cases originate through contact with infected animals during

food preparation, laboratory w o r k , slaughter or necropsy.

People involved in occupat ions leading to such contacts

(livestock workers , veterinarians, abattoir workers , laboratory

workers , housewives dealing wi th fresh meat , etc.) are

therefore mos t at r isk in these areas (99) . The virus

presumably gains entry to the h u m a n hos t via abraded skin,

wounds or mucus membranes , but aerosol and intranasal

infection have also b e e n demonstra ted (99) . L o w

concentrations of RVFV have b e e n found in m i l k and other

b o d y fluids of infected l ivestock, and consequent ly s o m e

authorities be l i eve that h u m a n infections m a y somet imes result f rom the consumpt ion of raw mi lk (99) .

Conclusion T h e influence of climatic variables o n the distribution and seasonal inc idence of arboviral diseases is of fundamental importance. In particular, temperature-dependent interactions b e t w e e n vectors and the arboviruses that they transmit are highly complex , and the ou tcomes can b e difficult to predict. Furthermore, these interactions are l ikely to b e profoundly influenced b y climate change, whether or not this is i nduced b y humans . Should the pred ic ted alterations in climate over the first part of the 21st Century b e confirmed, then the changes that h a v e b e e n outlined in this paper , particularly in relation to the distribution of AHSV and its vectors, m a y serve as an indication of impending changes in the ep idemio logy of a wide range of other vector-borne diseases of m a n and animals.

Influence du climat et de l'environnement sur les arboviroses et leurs vecteurs

P.S. Mellor & C.J. Leake

Résumé

Les auteurs décrivent les facteurs climatiques susceptibles d'avoir une influence significative sur la répartition géographique, l'incidence saisonnière et la prévalence de maladies transmises par des vecteurs. En se fondant sur les variations climatiques observées et sur des prévisions, ils donnent quelques exemples des types de changements susceptibles de se produire pour plusieurs arboviroses humaines et animales majeures au plan international, notamment la fièvre de la Vallée du Rift et la peste équine.

Mots-clés

Arbovirus - Répartition - Santé publique - Température - Variables climatiques -Vecteurs - Zoonoses.

50 Rev. sci. tech. Off. int Epiz., 19 (1)

Efectos del clima y la geografía sobre las infecciones arbovirales y sus vectores

P.S. Mellor & C.J. Leake

Resumen Los autores indican los componentes del clima más susceptibles de ejercer un efecto notorio sobre la distribución geográfica, la incidencia estacional y la prevalencia de enfermedades transmitidas por vectores. A partir de las variaciones climáticas, reales o predichas, los autores ofrecen ejemplos del tipo de cambios que cabe esperar en esos parámetros, aplicándolos a varias enfermedades arbovirales animales o humanas de importancia internacional, entre ellas la fiebre del Valle del Rift y la peste equina. Palabras clave Arbovirus - Distribución - Salud pública - Temperatura - Variables climáticas - Vectores - Zoonosis.

References

1. Abdel-Wahab K.S.E., El Baz L.M., El Tayeb E.M., Ornar H., Ossman M.A. & Yasin W. (1978). - Rift Valley fever virus infections in Egypt. Pathological and virological findings in man. Trans. roy. Soc. trop. Med. Hyg., 72, 392-396.

2. Acha P.N. & Szyfres B. (1987). - Rift Valley fever. In Zoonoses and communicable diseases common to man and animals, 2nd Ed. Pan American Health Organization, Washington, DC, 449-454.

3. Adhami J. & Reiter P. (1998). - Introduction and establishment of Aedes (Stegomyia) albopictus Skuse (Diptera: Culicidae) in Albania. J . Am. Mosq. Control Assoc., 14, 340-343.

