- Email: [email protected]
Effects of altitude, distance and waves of movement on the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae)
Preventive Veterinary Medicine 65 (2004) 135–145 www.elsevier.com/locate/prevetmed
Effects of altitude, distance and waves of movement on the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) Alan L. Bishop*, Lorraine J. Spohr, Idris M. Barchia NSW Agriculture, Locked Bag 26, Gosford, NSW 2250, Australia Received 24 June 2003; received in revised form 24 May 2004; accepted 21 June 2004
Abstract The dispersal of the biting midge and arbovirus vector Culicoides brevitarsis in the Bellinger, Macleay and Hastings river valleys and up the escarpment of the great dividing range (GDR) of midnorthern coastal New South Wales, Australia, from 1995 to 2003 was studied. The midge moved up these valleys from the endemic coastal plain in at least two waves between October and May, and both waves were modelled. Dispersal time can be explained by direct distance from the coast and the altitude of the sites. Dispersal times due to distance were similar at 18.2 2.2 (S.D.) and 15.9 2.6 weeks per 100 km for first- and second-occurrences at fixed altitude. Time of the first wave was extended 0.48 0.22 weeks for every 100-m rise in altitude and the second by 1.14 0.24 weeks for every 100-m rise for a set distance. Although C. brevitarsis can move up the escarpment of the GDR (and possibly transmit virus), vector dispersal, survival and establishment at and beyond the top of the range are limited. A third model showed that previously described slower movement of C. brevitarsis up the more-southerly Hunter valley relative to movements down the coastal plain also was related to increasing altitude. # 2004 Elsevier B.V. All rights reserved. Keywords: Culicoides brevitarsis; Models; Distance; Altitude; Vector; Arboviruses
* Corresponding author. Tel.: +61 2 4348 1928; fax: +61 2 4348 1910. E-mail address: [email protected] (A.L. Bishop). 0167-5877/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.prevetmed.2004.06.011
136
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
1. Introduction There are several species of biting midge from the genus Culicoides (Diptera: Ceratopogonidae) that transmit viruses to animals in Australia (Strandfast et al., 1984). Culicoides brevitarsis Kieffer is the main species responsible for the transmission of bluetongue and Akabane viruses to livestock in Australia (Muller et al., 1982). These viruses are a serious threat to animal health and to livestock-exports. In New South Wales (NSW), C. brevitarsis and the virus populations develop and move from over-wintering foci on the mid-northern/northern coastal plain in spring–summer (Bishop et al., 1995b, 1996a). Establishment and survival of the vector outside of the endemic area depend primarily on temperature (Allingham, 1991; Bishop et al., 1996b) to the south and, additionally, moisture to the west. C. brevitarsis usually has a coastal distribution and virus transmission occurs within its dispersive limits. Virus activity is detected in sentinel cattle herds by seroconversions. This commonly is detected first 2–7 months after the first record of the vector in light traps (Bishop et al., 1995b, 1996a)—although in some years, one or both of the viruses is absent (P.D. Kirkland, personal communication). Strong epidemiological evidence suggests that the long-distance spread of many viruses is related to the wind-borne dispersal of midge vectors (Braverman, 1992; Sellers et al., 1977, 1992). C. brevitarsis movement in NSW usually depends on temperature, wind speed, wind direction and the distance to be travelled from the endemic area (Bishop et al., 2000a). The result is gradual, seasonal movements down the coastal plain and up the Hunter valley until activity is stopped as temperatures decline towards winter (Bishop et al., 1995a). Bishop et al. (2000a) demonstrated that C. brevitarsis moves at different rates in different areas and suggested that the speed of dispersal was influenced by geographical features [e.g. increasing altitude, urban areas and the escarpment of the great dividing range (GDR)] acting as physical barriers. The Hunter valley is 140 km south of the endemic area and has been considered the most likely route of both vector and virus to the central-western slopes and plains of NSW where there are high concentrations of susceptible cattle and sheep. This valley has lower altitude, fewer geographical barriers and a greater availability of host animals than any other coastal valley; but C. brevitarsis rarely reaches its western and upper end. This was explained by the distance the vector must travel from the endemic area and slower movement relative to that down the coastal plain over a 10-year study period (Bishop et al., 2000a). The reason for the slower movement was not explained. There are also several valleys that originate in the eastern escarpment of the GDR and lead directly from the endemic coastal area towards the west. Movement to northwestern, inland regions via these valleys is feasible provided the GDR could be crossed. It was suggested that movements of C. brevitarsis probably happen regularly from areas where the vector is established, but the later movements within a season are masked by C. brevitarsis that already are present (Bishop et al., 2000a). More than one movement throughout the season would increase the potential for successful migration and for virus to be relocated by the vector. Virus activity and disease incidence at sites of destination ultimately would depend on the parous stage and infective status of the migrating C. brevitarsis population; and, beyond the top of the GDR, its ability to establish on arrival at altitudes >1000 m.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
137
Our aim was to model (using distance from the endemic coastal area and altitude) the observed dispersal of C. brevitarsis up three major coastal valleys leading from the endemic mid-north coast to the western slopes and plains of NSW for a number of sites and years. Parous stage data were recorded to consider the infective potential of the moving populations.
2. Materials and methods 2.1. Sampling C. brevitarsis exhibits crepuscular behaviour, and its activity and numbers are recorded from catches in standardised light traps modified by Standfast (personal communication) from that described by Dyce et al. (1971). Most individuals are trapped in the first 2 h after sunset in NSW because activity decreases with declining night temperatures (Bishop et al., 1995c). Each trap has a 3.5 V globe and a small downwardly directed fan driven by three Dcell batteries. A photoelectric cell automatically triggers operation at sunset. The traps were suspended from trees about 2 m above the ground in areas where cattle were consistently nearby. Collections were made into plastic bottles containing 70% alcohol. These were returned to the laboratory where Culicoides spp. were separated from other insects under a binocular microscope. C. brevitarsis was identified by its wing pattern, and the numbers, sex and parous stages (Dyce and Okey, 1981) were recorded. 2.1.1. Mid-northern coastal valleys A light trap was placed at 15 sites in, and on the GDR and adjacent slopes above, the Bellinger (sites 1–4), Macleay (sites 5–9) and Hastings (sites10–15) river valleys for eight seasons from 1995 to 2003 (Fig. 1). (Three major centres that frame the study area are marked on Fig. 1 and can be used as reference points.) The valleys were located on the narrowest section of the endemic coastal plain to give greatest definition of the valley systems and to be far enough south to avoid complications from C. brevitarsis possibly dispersing from the north (Queensland). The altitudes (m) and direct distances (km) from the coast were recorded for each site. Although lower, site 15 was assigned the same altitude as site 14 because the midge must cross the higher altitude to reach the morewesterly site 15 (Table 1). The sites were sampled over a 29-week period when C. brevitarsis was active (the second week in October to the second week in May) in each year. Catches were made over two nights, three times per month with the week of the full moon excluded because of competition with the light trap. Not all sites were sampled in all years (Table 1) or in all weeks due to bad weather and occasional trap failure. 2.1.2. Hunter valley Seven sites (16–22) were sampled from 1993 to 2003. These were used to test whether the slower movements previously recorded in the Hunter valley (Bishop et al., 2000a) were related to altitude. Direct distance (km) and altitude (m) were recorded from site 16 (Fig. 1).
