chapter 1 Reservoirs: environmental process, management and policy -- 1.1Introduction / chapter 1 6 UNEP policy on reservoirs -- chapter 2 Environmental indicators of healthy water resources -- 2.1The need for environmental indicators of water quality / chapter 3 Water resources development and health: the policy perspective -- 3.1Introduction / chapter 4 Water resources development: policy perspectives of the Food and Agricultural Organization in relation to food security / chapter References -- chapter 5 Context and principles of environmental and health impact assessment -- 5.1Introduction / chapter 6 Health opportunity assessment in water resource development -- 6.1Introduction / chapter 7 Health opportunities in water resources development: a course promoting intersectoral collaboration / chapter 7 5 The role of non-expert tutors, local resource personnel and the course director -- chapter 8 Ord River irrigation area: the effect of dam construction and irrigation on the incidence of Murray Valley encephalitis virus / chapter to April, rainfall is received erratically from cyclonic depressions. Several large rivers in the Kimberley and in the adjacent western area of the Northern Territory, including the Fitzroy, Ord, Victoria and Daly rivers, drain an area of 543,000 km and yield 20 per cent of the total -- chapter waterbird populations and provide additional mosquito breeding habitats, which would be conducive to increased arbovirus activity (Stanley 1972; 1975). Indeed the potential problems were expected to become more acute as the population in the area increased with the development of Kununurra township and nearby farming activities, and with increased tourism and mining opportunities. More than sixty-five arboviruses have been isolated in tropical Australia, but only a few have been implicated in human disease (Mackenzie and others 1994a). These include the flaviviruses Murray Valley encephalitis (MVE), Kunjin, Kokobera, Alfuy, Edge Hill and dengue; and the alphaviruses Ross River, Barmah Forest, and Sindbis (Mackenzie and others 1994a; 1994b). With respect to the Ord River irrigation area, the most important of these viruses is MVE, the major cause of Australian encephalitis. MVE virus has a natural biocenose between waterbirds, particularly members of the order Ciconiiformes, and mosquitoes, particularly the fresh-water breeding species, Culex annulirostris. MVE virus is a member of the Japanese encephalitis serological complex of flaviviruses, and is more closely related to Japanese encephalitis virus than are the other Australian members of the complex (Kunjin, Kokobera, Alfuy and Stratford viruses). -- chapter Little was known about MVE virus, its vertebrate hosts or its vectors before the establishment of the Ord River irrigation area. Early serological studies by Stanley and Choo (1961; 1964) on human sera collected in 1960 from Halls Creek in East Kimberley and Derby in West Kimberley had demonstrated that the virus was circulating in these areas. However, no clinical cases of encephalitis had been reported, which may have been due to the small human population in the region prior to 1960, to a lack of awareness by clinicians, to low virus carriage rates in mosquitoes, or to a combination of these factors. Similarly, no cases of encephalitis had been reported in the Northern Territory. The first clinical case of Murray Valley encephalitis (now known as Australian encephalitis) occurred in 1969 (Table 8.1), a fatal case that was acquired by a tourist south of the Ord River irrigation area (Cook and others 1970). Only limited information was available on the mosquito species prevalent in the Ord River area before 1972, although Culex annulirostris, believed to be the major vector for MVE virus from studies carried out by Doherty and colleagues in north Queensland (Doherty and others 1963), was found to be present (H. Paterson, personal communication to Stanley 1972), and was the dominant species (H. Paterson, personal communication to Stanley 1975). Thus prior to the completion of stage one of the Ord River irrigation area, serological evidence had been obtained to demonstrate that MVE virus caused subclinical human infections, but no clinical cases had been reported. Between the completion of stage one and stage two, the first clinical case of encephalitis was reported, and limited information on the mosquito fauna was obtained but without details of mosquito numbers or population dynamics. 8.3 Studies on Murray Valley encephalitis from 1972 8.3.1 Early studies, 1972-1976 A series of investigations on the ecology of MVE virus in the Ord River irrigation area and on the effect of the completion of the Ord River dam were initiated by Stanley and colleagues in 1972.