4. Akafekwa G.I. (1975). - Sanitary position and methods of control in Zambia. Bull. Off. int. Epiz., 83, 843-863.

5. Anon. (1998). - An outbreak of Rift Valley fever, eastern Africa, 1997-1998. Weekly epidemiol. Rec., 73, 105-109.

6. Anon. (1998). - Rift Valley fever. EMPRES transbound. anim. Dis. Bull., 7, 7-8.

7. Anwar M. & Qureshi M.A.A. (1971). - Control and eradication of African horse sickness in Pakistan. In Control and eradication of viral diseases in the CENTO region (M.M. Lawrence, ed.). Central Treaty Organisation, Ankara, 110-112.

8. Barnard B.J.H. & Botha M.J. (1977). - An inactivated Rift Valley fever vaccine. J. S. Afr. vet. Assoc., 48, 45-48.

9. Baylis M., Touti J., Bouyoune H., Moudni L., Taoufiq B. & El Hasnaoui H. (1998). - Studies of the mortality rate of Culicoides imicola in Morocco. Arch. Virol. Suppl., 14, 127-136.

10. Baylis M., Mellor P.S. & Meiswinkel R. (1999). - Horse sickness and ENSO in South Africa. Nature, 397, 574.

11. Boorman J. (1960). - Observations on the amount of virus present in the haemolymph of Aedes aegypti infected with Uganda S, yellow fever and Semliki Forest viruses. Trans. roy. Soc. trop. Med. Hyg., 54, 362-365.

12. Boorman J. (1990). - A review of Culicoides subgenus Avaritia species (Dipt.: Cerat.), vectors of viruses of sheep, cattle and horses, with particular reference to Culicoides imicola in Europe and the Mediterranean region. Report prepared for the Overseas Development Administration of the British Government. Natural History Museum, London, 54 pp.

13. Bowden J . & Gibbs D.G. (1973). - Light-trap and suction-trap catches of insects in the northern Gezira, Sudan, in the season of southward movement of the inter-tropical front. Bull. entomol. Res., 62, 571-596.

Rev. sci. tech. Off. int. Epiz., 19 (1) 51

14. Burke D.S. & Leake C.J. (1988). - Japanese encephalitis. In The arboviruses: epidemiology and ecology, Vol. III (T.P. Monath, ed.) . CRC Press, Boca Raton, Florida, 63-92.

15. Christie G.J. (1969). - Rift Valley fever. Rhod. Sci. News, 3,

238-240.

16. Daubney R., Hudson J.R. & Garnham P.C. (1931). -Enzootic hepatitis or Rift Valley fever. An undescribed virus disease of sheep, cattle and man from East Africa. J. Pathol. Bacteriol., 34, 545-579.

17. Davies F.G. & Highton R.B. (1980). - Possible vectors of Rift Valley fever in Kenya. Trans. roy. Soc. trop. Med. Hyg., 74 (6), 815-816.

18. Davies F.G., Koros J. & Mbugua H. (1985). - Rift Valley fever in Kenya: the presence of antibody to the virus in camels (Camelus dromedarius). J. Hyg. (Lond.), 94 (2), 241-244.

19. Diaz Montilla R. & Paños Marti P. (1967). - Epizootiologia de la peste equina en España. Bull. Off. int. Epiz., 68, 705-714.

20. Diaz Yubero M.A. (1987). - Situación de la peste equina en España. Bull. OIE, 99 (12), 47-48.

21. Digoutte J.P. & Peters C.J. (1989). - General aspects of the 1987 Rift Valley fever epidemic in Mauritania. Res. Virol., 140, 27-30.

22. Eisa M., Kheir El Sid E.D., Shomein A.M. & Meegan J.M. (1980). - An outbreak of Rift Valley fever in the Sudan -1976. Trans. roy. Soc. trop. Med. Hyg., 74, 417-419.

23. Elias S.A. (1994). - Beetles are better Palaeothermometers than bears. Smithsonian Institution Press, Washington, DC, 17 P P .

24. Elias S.A. (1995). - Insects and climate change: fossil evidence from the Rocky Mountains. BioScience, 4 1 , 552-559.