138
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Fig. 1. Numbered locations of light traps for C. brevitarsis in three coastal valleys (1995–2003) and in the Hunter valley (1993–2003) in relation to the approximate midline (heavy line) of the 1000-m section of the great dividing range of NSW, Australia (* = major centre for reference).
2.2. Compilation of data Data were compiled for the following purposes. 2.2.1. Waves of movement Data were expressed initially as the percentage of times that C. brevitarsis was observed and established at each site in the Bellinger, Macleay and Hastings river valleys for a first or second time over 8 years (Table 1). Site information is also in Table 1. 2.2.2. Mid-northern coastal valleys Data used for the models were compiled based on the presence or absence of the vector at each site rather than on density. Factors including temperature, wind speed, distance from hosts and moon phase can affect comparisons of midge density from light traps at
Site no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
No. of years of sampling
8 8 8 3 3 8 4 7 6 3 8 3 7 5 8
Distance (km)
3 14 36 51 10 17 40 100 145 4 23 58 93 143 203
Altitude (m)
5 105 745 1040 7 10 25 971 887 7 10 155 981 1061 1061
First-occurrence
Second-occurrence
Occurred (%)
Established (%)
Weeks to occurrence
Occurred (%)
Established (%)
Weeks to occurrence
100 100 100 67 100 100 100 100 17 100 100 100 100 0 0
100 100 88 0 100 100 100 29 0 100 100 67 14 0 0
2 1 7 18 3 2 3 11 27 6 1 4 17 – –
0 0 12 0 0 0 0 71 0 0 0 33 43 0 0
0 0 100 0 0 0 0 60 0 0 0 100 0 0 0
– – 10 – – – – 18 – – – 15 23 – –
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Table 1 Site information, percentage estimates of occurrence and of establishment following occurrence of C. brevitarsis for a first and second time; and the times of occurrences relative to the number of weeks after sampling commenced in and above the Bellinger (Sites 1–4), Macleay (Sites 5–9) and Hastings (Sites 10–15) river valleys of NSW, Australia, from 1995 to 2003
139
140
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
different locations and at different times (Bishop et al., 2000b; Murray, 1987a). The uncontrolled effect of these factors made comparisons of density difficult and unsuitable for modelling purposes. The endemic area was a relatively narrow band of the mid-northern coastal plain. A zone 10- to 20-km inland was used as a starting point for the coastal valleys. Unfavourable winds close to the coast in the evening rather than vector absence were responsible for delays in first-occurrence at sites 1, 5 and 10 which were <10 km from the coast (Table 1); making them unsuitable as starting points. These sites subsequently were omitted from the analyses. The valleys were oriented approximately east-to-west, and all movements were assumed to be in a westerly direction (with the prevailing winds). Three data sets were compiled. The first was based on the time that C. brevitarsis was present at a site regardless of any future event. The second was based on the times when C. brevitarsis was present continuously after its first record and when it occurred for a second time at a site in the same season. A second presence was defined as happening when C. brevitarsis was found at a site at least three zero-midge sampling weeks after the first. A 3-week period of zero-midge catches represented the generation time for the species and was long enough to assume that C. brevitarsis was not established during that time. The mean (S.D.) of these periods was 5.0 2.2 weeks for the 8 years and were recorded on ten occasions. A second presence was treated as a new event and was considered the result of a new wave of movement. A second presence was either followed by no further occurrences, by establishment (i.e. continuous presence after a first or a second-occurrence) or by a third presence (one occasion only at site 8). 2.2.3. Hunter valley The Hunter valley data were compiled as a separate and third set and were firstoccurrence data only. 2.2.4. Parous composition Parous stages present the first time C. brevitarsis was recorded outside of the endemic area were used to assess the potential of the moving population to transmit virus at the new site, but were not part of the modelling procedure. For example, unfed midges (nulliparous stage) are incapable of carrying virus (Allingham, 1990). Gravid or post-parous midges have possibly fed on infected blood but may require time to become infective (Bellis et al., 1994). A Chi-square test was used to assess the relation between year and parous stage, and the results were given as total counts for the 8 years. 2.3. Statistical modelling First and second-occurrence models were developed from their respective data sets for the Bellinger, Macleay and Hastings valleys. Hunter valley data were analysed separately. The methodology used the same principles as those used by Bishop et al. (2000a) except that it was based on weeks instead of days. (Years were considered independent because of differences in climates, over-wintering in endemic areas, etc.) The dispersal time (in weeks
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
141
and denoted by t) of C. brevitarsis at a site was the number of weeks elapsing between the first presence of C. brevitarsis at any site and the first or second presence at the other sites for each year and valley. Presence was the actual week C. brevitarsis was recorded or, because some observations of C. brevitarsis were missed, presence was represented by an interval variable C which contained the time when C. brevitarsis had first been observed (U) and the previous observation time (L) i.e. C = (L, U). When C. brevitarsis was caught within a week interval then t = U (exact time). When the catches were made more than a week apart then L < t < U. When C. brevitarsis failed to appear then t > L. The objective of the analysis was to model the dispersal data (t) as a linear function of site direct distance from the coast (D) and altitude (A). For the Hunter valley data, D = loge direct distance from site 16. The following model was proposed for each of the three data sets: t ¼ b0 þ b1 D þ b2 A þ e where b0, b1 and b2 are the coefficients of regression and e is the normally distributed error with mean zero and variance s2. The coefficients b1 and b2 were tested for their significance at a = 5% (two-sided likelihood-ratio x2). A generalised failure time analysis (Turnbull, 1974, 1976) was used to fit the model and run using S-PLUS (MathSoft, 1998). The predicted time for C. brevitarsis to be present can be calculated using the standard normal probability with the following form for each of the three data sets. z¼
t ½b0 þ b1 D þ b2 A s
3. Results 3.1. Waves of movement C. brevitarsis was found early and established quickly at sites in the endemic area (Table 1). Apparent delays were recorded at the sites closest to the coast despite suitable temperatures and known vector activity. The frequency at which C. brevitarsis was detected and established decreased and the time taken to be detected increased the further sites were from the coast. C. brevitarsis reached but failed to establish at two sites and failed to reach another two sites. Two waves of movement were observed in each year of the study and these were modelled. 3.2. Mid-northern coastal valleys 3.2.1. First-occurrence model The time that C. brevitarsis was recorded at the observation sites was related to the distance from the coast and the altitude of the site (Table 2). This suggested that it had moved out of the endemic area and moved up each valley. Dispersal time (S.E.) of C. brevitarsis was 18.2 2.2 weeks per 100 km at fixed altitude. The time taken for C. brevitarsis to travel a set distance was extended by 0.48 (0.22) weeks for every 100-m increase in altitude.