The major components comprised: regular mosquito collections obtained just before and immediately after the wet season to determine the number and proportion of each species at different sites, and for isolation of viruses; serological studies of animals and birds to investigate their roles as possible vertebrate or reservoir hosts; and serological studies of the human population, both Caucasian and Aboriginal, to determine subclinical infection rates and to assess potential risks. These studies yielded a number of important findings which have provided the basis for much of our knowledge of MVE ecology in north-western Australia. The major findings were as follows. Mosquitoes. Using live bait traps to collect mosquitoes, it appeared that there had been a significant increase in mosquito numbers since the construction of the diver- -- chapter sion dam (Stanley 1979). Most of the species collected in the bait traps were those associated with permanent and semipermanent fresh water breeding sites, and the dominant species was Culex annulirostris, which accounted for over 70 per cent of the collections (Liehne and others 1976a; Stanley 1979). Thus the major vector species for MVE virus was shown to be abundant in the Ord River irrigation area. The major mosquito breeding areas were in swampland adjacent to the diversion dam. Little breeding activity was found in the irrigation area probably due to the excessive use of insecticides applied by aerial spraying for controlling insect pests on cotton crops. However, cotton was discontinued as a crop in 1975, and an increased number of mosquitoes began to appear in 1976. Viruses. Pools of mosquitoes were processed for virus isolation by intracerebral inoc-ulation of macerated mosquito supernatants into suckling mice. A total of 195 strains of 16 arboviruses were isolated from 1075 pools, of which 29 were identified as MVE virus and 21 as Kunjin virus. The majority of the isolates were made from Culex annulirostris (153 of 195 isolations), including 28 of 29 identified as MVE. Thus the overall virus isolation rate was high (18 per cent). For MVE virus from Culex annulirostris, 3.5 per cent of pools yielded virus at an approximate rate of 1 infected mosquito per 1459 uninfected mosquitoes (Liehne and others 1976b; 1981). Serological studies of animals and birds. All the early serological investigations employed the haemagglutination-inhibition (HI) assay. Cattle sera obtained from the Ord River irrigation area exhibited a high incidence of antibody to MVE virus (80 per cent positive), but the incidence declined to 37 per cent positivity in sera collected elsewhere in the Kimberley region (Liehne and others 1976c). A very significant increase in the incidence of antibody to MVE was observed in cattle between 1972 and 1975 in the irrigation area and nearby cattle properties, with increases ranging from between 22 and 36 per cent to between 75 and 90 per cent (Stanley 1979). While the establishment of the irrigation area and the completion of the Ord River dam were undoubtedly responsible for some of this increase, it is probable that the very heavy wet season rainfall in 1973-74 also contributed. Of 335 sera collected from 31 avian species, 195 were found to have antibody to MVE virus. Although only a few species were sampled in moderate or large num-bers, it was interesting to note that the incidence of antibody was similar between waterbirds and non-waterbirds (56 and 59 per cent, respectively), and between differ-ent avian orders: Ciconiiformes (herons, egrets), 62 per cent; Anseriformes (ducks, grebes), 55 per cent; and Psittaciformes (parrots), 56 per cent, (Liehne and others 1976c).