25. Erasmus B.J., Swanepoel R.J., Williams R. & Verwoerd D.W. (1990). - Human infection with African horse sickness virus (AHSV). In Proc. 8th Congress of Virology, 26-31 August, Berlin. Institute for Clinical and Experimental Virology, Berlin, Abstract P16-005, 114 pp.

26. Gargan T.P., Bailey C.L., Higbee G.A., Gad A. & El Said S. (1983). - The effect of laboratory colonization on the vector-pathogen interactions of Egyptian Culex pipiens and Rift Valley fever virus. Am. J. trop. Med. Hyg., 32, 1154-1163.

27. Guillard M., Le Guenno B., Wilson M.L., Desoutter D., Gonzarlez J.P. & Digoutte J.P. (1988). - Prévalence d'anticorps contre le virus de la fièvre de la Vallée du Rift chez les petits ruminants du Sénégal. Ann. Inst. Pasteur/Virol., 139, 455-459.

28. Hack W.H. (1955). - Estudio sobre la biología del Triatoma infestens. Anales Inst. Med. reg. (Tucumán), 4, 125-102.

29. Hayashi K., Suzuki H. & Makino Y. (1979). - Notes on the transoceanic insects captured on the East China Sea in 1976-1978. Trop. Med., 21, 1-10.

30. Hoch A.L., Gargan T.P. & Bailey C.L. (1985). - Mechanical transmission of Rift Valley fever virus by haematophagous diptera. Am. J. trop. Med. Hyg., 34, 188-193.

31. Howell P.G. (1963). - African horse sickness. In Emerging diseases of animals. Food and Agriculture Organization, Rome, 73-108.

32. Hussein N.A., Chizyuka R.Z., Ksiazek T.G., Scott R.M. & Boulos B.A.M. (1987). - Epizootic of Rift Valley fever in Zambia, 1985. Vet. Rec., 121 (5), 111.

33. Intergovernmental Panel on Climate Change (IPCC) (1996). - Climate change 1995. Impacts, adaptations and mitigation of climate change: scientific-technical analyses. Contribution of Working Group II to the Second Assessment Report of the IPCC (R.T. Watson, M.C. Zinyowera & R.H. Moss, eds). Cambridge University Press, Cambridge, 878 pp.

34. Intergovernmental Panel on Climate Change (IPCC) (1996). - Climate change 1995. The science of climate change. Contribution of Working Group I to the Second Assessment Report of the IPCC (J.T. Houghton, L.G. Meiro Filho, B.A. Callander, N. Harris, A. Kattenberg & K. Maskell, eds). Cambridge University Press, Cambridge, 572 pp.

35. Jouan A., Le Guenno B., Digoutte J.P., Phillipe B., Riou O. & Adam F. (1988). - A Rift Valley fever epidemic in southern Mauritania. Ann. Inst. Pasteur/Virol, 139, 307-308.

36. Jouan A., Coulibaly I., Adam F., Phillipe B., Riou O., Le Guenno B., Christie R., Merzoug N.O., Ksiazek T. & Digoutte J.P. (1989). - Analytical study of a Rift Valley fever epidemic. Res. Virol, 40, 175-186.

37. Jupp P.G., Mcintosh B.M. & Thompson D.L. (1984). -Mechanical transmission of Rift Valley fever virus by mosquitoes. S. Afr. J . Sci., 80, 276.

38. Knudson A.B. (1995). - Geographic spread of Aedes albopictus in Europe and the concern among public health authorities. Eur. J. Epidemiol, 11, 345-348.

39. Ksiazek T.G., Jouan A., Meegan J.M., Le Guenno B., Wilson M.L., Peters C.J., Digoutte J.P., Guillaud M., Merzoug N.O. & Touray E.M. (1989). - Rift Valley fever among domestic animals in the recent West African outbreak. Res. Virol., 140, 67-77.