142
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Table 2 Effects of distance and altitude on the dispersal times of C. brevitarsis in three coastal valleys (1995–2003) and in the Hunter valley (1993–2003) of New South Wales, Australia Source of variation
Regression coefficient
d.f.
LR (x2)
P (x2)
b
S.E.
First-occurrence model Intercept Distance (km) Altitude (m) Residual
3.28 0.18 0.005 –
0.98 0.022 0.002 –
1 1 1 51
– 142.86 4.52 4.77 (variance)
– <0.001 0.03
Second-occurrence model Intercept Distance (km) Altitude (m) Residual
3.27 0.16 0.011 –
1.11 0.026 0.002 –
1 1 1 42
– 123.62 16.58 5.28 (variance)
– <0.001 <0.001
Hunter valley model Intercept Loge distance (km) Altitude (m) Residual
0.65 2.11 0.037 –
1.59 0.53 0.007 –
1 1 1 14
– 45.51 31.30 5.41 (variance)
– <0.001 0.001
3.2.2. Second-occurrence model The time taken for C. brevitarsis to be recorded a second time at observation sites also was related to the distance from the coast and the altitude of the site (Table 2). This suggested that C. brevitarsis had moved out of the endemic area and moved up each valley a second time. Dispersal time of C. brevitarsis was 15.9 2.6 weeks per 100 km at fixed altitude. The time taken for C. brevitarsis to travel a set distance was extended by 1.14 (0.2) weeks for every 100-m increase in altitude. 3.3. Hunter valley model The time that C. brevitarsis was observed at the observation sites was related to the loge distance from site 16 and the altitude of the sites (Table 2). The time taken for C. brevitarsis to travel a set distance was extended by 3.7 (0.7) weeks for every 100-m increase in altitude. 3.4. Parous composition The total numbers of each parous stage recorded at the times C. brevitarsis occurred outside of the endemic area for the first and second time over the 8 years were nulliparous (40), gravid (66), post-parous (36) and males (4). There was a relation between year and parous stage (x2 = 25.4, d.f. = 10, P = 0.013) with greatest difference due to numbers of gravid females. About 71.8% of the total number of individuals was capable of carrying viruses.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
143
4. Discussion Knowing areas of Australia where C. brevitarsis is permanently or seasonally absent is necessary for defining the epidemiology of bluetongue and Akabane viruses and is critical for establishing protocols for the Australian livestock-export industry. C. brevitarsis normally is limited to the coastal plains of NSW (Bishop et al., 1995a, 2000a). Occasional incursions into western regions of NSW are a serious threat to the maintenance of a certifiable virus-free area. The model of Murray and Nix (1987) predicts incursions into western and north-western NSW from the north. However, their model depends on the presence of over-wintering foci from which both vector and virus develop and their predictions would be weakened or negated by the failure of C. brevitarsis to survive west of the GDR in NSW in most years. Incursions from vector-endemic regions in northern coastal NSW are feasible and are examined here. Our models indicate that the effect of distance on dispersal is significant (as shown previously by Bishop et al., 2000a) and relatively constant. Altitude slows movement further (i.e. it extends first and second presence by about 5 and 11 weeks, respectively, for the GDR’s 1000 m altitude). The GDR therefore helps prevent C. brevitarsis from reaching more-westerly sites (e.g. sites 14 and 15) or delays movements until activity could be stopped by declining temperatures late in the season (e.g. site 9). In previous studies of C. brevitarsis (Bishop et al., 1995b, 1996a), possible waves of movement were masked by individuals already present after earlier establishment. In our study, C. brevitarsis did not always establish the first time that it was recorded outside of the endemic area. Presence often was detected a second time at intervals indicating another wave of movement, although the frequency and nature of movements might not be represented fully. More than one movement would best explain the progressive spread of viruses from endemic foci in the study area long after the vector has moved initially (Barchia et al., 2002; Bishop et al., 1995b; P.D. Kirkland, personal communication). Any impact on livestock from C. brevitarsis movements to the west would depend on the vector’s ability to transmit virus on arrival. In our study, all parous stages potentially were represented when C. brevitarsis was recorded for a first or second time outside of the endemic area and the potential was higher for individuals capable of being infected or infective. However, midges moving early in the season are unlikely to be carrying virus and the potential for infection increases as the season progresses (Barchia et al., 2002). This would be facilitated by later movements of the vector. Additionally, C. brevitarsis must survive long enough at the top of the GDR for infected individuals to become infective or for infective individuals to locate a susceptible host. Survival appeared limited on and above the escarpment of the GDR, because in only 42% of occasions when C. brevitarsis was present outside of the endemic area over 8 years, did it survive long enough to become established. Typical movement by C. brevitarsis up the Hunter valley is slower than that down the coastal plain (Bishop et al., 2000a,b) and this is due to its distance from the endemic area and the effects of increasing altitude. The assumption that the Hunter valley is a common route for C. brevitarsis to the central-west was unfounded except under extremely favourable or infrequent and chance weather conditions (Murray, 1987b; Murray and Kirkland, 1995).