Human serological studies. A total of 441 human sera were collected in the Ord River area, of which 293 were from Caucasians and 148 from Aboriginals. A very high incidence of MVE antibodies was observed in the Aboriginal population, with 96 per cent of adults and 77 per cent of children exhibiting antibodies. In the Caucasian pop-ulation, the incidence of MVE virus antibodies was 53 per cent in adults and 24 per -- chapter cent in children, but the length of residence in the Ord River area was an important determinant, with those who had lived in the area fewer than three years having a lower incidence (26 per cent) than those who had lived in the area for more than three years (64 per cent) (Liehne and others 1976c). Thus these early results demonstrated that the mosquito density and bird numbers had increased since the establishment of the Ord River irrigation project, particularly around the diversion dam and Lake Kununurra, that the major mosquito vector of MVE virus was the predominant species Culex annulirostris, and that MVE virus was actively circulating in the area. However, the serological results must be treated with caution as the HI test cannot differentiate clearly between MVE and Kunjin viruses, and therefore a number of seroconversions may have been due to infection with the latter. Nevertheless, the results suggested that MVE virus may have become enzootic in the Ord River irrigation area. A single case of Australian encephalitis occurred in Kununurra in 1974; this was the last case of the 1974 epidemic that affected all Australian mainland states (Table 8.1). The first cases to be reported in the Northern Territory also occurred during the 1974 epidemic. 8.3.2 Studies carried out between 1977 and 1995 The early studies between 1972 and 1976 laid the foundation for the more detailed investigations of MVE virus ecology in north-western Australia that have been undertaken over the past twenty years. These investigations became increasingly important as cases of Australian encephalitis became more frequent, particularly with respect to surveillance methodology to enable early warnings to be given of impending epidemic activity and to understand the spread and possible persistence of the virus. In addition, the apparently ideal conditions for arboviral ecology in the Ord River irrigation area have made it essential to monitor for possible incursant mosquito vector species and viruses that could potentially become established in the region. Improved methods for mosquito collection, virus isolation, and antibody detection have been introduced over the past twenty years, which have allowed a more accurate picture to emerge of the ecology of MVE virus and a more effective surveillance system to be established to provide an early warning of increased virus activity. Human cases of Australian encephalitis, surveillance for virus activity, virus isolations, factors affecting mosquito populations, and virus spread and persistence are discussed below. Human encephalitis cases Increasing numbers of Australian encephalitis cases have occurred in Western Australia and the Northern Territory since 1977 (Mackenzie and Broom 1995; Mackenzie and others 1993a; Smith and others 1993). Indeed the majority of cases reported in Australia since 1977, thirty of -- chapter 48 cases, have been from Western Australia, with a further thirteen cases from the Northern Territory. It is also interesting to note that the first confirmed case of encephalitis due to Kunjin virus occurred in Western Australia in 1978, and three additional cases have been diagnosed since, two from Western Australia in 1991 and 1995, and one in Victoria in 1984 (Table 8.1). Most of the cases of Australian encephalitis in Western Australia have occurred in areas distant from the Ord River irrigation area. Of particular significance was the spread of MVE virus from the Kimberley area south to the Pilbara and Gascoyne regions causing one case of encephalitis in 1978 and three cases in 1981. It is hypothesized that movement of virus to the Pilbara region in 1978 was due to an increase in viral activity in the West Kimberley area following heavy rainfall and flooding, and that with subsequent extensive cyclonic rainfall in the Pilbara region, viraemic waterbirds moved south down the narrow coastal strip, introducing the virus into Pilbara (Stanley 1979). It is probable that a similar mechanism may have occurred in 1981. Although there has been evidence (see next section), of MVE virus activity in the Pilbara region in recent years, there have been no further cases. Analysis of the cases of Australian encephalitis has indicated that Aboriginal infants, particularly male infants, are most at risk of fatal or severe disease (Mackenzie and others 1993a). However, tourists and visitors to the Kimberley region (and Northern Territory) have also been shown to have an increased risk of disease. Sentinel chicken surveillance Following the 1978 outbreak of Australian encephalitis, a number of sentinel chicken flocks were established in the Kimberley area. Six flocks had been established by 1981 and the number rose to twenty-four flocks in twenty-two regional centres in the Kimberley, Pilbara and Gascoyne regions by 1989 (Broom and others 1989; Mackenzie and others 1992; 1994c).

