Assessing the risk of windborne spread of bluetongue in the 2006 outbreak of disease in northern Europe

BLUETONGUE virus (BTV) is an insect-transmitted pathogen that causes bluetongue, and can infect all species of ruminant. Cattle are reservoirs and usually show little evidence of disease, but in sheep, particularly European breeds, the virus causes a disease of major significance. BTV is transmitted between its ruminant hosts by certain species of Culicoides biting midge, which are obligate vectors, and experimental work has shown that it is transmitted most efficiently by these midges at temperatures of approximately 28 to 30°C, much less efficiently at lower temperatures and probably not at all below 10°C (Purse and others 2005). BTV distribution is therefore limited to regions where vector midges occur and transmission is limited to periods of permissive temperature. However, driven by climate-change (Purse and others 2005), the major southern European vector Culicoides imicola is expanding its range northwards and is now found throughout southern Europe ranging as far north as southern France and southern Switzerland. These changes have caused increased overlap in the distribution of C imicola and more northerly Culicoides species, some of which, for example Culicoides obsoletus and Culicoides pulicaris are also able to transmit BTV. This has resulted in the further northerly extension of BTV. A major concern is that the C obsoletus/C pulicaris midges are common throughout central and northern Europe, including the UK. Vector Culicoides are approximately 1·5 mm long. They tend to be active around sunset/sunrise and during the night. Their normal flight range is less than 1 to 2 km, but they can also be transported on the wind over distances of several 100 km. This is believed to have led to the introduction of BTV into locations remote from the source, for example, the Spanish Balearics from North Africa and/or Sardinia. Circumstantial evidence suggests that such windborne transport is more efficient over sea than land. These windborne movements of vectors are of major concern and make BTV and other Culicoides-borne viruses difficult to control. Currently, such movements are almost impossible to predict because virtually no information exists about the complex interactions between Culicoides species and the meteorological conditions that facilitate their spread. On June 1, 2006, DEFRA commission a research project to investigate the spread of bluetongue by the windborne vector Culicoides species. The overall objectives of the work are to monitor developments of bluetongue in response to environmental change due to global warming and to develop an operational early warning scheme for the UK. This would enable the deployment of appropriate and targeted surveillance and control measures with the minimum of delay. On August 17, 2006, bluetongue was confirmed in an established and closed flock of 90 sheep in the Netherlands. The next day, disease was also confirmed in 11 outbreaks in Belgium (seven in cattle and four in sheep). On August 21, disease was also confirmed in both cattle and sheep in North Rhine, Germany. In total, 33 outbreaks were declared within five days, all of which were located on the borders of the Netherlands, Belgium and Germany, and were the first cases of BT ever reported in each of these three countries. Once the alert was raised, DEFRA asked three questions. First, was it possible for the initial introduction of BTV to be as a result of windborne infected midges? Secondly, what was the likelihood for further spread within the immediate area, and thirdly, what was the potential for disease spread to the UK. This short communication describes how the questions were addressed and presents the results as preliminary findings, which will be followed in due course by a full scientific paper and the introduction of an enhanced operational risk assessment system. At the time of the analysis, the only declared bluetongue outbreak area in southern Europe/Mediterranean was in Algeria (BTV serotype 1), although BTV had been active during 2005 in Spain, Portugal and Italy, but in these cases the causative viruses were known to be of serotypes different from the one in northern Europe during 2006 (BTV serotype 8). The Met Office operational atmospheric dispersion model (NAME) (Ryall and Maryon 1998, Jones and others 2007) was used as the prime tool for answering each of the questions raised by DEFRA. NAME was run for seven-day periods commencing on July 1, using meteorological data from the Met Office’s mesoscale weather forecast model, to determine if the air in the outbreak area had passed over a known infected area. The analysis showed that, for the vast majority of the time, air reaching the infected area in Europe had involved a passage of thousands of kilometres over the Atlantic Ocean or from countries to the east of the outbreak. On a few occasions air would have passed from Spain towards the infected area. However, even on those days, it was believed that the air had not passed close to Algeria. Consequently, it was believed highly unlikely that infected midges, carried by the wind, during the period of study were responsible for the introduction of infection to the Netherlands/Belgium/Germany. To model both the potential for local and longer distance transport of BTV-infected midges, it was necessary to make a number of simplistic assumptions about the behaviour of the midges. These were: take off was most likely from 18.00 to 21.00; activity increased with temperature (no activity below 10°C, rising to a maximum between 25 and 30°C); and activity was unlikely during wet weather, as this reduced the likelihood of take-off and increased the potential for washHigh Medium Low Very Low No risk 20°C Temperature 10-20°C?