144
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
5. Conclusions Based on 8 years of data, we concluded that the altitude barrier provided by the GDR in NSW significantly delays (by 5–11 weeks) movements inland from the endemic coastal region in addition to previously demonstrated delays due to the distance to be travelled. These movements can occur on more than one occasion. Extended incursions onto the slopes and plains of NW NSW from the NSW coast would be rare.
Acknowledgements We thank Mr H. McKenzie for his assistance and technical expertise and Drs R. Van de Ven and P.D. Kirkland for their comments on the manuscript. The significant contribution of the many cooperators that operated light traps is also acknowledged. Funding was partially received from the National Arbovirus Monitoring Program.
References Allingham, P.G., 1990. An investigation of transovarial transmission of Akabane virus in Culicoides brevitarsis. Aust. Vet. J. 67, 273–274. Allingham, P.G., 1991. Effect of temperature on late immature stages of Culicoides brevitarsis (Diptera: Ceratopogonidae). J. Med. Entomol. 28, 878–881. Barchia, Idris M., Spohr, Lorraine J., Bishop, Alan L., 2002. Relationships between Culicoides brevitarsis parous stage and arbovirus transmission in New South Wales. Final Report to the Australian Quarantine Inspection Services, July 2002. NSW Agriculture, Gosford, Australia. Bellis, G.A., Gibson, D.S., Polkingham, I.G., Johnson, S.J., Flanagan, M., 1994. Infection of Culicoides brevitarsis and C. wadai (Diptera: Ceratopogonidae) with four Australian serotypes of bluetongue virus. J. Med. Entomol. 31, 382–387. Bishop, A.L., Barchia, I.M., Harris, A.M., 1995a. Last occurrence and survival during winter of the arbovirus vector Culicoides brevitarsis at the southern limits of its distribution. Aust. Vet. J. 72, 53–55. Bishop, A.L., Kirkland, P.D., McKenzie, H.J., Spohr, L.J., Barchia, I.M., Muller, M.J., 1995b. Distribution and seasonal movements of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) at the southern limits and its distribution in New South Wales and their correlation with arboviruses affecting livestock. J. Aust. Ent. Soc. 34, 289–298. Bishop, A.L., McKenzie, H.J., Spohr, L.J., Barchia, I.M., 1995c. Daily activity of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in the Hunter valley, NSW. Gen. Appl. Ent. 26, 31–39. Bishop, A.L., Kirkland, P.D., McKenzie, H.J., Barchia, I.M., 1996a. The dispersal of Culicoides brevitarsis in eastern New South Wales and associations with the occurrences of arbovirus infections in cattle. Aust. Vet. J. 73, 174–178. Bishop, A.L., McKenzie, H.J., Barchia, I.M., Harris, A.M., 1996b. Effect of temperature regimes on the development, survival and emergence of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in bovine dung. Aust. J. Entomol. 35, 361–368. Bishop, A.L., Barchia, I.M., Spohr, L.J., 2000a. Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med. 47, 243–254. Bishop, A.L., McKenzie, H.J., Barchia, I.M., Spohr, L.J., 2000b. Moon phase and other factors affecting light-trap catches of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Aust. J. Entomol. 39, 29–32. Braverman, Y., 1992. The possible introduction to Israel of Culicoides (Diptera Ceratopogonidae) borne animal diseases by wind. In: Walton, T.E., Osburn, B.I. (Eds.), Proceedings of the Second International Symposium on Bluetongue, African Horse Sickness, and Related Orbiviruses, Boca Raton, Florida.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
145
Dyce, A.L., Okey, D.H., 1981. Determination of parity in pinned specimens in the biting midge Culicoides variipennis. Mosq. News 41, 797–798. Dyce, A.L., Standfast, H.A., Kay, B.H., 1971. Collection and preparation of biting midges (Fam. Ceratopogonidae) and other small Diptera for virus isolation. J. Aust. Ent. Soc. 11, 91–96. MathSoft, 1998. S-PLUS 5 Guide to Statistics. MathSoft, Inc. Seattle, USA. 1014 pp. Muller, M.J., Standfast, H.A., St. George, T.D., Cybinski, D.H., 1982. Culicoides brevitarsis (Diptera: Ceratopogonidae) as a vector of arboviruses in Australia. In: St. George, T.D., Kay, B.H. (Eds.), Proceedings of the 3rd Symposium on Arbovirus Research in Australia, Brisbane, pp. 43–49. Murray, M.D., 1987a. Local dispersal of the biting midge, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in south-eastern Australia. Aust. J. Zool. 35, 559–573. Murray, M.D., 1987b. Akabane epizootics in New South Wales: evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J. 64, 305–308. Murray, M.D., Kirkland, P.D., 1995. Bluetongue and Douglas virus activity in New South Wales in 1989: further evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J. 72, 56–57. Murray, M.D., Nix, H.A., 1987. Southern limits of distribution and abundance of the biting-midge Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in south-eastern Australia: an application of the GROWEST Model. Aust. J. Zool. 35, 575–585. Sellers, R.F., Pedgley, D.E., Tucker, M.R., 1977. Possible spread of African horse sickness on the wind. J. Hyg. 79, 279–298. Sellers, R.F., Walton, T.E., Osburn, B.I., 1992. Weather Culicoides and the distribution and spread of bluetongue and African horse sickness viruses. In: Walton, T.E., Osburn, B.I. (Eds.), Proceedings of the Second International Symposium on Bluetongue, African Horse Sickness, and Related Orbiviruses, CRC Press, Boca Raton, USA, pp. 284–290. Strandfast, H.A., Dyce, A.L., St. George, T.D., Muller, M.J., Doherty, R.L., Carley, J.G., Filippich, C., 1984. Isolation of arboviruses from insects collected at Beatrice Hill, Northern Territory of Australia, 1974–1976. Aust. J. Biol. Sci. 37, 351–366. Turnbull, B.W., 1974. Nonparametric estimation of a survivorship function with doubly censored data. J. Am. Stat. Assoc. 69, 169–173. Turnbull, B.W., 1976. The empirical distribution function with arbitrarily grouped, censored and truncated data. J. R. Stat. Soc. B. 38, 290–295.
