Host preference of Japanese encephalitis vectors in Chiangmai, Northern Thailand

All night mosquito collections by light traps and by a mouth aspirator for 5 minutes around animal sheds were conducted at Mae Joh University compound, northern Thailand on three nights between August and September, 1998. These were supplemented with choice/no choice experiments using nets in which mosquitoes collected from the rice field or a net with a cow or a pig were released. In the experimental nets a cow or a pig or both were tethered to evaluate precise host preference of Japanese encephalitis vectors. Light traps collected 34,708 female mosquitoes among which were 7,958 Cx. tritaeniorhynchus, 22,782 Cx. vishnui and 2,083 Cx. gelidus, the main JE vectors. Only 114 females of Cx. fuscocephala, another JE vector, were captured. The proportions of fed females showed that Cx. tritaeniorhynchus and Cx. vishnui obtained more meals from the pigs than from the cows whereas Cx. gelidus fed on the cows more than the pigs in significantly different proportions. A total of 1,811 (47.2% fed) and 1,701 (38.2% fed) females of JE vectors were picked by aspirators at the cowshed and pigsty respectively. There was no difference in the proportions of fed females of Cx. tritaeniorhynchus and Cx. vishnui between the cowshed and pigsty collections, and Cx. gelidus fed significantly more on the cows than the pigs. Light trap collections revealed the host range of the mosquitoes in nature, which depends on availability of blood meal sources. The choice/no-choice experiments revealed stronger preference of JE vectors for the cow to the pig. We considered this to be the precise host preference because the experiments were conducted under conditions of equal availability of hosts.