40. Kukla G. & Karl T.R. (1993). - Night-time warming and the greenhouse effect. Environ. Sci. Technol., 27, 1468-1474.

41 . Leake C.J. (1984). - Transovarial transmission of arboviruses by mosquitoes. In Vectors in virus biology (M.A. Mayo & K.A. Harrap, eds). Academic Press, New York, 63-91.

42. Leake C.J. (1998). - Mosquito-borne arboviruses. In Zoonoses: biology, clinical practice, and public health control (S.R. Palmer, E.J.L. Soulsby & D.I.H. Simpson, eds). Oxford University Press, London, 401-413.

4 3 . Leake C.J. & Johnson R.T. (1987). - The pathogenesis of Japanese encephalitis virus in Culex tritaeniorhynchus mosquitoes. Trans. roy. Soc. trop. Med. Hyg., 81, 681-685.

52 Rev. sci. tech. Off. int Epiz., 19 (1)

44. Lines J. (1995). - The effects of climatic and land-use changes on insect vectors of human disease. In Insects in a changing environment (R. Harrington & N.E. Stork, eds). Academic Press, New York, 157-175.

45. Linthicum K.J., Davies F.G., Bailey C.L. & Kairo A. (1983). -Mosquito species succession in a dambo in an East African forest. Mosq. News, 43, 464-470.

46. Linthicum K.J., Davies F.G., Bailey C.L. & Kairo A. (1984). -Mosquito species encountered in a flooded grassland dambo in Kenya. Mosq. News, 44, 228-232.

47. Linthicum K.J., Davies F.G. & Kairo A. (1984). -Observations of the biting activity of mosquitoes at a flooded dambo in Kenya. Mosq. News, 44, 595-598.

48. Linthicum K.J., Davies F.G., Kairo A. & Bailey C.L. (1985). -Rift Valley fever (family Bunyaviridae, genus Phlebovirus). Isolations from Diptera collected during an inter-epizootic period in Kenya. J. Hyg. (Camb.), 95, 197-209.

49. Linthicum K.J., Logan T.M., Bailey C.L., Dohm D.J. & Moulton J.R. (1989). - Transstadial and horizontal transmission of Rift Valley fever virus in Hyalomma truncation. Am. J. trop. Med. Hyg., 41, 491-496.

50. Lubroth J. (1988). - African horse sickness and the epizootic in Spain, 1987. Equine Pract., 10, 26-33.

51. Mcintosh B.M. (1972). - Rift Valley fever. I. Vector studies in the field. J. 5. Afr. vet. Assoc., 43, 391-395.

52. Mcintosh B.M. (1975). - Mosquitoes as vectors of viruses in southern Africa. Entomol. Mémoires S. Afr. Dept. Agric. Tech. Services, 43, 1-19.

53. Mcintosh B.M. (1980). - The epidemiology of arthropod-borne viruses in southern Africa. DSc Thesis, University of Pretoria, 348 pp.

54. Mcintosh B.M., Jupp P.G., Dos Santos I. & Barnard B.J.H. (1980). - Vector studies on Rift Valley fever virus in South Africa. S. Afr. med. J., 58, 127-132.

55. Mcintosh B.M., Jupp P.G., Dos Santos I. & Rowe A.C. (1983). - Field and laboratory evidence implicating Culex zambiensis and Aedes circumluteolus as vectors of Rift Valley fever virus in coastal South Africa. S. Afr. J. Sci., 79, 61-64.

56. McMichael A.J., Haines A., Slooff R. & Kovats S. (eds) (1996). - Climate change and human health. An assessment prepared by a task group on behalf of the World Health Organization (WHO), the World Meteorological Organization and the United Nations Environment Programme. WHO, Geneva, 297 pp.

57. Meegan J.M. (1979). - The Rift Valley fever epizootic in Egypt 1977-1978. I. Description of the epizootic and virological studies. Trans. roy. Soc. trop. Med. Hyg., 73, 618-623.