Effects of altitude, distance and waves of movement on the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) Alan L. Bishop*, Lorraine J. Spohr, Idris M. Barchia NSW Agriculture, Locked Bag 26, Gosford, NSW 2250, Australia Received 24 June 2003; received in revised form 24 May 2004; accepted 21 June 2004
Abstract The dispersal of the biting midge and arbovirus vector Culicoides brevitarsis in the Bellinger, Macleay and Hastings river valleys and up the escarpment of the great dividing range (GDR) of midnorthern coastal New South Wales, Australia, from 1995 to 2003 was studied. The midge moved up these valleys from the endemic coastal plain in at least two waves between October and May, and both waves were modelled. Dispersal time can be explained by direct distance from the coast and the altitude of the sites. Dispersal times due to distance were similar at 18.2 2.2 (S.D.) and 15.9 2.6 weeks per 100 km for first- and second-occurrences at fixed altitude. Time of the first wave was extended 0.48 0.22 weeks for every 100-m rise in altitude and the second by 1.14 0.24 weeks for every 100-m rise for a set distance. Although C. brevitarsis can move up the escarpment of the GDR (and possibly transmit virus), vector dispersal, survival and establishment at and beyond the top of the range are limited. A third model showed that previously described slower movement of C. brevitarsis up the more-southerly Hunter valley relative to movements down the coastal plain also was related to increasing altitude. # 2004 Elsevier B.V. All rights reserved. Keywords: Culicoides brevitarsis; Models; Distance; Altitude; Vector; Arboviruses
* Corresponding author. Tel.: +61 2 4348 1928; fax: +61 2 4348 1910. E-mail address: [email protected] (A.L. Bishop). 0167-5877/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.prevetmed.2004.06.011
136
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
1. Introduction There are several species of biting midge from the genus Culicoides (Diptera: Ceratopogonidae) that transmit viruses to animals in Australia (Strandfast et al., 1984). Culicoides brevitarsis Kieffer is the main species responsible for the transmission of bluetongue and Akabane viruses to livestock in Australia (Muller et al., 1982). These viruses are a serious threat to animal health and to livestock-exports. In New South Wales (NSW), C. brevitarsis and the virus populations develop and move from over-wintering foci on the mid-northern/northern coastal plain in spring–summer (Bishop et al., 1995b, 1996a). Establishment and survival of the vector outside of the endemic area depend primarily on temperature (Allingham, 1991; Bishop et al., 1996b) to the south and, additionally, moisture to the west. C. brevitarsis usually has a coastal distribution and virus transmission occurs within its dispersive limits. Virus activity is detected in sentinel cattle herds by seroconversions. This commonly is detected first 2–7 months after the first record of the vector in light traps (Bishop et al., 1995b, 1996a)—although in some years, one or both of the viruses is absent (P.D. Kirkland, personal communication). Strong epidemiological evidence suggests that the long-distance spread of many viruses is related to the wind-borne dispersal of midge vectors (Braverman, 1992; Sellers et al., 1977, 1992). C. brevitarsis movement in NSW usually depends on temperature, wind speed, wind direction and the distance to be travelled from the endemic area (Bishop et al., 2000a). The result is gradual, seasonal movements down the coastal plain and up the Hunter valley until activity is stopped as temperatures decline towards winter (Bishop et al., 1995a). Bishop et al. (2000a) demonstrated that C. brevitarsis moves at different rates in different areas and suggested that the speed of dispersal was influenced by geographical features [e.g. increasing altitude, urban areas and the escarpment of the great dividing range (GDR)] acting as physical barriers. The Hunter valley is 140 km south of the endemic area and has been considered the most likely route of both vector and virus to the central-western slopes and plains of NSW where there are high concentrations of susceptible cattle and sheep. This valley has lower altitude, fewer geographical barriers and a greater availability of host animals than any other coastal valley; but C. brevitarsis rarely reaches its western and upper end. This was explained by the distance the vector must travel from the endemic area and slower movement relative to that down the coastal plain over a 10-year study period (Bishop et al., 2000a). The reason for the slower movement was not explained. There are also several valleys that originate in the eastern escarpment of the GDR and lead directly from the endemic coastal area towards the west. Movement to northwestern, inland regions via these valleys is feasible provided the GDR could be crossed. It was suggested that movements of C. brevitarsis probably happen regularly from areas where the vector is established, but the later movements within a season are masked by C. brevitarsis that already are present (Bishop et al., 2000a). More than one movement throughout the season would increase the potential for successful migration and for virus to be relocated by the vector. Virus activity and disease incidence at sites of destination ultimately would depend on the parous stage and infective status of the migrating C. brevitarsis population; and, beyond the top of the GDR, its ability to establish on arrival at altitudes >1000 m.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
137
Our aim was to model (using distance from the endemic coastal area and altitude) the observed dispersal of C. brevitarsis up three major coastal valleys leading from the endemic mid-north coast to the western slopes and plains of NSW for a number of sites and years. Parous stage data were recorded to consider the infective potential of the moving populations.
2. Materials and methods 2.1. Sampling C. brevitarsis exhibits crepuscular behaviour, and its activity and numbers are recorded from catches in standardised light traps modified by Standfast (personal communication) from that described by Dyce et al. (1971). Most individuals are trapped in the first 2 h after sunset in NSW because activity decreases with declining night temperatures (Bishop et al., 1995c). Each trap has a 3.5 V globe and a small downwardly directed fan driven by three Dcell batteries. A photoelectric cell automatically triggers operation at sunset. The traps were suspended from trees about 2 m above the ground in areas where cattle were consistently nearby. Collections were made into plastic bottles containing 70% alcohol. These were returned to the laboratory where Culicoides spp. were separated from other insects under a binocular microscope. C. brevitarsis was identified by its wing pattern, and the numbers, sex and parous stages (Dyce and Okey, 1981) were recorded. 2.1.1. Mid-northern coastal valleys A light trap was placed at 15 sites in, and on the GDR and adjacent slopes above, the Bellinger (sites 1–4), Macleay (sites 5–9) and Hastings (sites10–15) river valleys for eight seasons from 1995 to 2003 (Fig. 1). (Three major centres that frame the study area are marked on Fig. 1 and can be used as reference points.) The valleys were located on the narrowest section of the endemic coastal plain to give greatest definition of the valley systems and to be far enough south to avoid complications from C. brevitarsis possibly dispersing from the north (Queensland). The altitudes (m) and direct distances (km) from the coast were recorded for each site. Although lower, site 15 was assigned the same altitude as site 14 because the midge must cross the higher altitude to reach the morewesterly site 15 (Table 1). The sites were sampled over a 29-week period when C. brevitarsis was active (the second week in October to the second week in May) in each year. Catches were made over two nights, three times per month with the week of the full moon excluded because of competition with the light trap. Not all sites were sampled in all years (Table 1) or in all weeks due to bad weather and occasional trap failure. 2.1.2. Hunter valley Seven sites (16–22) were sampled from 1993 to 2003. These were used to test whether the slower movements previously recorded in the Hunter valley (Bishop et al., 2000a) were related to altitude. Direct distance (km) and altitude (m) were recorded from site 16 (Fig. 1).