Each flock contains twelve chickens which are bled at two weekly intervals between November and June, the period of increased risk of virus transmission, and monthly at other times. The sera are then assayed for antibody to MVE and Kunjin viruses in our laboratory in Perth to provide an early warning system of increased virus activity. Initially sera were tested by HI for the presence of antibody, and positive sera were then subjected to neutralization assay to determine the identity of the infecting virus. A more rapid enzyme-linked immunosorbent assay (ELISA) was introduced in 1986 (Broom and others 1987), and more recently a competitive ELISA using specific monoclonal antibodies to identify the virus is being used (Hall and others 1992; 1995). Sentinel chicken flocks were also established in 1992 in the Northern Territory to monitor MVE activity (Aldred and others 1992). The sentinel chicken programme has clearly shown that MVE virus is enzootic in several areas of the Kimberley region, particularly in the Ord River area at Kununurra. Seroconversions in sentinel chickens occur every year during the latter half of the wet season -- chapter in Kununurra; indeed, occasional seroconversions have been recorded in every month of the year. Elsewhere in the Kimberley region, seroconversions occur in most years towards the end of the wet season at all sites monitored, but the overall frequency tends to be less than that observed in Kununurra, except when flooding is extensive and widespread. Until about 1990, most seroconversions in sentinel chickens in the Pilbara region were due to infections with Kunjin virus, but over the next three years seroconversions to MVE virus showed a significant increase in incidence, suggesting that virus movement from the Kimberley region may be occurring more often. Since 1993, however, Kunjin virus activity has once again become more prevalent in the Pilbara area. Mosquito collections Continuing studies in 1976 and 1977 in the Ord River area using bait traps showed that while Culex annulirostris continued to dominate the mosquito fauna of the area, other species such as Coquillettidia xanthogaster, Mansonia uniformis and Anopheles bancroftii increased in number following stabilization of the margins of Lake Kununurra and the Profilific growth of aquatic plant species (Wright 1981). Studies in the West Kimberley area in 1977 in the Derby area also found that Culex annulirostris was the dominant mosquito species (Wright and others 1981). A major advance in mosquito trapping in the north of Western Australia was the introduction of the EVS-CO light trap in 1978, which replaced the use of bait traps after 1979. This resulted in a ninefold increase in the number of mosquitoes being collected, and a significant increase in the species diversity, although Culex annulirostris remained the dominant species (Stanley 1979). Annual mosquito collections have continued to be undertaken in the Ord River area and at other sites in the Kimberley region since 1978, particularly at the end of the wet season although also at other times if unusual environmental conditions such as cyclones or early wet season flooding have occurred. With the stabilization of Lakes Argyle and Kununurra and of the area under irrigation, the results obtained have provided a clearer association between environmental conditions, mosquito numbers and virus activity (see below). Although the mosquito density, and thus the number collected, is always relatively high in the Ord River area, heavy wet season rainfall and flooding result in a significant increase in the mosquito density. In other areas of the Kimberley, a similar pattern has emerged but the increase in the mosquito density is often more marked than in the Ord River area, and the proportion of different mosquito species tends to vary considerably. Nevertheless, regardless of the study area, Culex annulirostris dominates after widespread heavy rainfall and flooding, but if the rainfall is more localized, other floodplain breeding species such as Aedes normanensis may dominate initially (e.g. Broom and others 1992). -- chapter Virus isolations Mosquito collections obtained during most field trips to the north-west of Western Australia have been processed for virus isolation. Until 1985, virus isolation was undertaken by intracerebral inoculation of suckling mice, but this was then replaced by cell culture using C6/36 mosquito, PSEK, BHK and Vero cells. The use of cell culture has significantly reduced the overall virus isolation rate by largely excluding arboviruses, rhabdoviruses and most bunyaviruses, but is as effective as suckling mice for the isolation of flaviviruses and alphaviruses. MVE virus has been isolated every year that significant numbers of adult mosquitoes have been processed except 1983 (Broom and others 1989; Broom and others 1992; Mackenzie and others 1994c). Isolations of MVE, Kunjin and other flaviviruses are shown in Table 8.2. There was a strong correlation between the number of virus isolates in any given year and the prevailing environmental conditions. Thus those years with a heavy, above average wet season rainfall and subsequent widespread flooding yielded large numbers of virus isolates (1981, 1991, 1993) compared with years with average or below average rainfall and with only localized flooding. Although most MVE virus isolates were obtained from Culex annulirostris mosquitoes, occasional isolates were also obtained from a variety of other species, including Culex quinquefasciatus, Culex palpalis, Aedes normanensis, Aedes pseudonormanensis, Aedes eidvoldensis, Aedes tremulus, Anopheles annulipes, Anopheles bancroftii, Anopheles amictus and Mansonia uniformis (cited in Mackenzie and others 1994b; Mackenzie and Broom 1995), although the role of these species in natural.

