Mosquito Vector Competence Tests for West Nile Virus: What do they Mean for Florida?

The finding of a West Nile positive crow in Jefferson County reported by the Florida Department of Health on July 6, 2001 heralds the arrival of WN virus in Florida. The next few months will certainly be interesting as we try to understand the implications of WN in Florida. The FMEL will continue to monitor and report on the situation on its Encephalitis Information System (

Some individuals have suggested that information on the vector competence of Florida mosquitoes for WN virus is essential to our ability to mitigate WN's impact in Florida. As someone who has conducted research on arthropod arboviral vector competence, I do not agree that vector competence experiments are a high priority at this time.

Recall that vector competence is just one of the suite of characters that define vector ability or "capacity" to transmit a pathogen. Competence includes a vector's susceptibility to infection with a pathogen, and the capacity of the infected vector to transmit that pathogen to a susceptible host. In addition to competence, vector capacity includes a variety of other essential traits, i.e., host preference, longevity, extrinsic incubation time, and abundance, all critical to determining whether a species is an efficient disease vector. Therefore, even if one agreed that laboratory experiments could accurately evaluate vector competence, one would still be unable to evaluate vector potential in the absence of similar information for other vector capacity traits.

For purposes of discussion, I will consider only vector susceptibility to infection with WN virus. Is laboratory-collected information useful in predicting vector susceptibility to infection in nature? Like many other complex traits, mosquito susceptibility to WN virus infection is influenced both by genetic and environmental factors. There is a large body of literature devoted to distinguishing the influence of genetic and environmental components to variation within and between mosquito species for arboviruses, e.g., Aedes aegypti for yellow fever and dengue, and Culicoides sonorensis for bluetongue virus. If you are interested in the details see Tabachnick, WJ. 1994. Genetics of insect vector competence for arboviruses. Advances in Disease Vector Research 10: 93 - 108.

There are several difficulties using laboratory-generated information on vector competence to predict what will occur in nature. Laboratory methods are straightforward. Potential vectors, either colonies or field-collected insects, are exposed to the virus through blood feeding. Blood fed individuals are held for an extrinsic incubation period, and then assayed to determine the proportion containing living virus. Most experts agree that the use of colony material for such tests should be avoided because colonized insects are often not representative of field populations. However, even field-collected populations may not represent the true genetic diversity of the species. Hence numerous populations need to be tested. Laboratory tests also fail to account for the effects of environmental effects on wild mosquitoes or the possibility that the behavior of a species in the laboratory has been influenced by the artificial environmental parameters of the test. For readers of this column, recall the experiments on Aedes aegypti by Jesse Lazear and James Carroll in Walter Reed's yellow fever research group (Tabachnick, WJ. 2000. Mosquito Variability: Did the mosquito do it? Buzz Words, Aug./Sept. 2000). Initially, Ae. aegypti failed to transmit yellow fever in their study because they used the wrong environment (too short an incubation period), and possibly the wrong genetic material since we now know there are indeed genetic differences between Ae. aegypti females for their ability to transmit yellow fever virus.

So what will vector competence information for WN virus in Florida mosquitoes tell us? Species that are capable of being infected in the laboratory are certainly candidates. However, it remains uncertain whether they will be capable vectors in nature. What about species that are not susceptible in the laboratory? Would you consider them less likely to be vectors in nature? I submit that one needs to be careful here. What would be your response if a species that is not readily infected in the laboratory, proved to be an epidemic vector? Clearly, possible explanations are that the vector competence tests done in the laboratory or the tested mosquitoes were not representative of circumstances in the field. This is my point: there is little predictability that will result from laboratory vector competence studies for "anticipating" WN vectors in Florida.

Vector competence tests can be invaluable, but not as "predictors" of what will occur during epidemics. Predictability is tricky and I wish we were better at it, but we are not. Now that WN virus has been found in Florida we need epidemiological and field data on its distribution in Florida. Analyses of geographic variation in vector competence may then provide an understanding on the causes of any geographic differences in the risk for virus transmission in Florida. This was used for C. sonorensis vector competence and the distribution of bluetongue viruses in North America. However, this analysis relied on more then 30 years of epidemiological information on virus distribution (Tabachnick, WJ. 1996. The genetics of Culicoides variipennis and the epidemiology of bluetongue disease in North America. Ann. Rev. of Entomol. 45:20-40).

What can Florida do with vector competence information at this time? It would be foolhardy for any mosquito control program to base their control efforts on laboratory results showing which vector species appear to be the most efficient vector of WN virus. So what should we do? I submit that vector competence assays of different species are not high priority at this time for Florida. Now that WN virus has been introduced into Florida, the FMEL, PHEREC, DOACS, DOH and Florida Mosquito Control Districts must work efficiently to identify the actual Florida WN virus vectors. This information might be consistent with any laboratory results, or it might not. However, once the WN virus vectors are identified within actual transmission zones, any prior laboratory vector competence studies will be irrelevant other then in the context of geographic variation.

I suggest that a high priority issue for Florida is to develop a real-time data collection system that will quickly and efficiently identify those mosquitoes responsible for actual transmission of WN virus in various locations throughout Florida soon after the virus is detected in a region. The FMEL is working to develop such a system - we call it the Arbovirus Rapid Deployment System. WN virus in Jefferson County means it is time for action.

Walter J. Tabachnick,
Director, FMEL