58. Meegan J.M. (1981). - Rift Valley fever in Egypt: an overview of the epizootic in 1977 and 1978. Contrib. Epidemiol. Biostat., 3, 100-113.

59. Meegan J.M., Khalil G.M., Hoogstraal H. & Adham F.K. (1980). - Experimental transmission and field isolation studies implicating Culex pipiens as a vector of Rift Valley fever virus in Egypt. Am. J. trop. Med. Hyg., 29, 1405-1410.

60. Meegan J.M. & Bailey C.L. (1989). - Rift Valley fever. In The arboviruses: epidemiology and ecology, Vol. IV (T.P. Monath, ed.). CRC Press, Boca Raton, Florida, 51-76.

61. Meiswinkel R., Nevill E.M. & Venter G.J. (1994). - Vectors: Culicoides spp. In Infectious diseases of livestock with special reference to southern Africa, Vol. I. (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, 68-89.

62. Mellor P.S. (1993). - African horse sickness: transmission and epidemiology. Vet. Res., 24, 199-212.

63. Mellor P.S., Boned J., Hamblin C. & Graham S. (1990). -Isolations of African horse sickness virus from vector insects made during the 1988 epizootic in Spain. Epidemiol. Infect., 105, 447-454.

64. Mellor P.S., Capela R., Hamblin C , Hooghuis H., Mertens P.P.C., Plateau E. & Vizcaino J.M. (1994). -African horse sickness in Europe: epidemiology. Equine infect. Dis., 7, 61-63.

65. Mellor P.S. & Boorman J. (1995). - The transmission and geographical spread of African horse sickness and bluetongue viruses. Ann. trop. Med. Parasitai., 89, 1-15.

66. Mellor P.S., Rawlings P., Baylis M. & Wellby M.P. (1998). -Effect of temperature on African horse sickness virus infection in Culicoides. Arch. Virol., Suppl., 14, 115-163.

67. Mellor P.S., Baylis M., Rawlings P. & Wellby M.P. (1999). -Effect of temperature on African horse sickness virus infection of and transmission by vector species of Culicoides (Diptera: Ceratopogonidae). Equine infect. Dis., 8, 246-251.

68. Meyer R.P., Hardy J.L. & Reisen W.K. (1990). - Diel changes in adult mosquito microhabitat temperatures and their relationship to the extrinsic incubation of arboviruses in mosquitoes in Kern County, California. J. med. Entomol., 27 (4), 607-614.

69. Ming J.-G., Jin H., Riley J.R., Reynolds D.R., Smith A.D., Wang R.-L., Cheng J.-Y. & Cheng X.-N. (1993). - Autumn southward 'return' migration of the mosquito Culex tritaeniorhynchus in China. Med. vet. Entomol., 7, 323-327.

70. Mirchamsy H. & Hazrati A. (1973). - A review on the aetiology and pathology of African horse sickness. Arch. Inst Razi, 25, 23-46.

71. Mohammed M.E.H. & Mellor P.S. (1990). - Further studies on bluetongue and bluetongue-related orbiviruses in the Sudan. Epidemiol. Infect., 105, 619-632.

72. Montgomery R.E. (1912). - Nairobi sheep disease. Annual Report of the Department of Agriculture, British East Africa, 1910-1911.

73. Moore C.G. & Mitchell C.J. (1997). - Aedes albopictus in the United States: ten-year presence and public health implications. Emerg. infect Dis., 3, 329-334.

Rev. sci. tech. Off. int. Epiz., 19(1) 53

74. Mundel B. & Gear J. (1951). - Rift Valley fever. The occurrence of human cases in Johannesburg. S. Afr. med. J., 25, 797-800.

75. Murphy F.A., Fauquet C.M., Bishop D.H.L., Ghabrial S.A., Jarvis A.W., Martelli G.P., Mayo M.A. & Summers M.D. (eds) (1995). - Virus taxonomy: classification and nomenclature of viruses. Sixth Report of the International Committee on Taxonomy of Viruses. Springer-Verlag, Vienna and New York, 586 pp.

