Date Published: November 17, 2009
Publisher: Public Library of Science
Author(s): Jennifer E. Simpson, Corrine M. Folsom-O’Keefe, James E. Childs, Leah E. Simons, Theodore G. Andreadis, Maria A. Diuk-Wasser, Colin J. Sutherland. http://doi.org/10.1371/journal.pone.0007861
Abstract: Evidence from field studies suggests that Culex pipiens, the primary mosquito vector of West Nile virus (WNV) in the northeastern and north central United States, feeds preferentially on American robins (Turdus migratorius). To determine the contribution of innate preferences to observed preference patterns in the field, we conducted host preference trials with a known number of adult female C. pipiens in outdoor cages comparing the relative attractiveness of American robins with two common sympatric bird species, European starling, Sternus vulgaris and house sparrow, Passer domesticus. Host seeking C. pipiens were three times more likely to enter robin-baited traps when with the alternate host was a European starling (n = 4 trials; OR = 3.06; CI [1.42–6.46]) and almost twice more likely when the alternative was a house sparrow (n = 8 trials; OR = 1.80; CI = [1.22–2.90]). There was no difference in the probability of trap entry when two robins were offered (n = 8 trials). Logistic regression analysis determined that the age, sex and weight of the birds, the date of the trial, starting-time, temperature, humidity, wind-speed and age of the mosquitoes had no effect on the probability of a choosing a robin over an alternate bird. Findings indicate that preferential feeding by C. pipiens mosquitoes on certain avian hosts is likely to be inherent, and we discuss the implications innate host preferences may have on enzootic WNV transmission.
Partial Text: Heterogeneities in contact rates between arthropod vectors and hosts are important to vector-borne disease dynamics, because they can result in increased disease transmission if vector blood meals occur more commonly on pathogen-competent hosts. In contrast, transmission may be reduced if blood meals are ‘diluted’ by feeding on non-competent hosts.
Host-seeking C. pipiens were significantly more attracted to robins than to either sparrows or starlings (Table 1), but no difference was detected when trials were conducted with two robins (Additional information for individual trials are provided in Supplementary Tables S1–S3). C. pipiens were three times more likely to enter robin-baited traps when paired with starling-baited traps (OR = 3.06; CI [1.42–6.46]) and almost twice more likely when paired with sparrow-baited traps (OR = 1.80; CI = [1.22–2.90]).
Our results indicate that C. pipiens display preference for robins when offered a choice between a robin and one of two locally common, sympatric bird species. The degree of preference varied depending on the alternate species, with robins selected approximately three times over European starlings and two times over house sparrows. This preference was not affected by potential confounders such as age of the mosquito cohort, birds’ weight, age and sex and differences in body weight between the trial pair and the environmental conditions (temperature, relative humidity and wind speed) in which trials were held. Furthermore no location bias in trap selection was indicated by the equal probability of mosquitoes entering either trap when both were baited with robins. The only covariate which significantly influenced host choice was the species with which the robin was paired. We acknowledge that the current investigation evaluated the attractiveness of three avian species only, and that further experiments with larger numbers will need to be conducted to determine whether a preference for robins is maintained when paired with other common Passeriform species upon which C. pipiens is known to feed.