transmission cycles has still to be determined. Virus carriage rates in Culex annulirostris mosquitoes are shown in Table 8.3 for the Ord River area (Kununurra�Wyndham) and Balgo and Billiluna in south-east Kimberley. Very high mosquito infection rates were observed in those years with above average rainfall. Virus spread and persistence Stanley (1979) suggested that viraemic waterbirds, which are often nomadic, may generate epidemic activity of MVE in south-east Australia and in the Pilbara region. In an attempt to understand the genesis of epidemic activity better, our laboratory initiated a long-term study in the arid south-east Kimberley area at Billiluna and Balgo, two Aboriginal communities on the northern edge of the Great Sandy Desert. Occasional cases of Australian encephalitis had occurred in both communities (1978, 1981). The studies have clearly shown that MVE virus activity only occurs following very heavy, widespread rainfall both locally and in the catchment area of the nearby watercourse, Sturt Creek, which results in extensive flooding across its floodplain (Broom and others 1992). Localized flooding is insufficient to generate virus activity. Two possible explanations can be proposed to account for the reappearance of MVE virus activity when environmental conditions are suitable: either virus can be reintroduced into the area by viraemic waterbirds arriving from enzootic areas further north; or virus may -- chapter from Halls Creek in the East Kimberley region and Derby in West Kimberley in 1960 had demonstrated that subclinical infections with both MVE and Kunjin viruses had occurred in the human population (Stanley and Choo, 1961; 1964), there had been no reported cases of Australian encephalitis in Western Australia or in the Northern Territory. Unfortunately no baseline studies were undertaken on either mosquito densities or virus incidence before the completion of stage one of the irrigation project; indeed no studies were initiated until completion of stage two, the construction of the Ord River dam. While the Ord River irrigation area undoubtedly had enormous and profound effects on the ecology of the region, most of the evidence for increases in mosquito densities and waterbird populations is circumstantial. The climate in the Kimberley and adjacent areas of the Northern Territory comprises a relatively short (four month) monsoonal wet season during which heavy rainfall events occur and the major rivers extend across vast floodplains, and a very dry dry season during which most of the country becomes arid and, in the latter half, even large rivers cease to flow. Results from studies at various locations, such as Billiluna and Halls Creek, suggest that MVE virus is occasionally epizootic in many arid areas of the Kimberley. It is probable, therefore, that the area in which the Ord River irrigation area was established was similar and, consequently, that prior to the irrigation scheme being implemented, MVE was also epizootic. Since 1972, our studies in the Ord River irrigation area and elsewhere in the Kimberley region on virus isolations from mosquitoes, on serological investigations of humans, animals and sentinel chickens, and on human cases of Australian encephalitis, have clearly shown that MVE virus is now enzootic in the Ord River area and probably in other foci such as the Derby and Broome areas of the West Kimberley region. Elsewhere, in arid areas of the Kimberley and in the Pilbara, MVE virus is epizootic and virus activity is probably initiated either by virus reactivation from desiccation-resistant mosquito eggs or by introduction through viraemic vertebrate hosts. The situation in the Northern Territory is less clear as insufficient data have been accumulated. However, it is probable that MVE is enzootic in the wetlands in the north of the Northern Territory, but epizootic in the more arid areas further south extending east from the Kimberley border. Since 1978 there has been a substantial increase in the number of cases of Australian encephalitis throughout the Kimberley and Northern Territory that cannot be ascribed to either an increase in population or a heightened awareness among clinicians. Thus, although based largely on circumstantial evidence, we believe that the Ord River Irrigation Area has