76. Murray M.D. (1987). - Akabane epizootics in New South Wales: evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. vet. J., 64, 305-308.

77. Nawrocki S.J. & Hawley W.A. (1987). - Estimation of the northern limits of distribution of Aedes albopictus in North America. J. Am. Mosq. Control Assoc., 3, 314-317.

78. Peters C.J. & Meegan J.M. (1981). - Rift Valley fever. In CRC handbook series in zoonoses (G. Beran, ed.). CRC Press, Boca Raton, Florida, 403-419.

79. Peters R.L. (1993). - Conservation of biological diversity in the face of climate change. In Global warming and biological diversity (R.L. Peters & T.E. Lovejoy, eds). Yale University Press, New Haven, 15-30.

80. Petisca J.L.N. & Serra J.J.B.L. (1971). - Anatomia patoligica de algumas doencas dos anumals domesticos, VI. Febre do Vale de Rift. Vet. Mocambicana, 4, 69-73.

81. Rafyi A. (1961). - Horse sickness (translation). Bull. Off. int. Epiz., 56, 216-250.

82. Rawlings P., Pro M.-J., Pena I., Ortega M.-D. & Capela R. (1997). - Spatial and seasonal distribution of Culicoides imicola in Iberia in relation to the transmission of African horse sickness virus. Med. vet. Entomol., 11, 49-57.

83. Reeves W.C., Hardy J., Reisen W.K. & Millby M.M. (1994). - Potential effect of global warming on mosquito-borne arboviruses. J. med. Entormol., 31, 323-332.

84. Reisen W.K., Aslamkhan M. & Basio R.B. (1976). - The effects of climatic patterns and agricultural practices on the population dynamics of Culex tritaeniorhynchus in Asia. Southeast Asian J. trop. Med. public Hlth, 7, 61-71

85. Reynolds D.R., Smith A.D., Mukhopadhyay S., Chowdhury A.K., De B.K., Nath P.S., Mondai S.K., Das B.K. & Mukhopadhyay S. (1996). - Atmospheric transport of mosquitoes in northeast India. Med. vet. Entomol., 10 (2), 185-186.

86. Rosen L. (1986). - The natural history of Japanese encephalitis virus. Annu. Rev. Microbiol., 40, 395-414.

87. Sall A.A., de A. Zanotto P.M., Vialat P., Sene O.K. & Bouloy M. (1998). - Origin of 1997-98 Rift Valley fever outbreak in East Africa. Lancet, 352 (9140), 1596-1597.

88. Schihijo A., Mifune K. & Hayashi K. (1972). - Experimental infection of Culex tritaeniorhynchus summorosus mosquitoes reared in biotron with Japanese encephalitis virus. Trop. Med., 14, 218-229.

89. Schultz K.C.A. (1951). - The pathology of Rift Valley fever or enzootic hepatitis in South Africa. J. S. Afr. vet. med. Assoc., 22, 113-120.

90. Sellers R.F. (1980). - Weather, host and vector - their interplay in the spread of insect-borne animal virus diseases. J. Hyg. (Camb.), 85, 65-102.

91. Sellers R.F., Pedgley D.E. & Tucker M.R. (1982). - Rift Valley fever, Egypt - 1977: disease spread by wind-borne insect vectors? Vet. Rec., 110, 73-77.

92. Sellers R.F. & Mellor P.S. (1993). - Temperature and the persistence of viruses in Culicoides spp. during adverse conditions. Rev. sci. tech. Off. int. Epiz., 12 (3), 733-755.

93. Shimshony A. & Barzilai R. (1983). - Rift Valley fever. Adv. vet. Sci. comp. Med., 27, 347-425.

94. Shone D.K. (1958). - Rift Valley fever in Southern Rhodesia. Central Afr. J. Med., 4, 284-286.

95. Stordy R.J. (1913). - Mortality among lambs. Annual Report of the Department of Agriculture, British East Africa, 1912-1913.