138
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Fig. 1. Numbered locations of light traps for C. brevitarsis in three coastal valleys (1995–2003) and in the Hunter valley (1993–2003) in relation to the approximate midline (heavy line) of the 1000-m section of the great dividing range of NSW, Australia (* = major centre for reference).
2.2. Compilation of data Data were compiled for the following purposes. 2.2.1. Waves of movement Data were expressed initially as the percentage of times that C. brevitarsis was observed and established at each site in the Bellinger, Macleay and Hastings river valleys for a first or second time over 8 years (Table 1). Site information is also in Table 1. 2.2.2. Mid-northern coastal valleys Data used for the models were compiled based on the presence or absence of the vector at each site rather than on density. Factors including temperature, wind speed, distance from hosts and moon phase can affect comparisons of midge density from light traps at
Site no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
No. of years of sampling
8 8 8 3 3 8 4 7 6 3 8 3 7 5 8
Distance (km)
3 14 36 51 10 17 40 100 145 4 23 58 93 143 203
Altitude (m)
5 105 745 1040 7 10 25 971 887 7 10 155 981 1061 1061
First-occurrence
Second-occurrence
Occurred (%)
Established (%)
Weeks to occurrence
Occurred (%)
Established (%)
Weeks to occurrence
100 100 100 67 100 100 100 100 17 100 100 100 100 0 0
100 100 88 0 100 100 100 29 0 100 100 67 14 0 0
2 1 7 18 3 2 3 11 27 6 1 4 17 – –
0 0 12 0 0 0 0 71 0 0 0 33 43 0 0
0 0 100 0 0 0 0 60 0 0 0 100 0 0 0
– – 10 – – – – 18 – – – 15 23 – –
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Table 1 Site information, percentage estimates of occurrence and of establishment following occurrence of C. brevitarsis for a first and second time; and the times of occurrences relative to the number of weeks after sampling commenced in and above the Bellinger (Sites 1–4), Macleay (Sites 5–9) and Hastings (Sites 10–15) river valleys of NSW, Australia, from 1995 to 2003
139
140
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
different locations and at different times (Bishop et al., 2000b; Murray, 1987a). The uncontrolled effect of these factors made comparisons of density difficult and unsuitable for modelling purposes. The endemic area was a relatively narrow band of the mid-northern coastal plain. A zone 10- to 20-km inland was used as a starting point for the coastal valleys. Unfavourable winds close to the coast in the evening rather than vector absence were responsible for delays in first-occurrence at sites 1, 5 and 10 which were <10 km from the coast (Table 1); making them unsuitable as starting points. These sites subsequently were omitted from the analyses. The valleys were oriented approximately east-to-west, and all movements were assumed to be in a westerly direction (with the prevailing winds). Three data sets were compiled. The first was based on the time that C. brevitarsis was present at a site regardless of any future event. The second was based on the times when C. brevitarsis was present continuously after its first record and when it occurred for a second time at a site in the same season. A second presence was defined as happening when C. brevitarsis was found at a site at least three zero-midge sampling weeks after the first. A 3-week period of zero-midge catches represented the generation time for the species and was long enough to assume that C. brevitarsis was not established during that time. The mean (S.D.) of these periods was 5.0 2.2 weeks for the 8 years and were recorded on ten occasions. A second presence was treated as a new event and was considered the result of a new wave of movement. A second presence was either followed by no further occurrences, by establishment (i.e. continuous presence after a first or a second-occurrence) or by a third presence (one occasion only at site 8). 2.2.3. Hunter valley The Hunter valley data were compiled as a separate and third set and were firstoccurrence data only. 2.2.4. Parous composition Parous stages present the first time C. brevitarsis was recorded outside of the endemic area were used to assess the potential of the moving population to transmit virus at the new site, but were not part of the modelling procedure. For example, unfed midges (nulliparous stage) are incapable of carrying virus (Allingham, 1990). Gravid or post-parous midges have possibly fed on infected blood but may require time to become infective (Bellis et al., 1994). A Chi-square test was used to assess the relation between year and parous stage, and the results were given as total counts for the 8 years. 2.3. Statistical modelling First and second-occurrence models were developed from their respective data sets for the Bellinger, Macleay and Hastings valleys. Hunter valley data were analysed separately. The methodology used the same principles as those used by Bishop et al. (2000a) except that it was based on weeks instead of days. (Years were considered independent because of differences in climates, over-wintering in endemic areas, etc.) The dispersal time (in weeks
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
141
and denoted by t) of C. brevitarsis at a site was the number of weeks elapsing between the first presence of C. brevitarsis at any site and the first or second presence at the other sites for each year and valley. Presence was the actual week C. brevitarsis was recorded or, because some observations of C. brevitarsis were missed, presence was represented by an interval variable C which contained the time when C. brevitarsis had first been observed (U) and the previous observation time (L) i.e. C = (L, U). When C. brevitarsis was caught within a week interval then t = U (exact time). When the catches were made more than a week apart then L < t < U. When C. brevitarsis failed to appear then t > L. The objective of the analysis was to model the dispersal data (t) as a linear function of site direct distance from the coast (D) and altitude (A). For the Hunter valley data, D = loge direct distance from site 16. The following model was proposed for each of the three data sets: t ¼ b0 þ b1 D þ b2 A þ e where b0, b1 and b2 are the coefficients of regression and e is the normally distributed error with mean zero and variance s2. The coefficients b1 and b2 were tested for their significance at a = 5% (two-sided likelihood-ratio x2). A generalised failure time analysis (Turnbull, 1974, 1976) was used to fit the model and run using S-PLUS (MathSoft, 1998). The predicted time for C. brevitarsis to be present can be calculated using the standard normal probability with the following form for each of the three data sets. z¼
t ½b0 þ b1 D þ b2 A s
3. Results 3.1. Waves of movement C. brevitarsis was found early and established quickly at sites in the endemic area (Table 1). Apparent delays were recorded at the sites closest to the coast despite suitable temperatures and known vector activity. The frequency at which C. brevitarsis was detected and established decreased and the time taken to be detected increased the further sites were from the coast. C. brevitarsis reached but failed to establish at two sites and failed to reach another two sites. Two waves of movement were observed in each year of the study and these were modelled. 3.2. Mid-northern coastal valleys 3.2.1. First-occurrence model The time that C. brevitarsis was recorded at the observation sites was related to the distance from the coast and the altitude of the site (Table 2). This suggested that it had moved out of the endemic area and moved up each valley. Dispersal time (S.E.) of C. brevitarsis was 18.2 2.2 weeks per 100 km at fixed altitude. The time taken for C. brevitarsis to travel a set distance was extended by 0.48 (0.22) weeks for every 100-m increase in altitude.