had a profound effect on MVE virus activity and indeed has resulted in the virus becoming enzootic in the area. We also believe that this large, stable enzootic focus has provided the source for regular epizootic incursions to other areas of the Kimberley and adjacent arid areas of the Northern Territory, and to the Pilbara, and has probably established smaller enzootic foci in the West Kimberley. As virus can persist in desiccation-resistant mosquito eggs, it is probable that most areas of the Kimberley and adjacent areas of the -- chapter Northern Territory, and possibly parts of the Pilbara, have been 'seeded with virus which could result in epizootic activity when appropriate environmental conditions occur. Our conclusions could have important health implications as the population in north-western Australia increases through intensive agriculture, mining, service industries and tourism and, in the longer term, through possible effects of climate change (Mackenzie and others 1993b; Lindsay and Mackenzie 1997). Furthermore, increased virus activity could be exacerbated as new irrigation areas are developed in the Wyndham�East Kimberley shire and the adjacent part of the Northern Territory. Finally, there is little doubt that the profound ecological changes resulting from the establishment of the Ord River irrigation area have provided ideal conditions for increased arboviral activity. These conditions are also suitable for other exotic arboviruses, such as Japanese encephalitis and chikungunya viruses, and exotic mosquito vectors, such as Aedes albopictus. Indeed an unusual strain of MVE has been isolated from the Ord River area, which was believed to have been introduced from the Indonesian archipelago (Mackenzie and others 1991). Further-more, the recent incursion of Japanese encephalitis virus into islands in the Torres Strait and Cape York, and its possible enzootic presence in the south of Papua New Guinea, provide additional cause for concern. It is therefore essential that monitoring and surveillance of mosquitoes and arboviruses is continued so that exotic virus or vector incursions can be rapidly detected. Acknowledgments We would like to thank our many colleagues who have contributed to these studies of MVE virus activity in the north-west of Western Australia. We would also like to acknowledge the support of the Health Department of Western Australia and the National Health and Medical Research Council, and the Commonwealth Department of Health.

References -- chapter 9 Safe management of the Ross River dam, Northern Australia, in relation to recreation and health -- 9.1Background / chapter The report recognized the need to minimize disturbance of fauna and flora and suggested that 'swimmer's itch, caused by avian schistosome cercariae, and mosquito-borne viruses should be investigated. Because the 26 km northern boundary, e.g. Big Bay, Antill Creek, had steeper foreshores and deeper water, it was recommended as a primary site for public access. The 7 km western boundary formed by the dam wall was seen as ideal for viewing opportunities of the lake and surrounding hills and mountains, and for water sports. Because of inaccessibility, potential management difficulties and shallowness, the 47 km southern and eastern margins did not offer significant recreational opportunities. 9.3 Tropical itch mite The stage 1 lake was surrounded with open schlerophyll woodland which afforded kangaroos and wallabies shelter during the hottest times of the day. Part of their exoparasitic fauna is the mite Eutrombicula macropus, whose offspring spend part of their life-cycle hanging off grass stems and other vegetative matter waiting to encounter a new host. Much to their misfortune, campers and bushwalkers consequently often find themselves with an itchy rash called tropical itch, often around the lines of underclothing. Prior to the filling of the stage 2 lake, the land in the zone between the stage 1 and stage 2 margins was selectively cleared. This probably diverted the macro-pods to other wooded habitat. From November 1990 to 1992, 350 litter samples were processed using Berlese funnels and 40 W incandescent bulbs to drive any inhabitants into sample bottles containing 70 per cent alcohol. No Eutrombicula macropus were collected. Thus clearing would seem to present an effective management option against this pest, as well as having the other benefits detailed below. 