96. Swanepoel R. (1976). - Studies on the epidemiology of Rift Valley fever. J. S. Afr. vet. med. Assoc., 47, 93-94.

97. Swanepoel R. (1981). - Observations on Rift Valley fever in Zimbabwe. Contrib. Epidemiol. Biostat., 3, 83-91.

98. Swanepoel R. (1999). - Rift Valley fever. EMPRES transbound. anim. Dis. Bull., 9, 6-8.

99. Swanepoel R. & Coetzer J.A.W. (1994). - Rift Valley fever. In Infectious diseases of livestock with special reference to southern Africa, Vol II 0.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, 688-717.

100. Takahashi M. (1976). - The effects of environmental and physiological conditions of Culex tritaeniorhynchus on the pattern of transmission of Japanese encephalitis virus. J. med. Entomol., 13, 275-284.

101. Taylor M.B., Van Der Meyden C.H., Dreyer L. & Erasmus B.J. (1993). - African horse sickness virus infection in humans: experimental work in primates. In Proc. 9th International Congress of Virology, 8-13 August, Glasgow. International Congress of Virology, Glasgow, Abstract No. P55-16, 306.

102. Turell M.J., Gargan T.P. & Bailey C.L. (1984). - Replication and dissemination of Rift Valley fever virus in Culex pipiens. Am. J., trop. Med. Hyg., 33, 176-181.

103. Valadao F.G. (1969). - Nota previa sobre a ocorrencia de uma nova doenca em Mocambique - a febre do Vale de Rift. Vet. Mocambicana, 2, 13-20.

104. Van Der Linde N.T. (1953). - A recent epidemic of Rift Valley fever in the Orange Free State. J. S. Afr. vet. med. Assoc., 24, 145-148.

105. Van Velden D.J.J., Meyer J.D., Oliver J., Gear J.H.S. & Mcintosh B. (1977). - Rift Valley fever affecting humans in South Africa. S. Afr. med. J., 51, 867-871.

54 Rev. sci. tech. Off. int Epiz., 19(1)

106. Walker J.S., Remmele N., Carter R.C., Mitten J.Q., Schuh L.G., Stephen E.L. & Klein F. (1970). - The clinical aspects of Rift Valley fever virus in household pets. I: Susceptibility of the dog. J. infect. Dis., 121, 9-18.

107. Walker J.S., Stephen E.L., Remmele N.S., Carter R.C., Mitten J.Q., Schuh L G . & Klein F. (1970). - The clinical aspects of Rift Valley fever virus in household pets. II: Susceptibility in the cat. J. infect. Dis., 121, 19-24.

108. Weiss K.E. (1957). - Rift Valley fever - a review. Bull. epiz. Dis. Afr., 5, 431-458.

109. Wellby M.P., Baylis M., Rawlings P. & Mellor P.S. (1996). -Effect of temperature on the rate of virogenesis of African horse sickness virus in Culicoides (Diptera: Ceratopogonidae) and its significance in relation to the epidemiology of the disease. Med. vet. Entomol., 86, 715-720.

110. Wilke C. (1995). - When winters go frost free, it's bugs easy. The Boston Globe, 29 May, 3 pp.

111. Wittmann E.J., Baylis M. & Mellor P.S. (1998). - Higher immature rearing temperatures induce vector competence for bluetongue virus in Culicoides nubeculosus Meigen (Ceratopogonidae). In Proc. 4th International Congress on Dipterology, 6-13 September, Oxford. Oxford International, Oxford, Abstracts Vol., 248-249.

112. World Health Organization (WHO) (1998). - El Niño and its health impacts. WHO Fact Sheet No. 192. WHO, Geneva, 1-4.

113. Yedloutschnig R.J., Dardiri A.H. & Walker J.S. (1981). - Rift Valley fever infection in sheep by contact exposure. Contrib. Epidemiol. Biostat., 3, 53-59.