142
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
Table 2 Effects of distance and altitude on the dispersal times of C. brevitarsis in three coastal valleys (1995–2003) and in the Hunter valley (1993–2003) of New South Wales, Australia Source of variation
Regression coefficient
d.f.
LR (x2)
P (x2)
b
S.E.
First-occurrence model Intercept Distance (km) Altitude (m) Residual
3.28 0.18 0.005 –
0.98 0.022 0.002 –
1 1 1 51
– 142.86 4.52 4.77 (variance)
– <0.001 0.03
Second-occurrence model Intercept Distance (km) Altitude (m) Residual
3.27 0.16 0.011 –
1.11 0.026 0.002 –
1 1 1 42
– 123.62 16.58 5.28 (variance)
– <0.001 <0.001
Hunter valley model Intercept Loge distance (km) Altitude (m) Residual
0.65 2.11 0.037 –
1.59 0.53 0.007 –
1 1 1 14
– 45.51 31.30 5.41 (variance)
– <0.001 0.001
3.2.2. Second-occurrence model The time taken for C. brevitarsis to be recorded a second time at observation sites also was related to the distance from the coast and the altitude of the site (Table 2). This suggested that C. brevitarsis had moved out of the endemic area and moved up each valley a second time. Dispersal time of C. brevitarsis was 15.9 2.6 weeks per 100 km at fixed altitude. The time taken for C. brevitarsis to travel a set distance was extended by 1.14 (0.2) weeks for every 100-m increase in altitude. 3.3. Hunter valley model The time that C. brevitarsis was observed at the observation sites was related to the loge distance from site 16 and the altitude of the sites (Table 2). The time taken for C. brevitarsis to travel a set distance was extended by 3.7 (0.7) weeks for every 100-m increase in altitude. 3.4. Parous composition The total numbers of each parous stage recorded at the times C. brevitarsis occurred outside of the endemic area for the first and second time over the 8 years were nulliparous (40), gravid (66), post-parous (36) and males (4). There was a relation between year and parous stage (x2 = 25.4, d.f. = 10, P = 0.013) with greatest difference due to numbers of gravid females. About 71.8% of the total number of individuals was capable of carrying viruses.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
143
4. Discussion Knowing areas of Australia where C. brevitarsis is permanently or seasonally absent is necessary for defining the epidemiology of bluetongue and Akabane viruses and is critical for establishing protocols for the Australian livestock-export industry. C. brevitarsis normally is limited to the coastal plains of NSW (Bishop et al., 1995a, 2000a). Occasional incursions into western regions of NSW are a serious threat to the maintenance of a certifiable virus-free area. The model of Murray and Nix (1987) predicts incursions into western and north-western NSW from the north. However, their model depends on the presence of over-wintering foci from which both vector and virus develop and their predictions would be weakened or negated by the failure of C. brevitarsis to survive west of the GDR in NSW in most years. Incursions from vector-endemic regions in northern coastal NSW are feasible and are examined here. Our models indicate that the effect of distance on dispersal is significant (as shown previously by Bishop et al., 2000a) and relatively constant. Altitude slows movement further (i.e. it extends first and second presence by about 5 and 11 weeks, respectively, for the GDR’s 1000 m altitude). The GDR therefore helps prevent C. brevitarsis from reaching more-westerly sites (e.g. sites 14 and 15) or delays movements until activity could be stopped by declining temperatures late in the season (e.g. site 9). In previous studies of C. brevitarsis (Bishop et al., 1995b, 1996a), possible waves of movement were masked by individuals already present after earlier establishment. In our study, C. brevitarsis did not always establish the first time that it was recorded outside of the endemic area. Presence often was detected a second time at intervals indicating another wave of movement, although the frequency and nature of movements might not be represented fully. More than one movement would best explain the progressive spread of viruses from endemic foci in the study area long after the vector has moved initially (Barchia et al., 2002; Bishop et al., 1995b; P.D. Kirkland, personal communication). Any impact on livestock from C. brevitarsis movements to the west would depend on the vector’s ability to transmit virus on arrival. In our study, all parous stages potentially were represented when C. brevitarsis was recorded for a first or second time outside of the endemic area and the potential was higher for individuals capable of being infected or infective. However, midges moving early in the season are unlikely to be carrying virus and the potential for infection increases as the season progresses (Barchia et al., 2002). This would be facilitated by later movements of the vector. Additionally, C. brevitarsis must survive long enough at the top of the GDR for infected individuals to become infective or for infective individuals to locate a susceptible host. Survival appeared limited on and above the escarpment of the GDR, because in only 42% of occasions when C. brevitarsis was present outside of the endemic area over 8 years, did it survive long enough to become established. Typical movement by C. brevitarsis up the Hunter valley is slower than that down the coastal plain (Bishop et al., 2000a,b) and this is due to its distance from the endemic area and the effects of increasing altitude. The assumption that the Hunter valley is a common route for C. brevitarsis to the central-west was unfounded except under extremely favourable or infrequent and chance weather conditions (Murray, 1987b; Murray and Kirkland, 1995).