9.4 Mosquitoes and arboviruses 9.4.1 Mosquitoes From April 1984 to September 1985 (stage 1), the primary questions related to definition of mosquito taxa and the suitability of different methods of catching adult mosquitoes for surveillance purposes. Twenty-six taxa were collected by all night carbon dioxide supplemented light traps or by human bait collections for one hour after sunset (Barker-Hudson and others 1993; Jones and others 1991). The numerically dominant species were Culex annulirostris and Anopheles annulipes (both species groups), which are traditionally associated with temporary fresh water pools along the lake margins, often among emergent vegetation. Of considerable surprise during September 1985 was the discovery of immatures of these species, plus Aedeomyia catasticta, utilizing extensive floating mats of the aquatic weed Hydrilla verticillata which sometimes covered 37 per cent of the surface of the lake. -- chapter This will be discussed later. Two species, Mansonia uniformis and Mansonia septempunctata, which breed in association with macrophytes such as water hyacinth Eichhornia crassipes, became less common from stage 1 to 2. The saltmarsh species Aedes vigilax was also collected in reasonable numbers at all localities around the reservoir. This species is known for its wide dispersal powers and was undoubtedly blown in from the extensive intertidal wetlands on the coast. Thus on the basis of abundance, two taxa Culex annulirostris and Anopheles annulipes s.1. warranted further consideration. The former species is considered to be the major vector of arboviruses in Australia (Russell 1995), transmitting Ross River, Barmah Forest, Kunjin, Kokobera, Alfuy and Edge Hill viruses and Murray Valley encephalitis, as well as dog heartworm. Of these, Ross River is by far the most common arbovirus in coastal northern Queensland, with morbidity approximating 400 cases per 100,000 population. Thus from first principles, this arbovirus and perhaps Barmah Forest, about which little is known, would constitute the greatest hazard to recreational use. Although Anopheles annulipes has previously been implicated in malaria transmission at Sellheim during the Second World War, this species group has returned isolated positives of Ross River and Barmah Forest viruses and Murray Valley encephalitis from other parts of Australia. However, no transmission studies have been done on the population from the reservoir. Thus on the evidence to date, it could not be regarded as a major concern at the Ross River dam. Both Culex annulirostris and Anopheles annulipes were shown to have seasonal peaks of abundance during the late post-wet season (March to May), with populations building up with the onset of spring (September to October). Spatially, the trapping programme was designed to compare mosquito numbers on the foreshore of the stage 1 lake with two localities expected to be on the margins of the stage 2A lake, with two remote localities (and therefore theoretically unaffected by any water resource project activity) as negative controls. Mosquito numbers (i.e. for those species known to breed at the dam) decreased with distance away from the Ross River dam. Both light trapping and human bait collections carried out twice per month were reasonable indicators of broad seasonal trends in mosquito abundance. However, the statistical analysis indicated that occasionally the light traps could miss short periods of high biting activity (Jones and others 1991). If greater resolution was required, it was recommended that light traps could be supplemented with animal baited traps, although it is probable that this could be rectified by intensifying the light trapping regimen. Cluster analyses of dam breeding species in both 1984-85 and 1991-93 indicated that light trap catches along the northern (Big Bay, Ti-Tree Bay, Round Island) and western sides (Ross River) gave similar patterns, but the profile towards the east (Antill Creek, Toonpan, Oak Valley) was somewhat different (Barker-Hudson and others 1993; Hearnden and Kay 1995). On this basis, adult mosquito surveillance would therefore need to be based on two localities at either en.