Japanese encephalitis in south Arcot district, Tamil Nadu, India: a three-year longitudinal study of vector abundance and infection frequency.

In the South Arcot district, an area endemic for Japanese encephalitis in Tamil Nadu, Culex tritaeniorhynchus Giles, Culex vishnui Theobald, Culex gelidus Theobald and Culex fuscocephala Theobald constituted 93.6% of 422,621 adult females representing 27 culicine species collected between August 1991 and July 1994. Vector abundance was lowest in the hot and dry season (April-June) and highest in the cool and wet season (October-December). Overall, 285,531 adult female mosquitoes (5,710 pools) were tested for virus using an enzyme-linked immunosorbent assay or by inoculation into larvae of Toxorhynchites splendens Wiedemann and identification by immunofluorescent test using JE virus specific monoclonal antibody or by both. In total, 91 isolations were made, of which 80 (88%) were identified as JE virus; 58 isolations were from Cx. tritaeniorhynchus, 22 from Cx. vishnui, 6 from Cx. fuscocephala and 5 from Cx. gelidus, giving similar minimum infection rates (MIR) of 0.28, 0.41, 0.39, and 0.52, respectively. Vector abundance and MIR increased from July concurrently with the initiation of rice cultivation. MIR peaked in September followed by a decrease in October, but mosquitoes remained abundant until March. The decrease in MIR from October onward coincided with rising herd immunity in pigs. Although MIRs in October (0.47) and November (0.42) were lower than in September (0.92), a comparable high risk of infection for humans continued because of high vector abundance and human biting rates. In the South Arcot district, the probability of a child receiving an infective bite was 0.53 per JE transmission season.