Comparaison de pièges à Culicoides en Afrique du Sud

Résumé

The primary monitoring tools for collecting Culicoides midges (Diptera: Ceratopogonidae) are various models of light traps. To facilitate data comparison, four downdraught light traps were used in summer under South African conditions. These were the Onderstepoort trap [220 V, ultraviolet (UV)], the Rieb trap (12 V, UV) from France, the mini-CDC trap (6 V, UV) from the USA , and the Pirbright trap (220 V, white light) from the United Kingdom. Traps were deployed in three replicates of a 4x4 randomized Latin square design, so that treatment means were independent of any effects caused by the site or the occa­sion. Trapping was conducted during 12 nights in January 2008 (summer) and comprised 48 collections with 643,374 Culicoides midges collected. Eighteen different Culicoides species, of which only six were found in all four traps, were collected. All four traps indicated Culicoides imicola to be the most abundant species. Its abundance ranged from 91.8% (Rieb) to 95.0% (Onderstepoort). Statistically significant differences were found in the total number of Culicoides collected by each of these traps. The Onderstepoort trap (407,411) collected statistically significantly more midges than the CDC (167,794), Pirbright (39,128) or Rieb trap (32,041). Significant differences were also found between the traps in the parous rate, sex ratio and number of Culicoides midges compared to those of other insects. When comparisons were repeated in winter, when Culicoides numbers were relatively low compared to summer, the Onderstepoort trap still performed significantly better. All four traps, however, still captured Culicoides midges. In addition, the BG-sentinel mosquito trap, from Germany, was compared with the other four traps in two replicates of a 5x5 Latin square. These com­parisons, which were done in spring, showed that the number of Culicoides midges collected with the BG-sentinel was not significantly different from that collected with the CDC trap. Although the Onderstepoort trap increased monitoring sensitivity in areas where vector abundances were low, results highlighted the notion that biases in trapping methods need to be evaluated and measured.