144
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
5. Conclusions Based on 8 years of data, we concluded that the altitude barrier provided by the GDR in NSW significantly delays (by 5–11 weeks) movements inland from the endemic coastal region in addition to previously demonstrated delays due to the distance to be travelled. These movements can occur on more than one occasion. Extended incursions onto the slopes and plains of NW NSW from the NSW coast would be rare.
Acknowledgements We thank Mr H. McKenzie for his assistance and technical expertise and Drs R. Van de Ven and P.D. Kirkland for their comments on the manuscript. The significant contribution of the many cooperators that operated light traps is also acknowledged. Funding was partially received from the National Arbovirus Monitoring Program.
References Allingham, P.G., 1990. An investigation of transovarial transmission of Akabane virus in Culicoides brevitarsis. Aust. Vet. J. 67, 273–274. Allingham, P.G., 1991. Effect of temperature on late immature stages of Culicoides brevitarsis (Diptera: Ceratopogonidae). J. Med. Entomol. 28, 878–881. Barchia, Idris M., Spohr, Lorraine J., Bishop, Alan L., 2002. Relationships between Culicoides brevitarsis parous stage and arbovirus transmission in New South Wales. Final Report to the Australian Quarantine Inspection Services, July 2002. NSW Agriculture, Gosford, Australia. Bellis, G.A., Gibson, D.S., Polkingham, I.G., Johnson, S.J., Flanagan, M., 1994. Infection of Culicoides brevitarsis and C. wadai (Diptera: Ceratopogonidae) with four Australian serotypes of bluetongue virus. J. Med. Entomol. 31, 382–387. Bishop, A.L., Barchia, I.M., Harris, A.M., 1995a. Last occurrence and survival during winter of the arbovirus vector Culicoides brevitarsis at the southern limits of its distribution. Aust. Vet. J. 72, 53–55. Bishop, A.L., Kirkland, P.D., McKenzie, H.J., Spohr, L.J., Barchia, I.M., Muller, M.J., 1995b. Distribution and seasonal movements of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) at the southern limits and its distribution in New South Wales and their correlation with arboviruses affecting livestock. J. Aust. Ent. Soc. 34, 289–298. Bishop, A.L., McKenzie, H.J., Spohr, L.J., Barchia, I.M., 1995c. Daily activity of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in the Hunter valley, NSW. Gen. Appl. Ent. 26, 31–39. Bishop, A.L., Kirkland, P.D., McKenzie, H.J., Barchia, I.M., 1996a. The dispersal of Culicoides brevitarsis in eastern New South Wales and associations with the occurrences of arbovirus infections in cattle. Aust. Vet. J. 73, 174–178. Bishop, A.L., McKenzie, H.J., Barchia, I.M., Harris, A.M., 1996b. Effect of temperature regimes on the development, survival and emergence of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in bovine dung. Aust. J. Entomol. 35, 361–368. Bishop, A.L., Barchia, I.M., Spohr, L.J., 2000a. Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med. 47, 243–254. Bishop, A.L., McKenzie, H.J., Barchia, I.M., Spohr, L.J., 2000b. Moon phase and other factors affecting light-trap catches of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Aust. J. Entomol. 39, 29–32. Braverman, Y., 1992. The possible introduction to Israel of Culicoides (Diptera Ceratopogonidae) borne animal diseases by wind. In: Walton, T.E., Osburn, B.I. (Eds.), Proceedings of the Second International Symposium on Bluetongue, African Horse Sickness, and Related Orbiviruses, Boca Raton, Florida.
A.L. Bishop et al. / Preventive Veterinary Medicine 65 (2004) 135–145
145
Dyce, A.L., Okey, D.H., 1981. Determination of parity in pinned specimens in the biting midge Culicoides variipennis. Mosq. News 41, 797–798. Dyce, A.L., Standfast, H.A., Kay, B.H., 1971. Collection and preparation of biting midges (Fam. Ceratopogonidae) and other small Diptera for virus isolation. J. Aust. Ent. Soc. 11, 91–96. MathSoft, 1998. S-PLUS 5 Guide to Statistics. MathSoft, Inc. Seattle, USA. 1014 pp. Muller, M.J., Standfast, H.A., St. George, T.D., Cybinski, D.H., 1982. Culicoides brevitarsis (Diptera: Ceratopogonidae) as a vector of arboviruses in Australia. In: St. George, T.D., Kay, B.H. (Eds.), Proceedings of the 3rd Symposium on Arbovirus Research in Australia, Brisbane, pp. 43–49. Murray, M.D., 1987a. Local dispersal of the biting midge, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in south-eastern Australia. Aust. J. Zool. 35, 559–573. Murray, M.D., 1987b. Akabane epizootics in New South Wales: evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J. 64, 305–308. Murray, M.D., Kirkland, P.D., 1995. Bluetongue and Douglas virus activity in New South Wales in 1989: further evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J. 72, 56–57. Murray, M.D., Nix, H.A., 1987. Southern limits of distribution and abundance of the biting-midge Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in south-eastern Australia: an application of the GROWEST Model. Aust. J. Zool. 35, 575–585. Sellers, R.F., Pedgley, D.E., Tucker, M.R., 1977. Possible spread of African horse sickness on the wind. J. Hyg. 79, 279–298. Sellers, R.F., Walton, T.E., Osburn, B.I., 1992. Weather Culicoides and the distribution and spread of bluetongue and African horse sickness viruses. In: Walton, T.E., Osburn, B.I. (Eds.), Proceedings of the Second International Symposium on Bluetongue, African Horse Sickness, and Related Orbiviruses, CRC Press, Boca Raton, USA, pp. 284–290. Strandfast, H.A., Dyce, A.L., St. George, T.D., Muller, M.J., Doherty, R.L., Carley, J.G., Filippich, C., 1984. Isolation of arboviruses from insects collected at Beatrice Hill, Northern Territory of Australia, 1974–1976. Aust. J. Biol. Sci. 37, 351–366. Turnbull, B.W., 1974. Nonparametric estimation of a survivorship function with doubly censored data. J. Am. Stat. Assoc. 69, 169–173. Turnbull, B.W., 1976. The empirical distribution function with arbitrarily grouped, censored and truncated data. J. R. Stat. Soc. B. 38, 290–295.