To mitigate the effects of West Nile virus (WNV) and eastern equine encephalitis virus (EEEV), the state of Florida conducts a serosurveillance program that uses sentinel chickens operated by mosquito control programs at numerous locations throughout the state. Coop locations were initially established to detect St. Louis encephalitis virus (SLEV), and coop placement was determined based on the location of human SLEV infections that occurred between 1959 and 1977. Since the introduction of WNV into Florida in 2001, WNV has surpassed SLEV as the primary arbovirus in Florida. Identifying high probability locations for WNV and EEEV transmission and relocating coops to areas of higher arbovirus activity would improve the sensitivity of the sentinel chicken surveillance program. Using 2 existing models, this study conducted an overlay analysis to identify areas with high probability habitats for both WNV and EEEV activity. This analysis identified approximately 7,800 km2 (about 4.5% of the state) as high probability habitat for supporting both WNV and EEEV transmission. Mosquito control programs can use the map resulting from this analysis to improve their sentinel chicken surveillance programs, increase the probability of virus detection, reduce operational costs, and allow for a faster, targeted response to virus detection.

West Nile virus (WNV) has the greatest distribution worldwide (Ciota and Kramer 2013) and is the number 1 cause of domestically acquired arboviral disease in the continental USA (McDonald et al. 2019). First isolated in the West Nile district of Uganda in 1937 (Smithburn et al. 1940), WNV spread globally, arriving in the USA in New York City in 1999, where it rapidly swept through all 48 contiguous states (Curren et al. 2018). Approximately 80% of people infected with WNV will remain asymptomatic (Chancey et al. 2015) with approximately 20% of those infected developing flu-like symptoms (headache, fever, muscle and joint pain, nausea, vomiting, and occasionally a rash) of varying intensity, a condition known as West Nile fever (Watson et al. 2004). These cases go largely unreported. Approximately 1% of infections will result in a neuroinvasive condition (Rossi et al. 2010), generally encephalitis or meningitis. From 1999 to 2019, the Centers for Disease Control and Prevention (CDC) has recorded 51,801 WNV cases and 2,390 fatalities (CDC 2021b). Of these cases, 25,290 were neuroinvasive with 2,259 fatalities (CDC 2021b).

West Nile virus is maintained through an enzootic cycle between mosquito vectors with birds acting as amplifying hosts (Chancey et al. 2015). In Florida, the common vectors of WNV are Culex quinquefasciatus (Say) and Cx. nigripalpus (Theobald) (Burkett-Cadena 2013; Turell et al. 2005). These species of mosquito act as both a vector for maintenance of the enzootic cycle and as a bridge vector for transmission to humans and other mammals (Turell et al. 2005).

The first human cases of eastern equine encephalitis virus (EEEV) in the USA were documented in 1938 following an outbreak in southeastern Massachusetts (Feemster 1938). Eastern equine encephalitis virus activity is found in the eastern portion of the USA, with the majority of infections occurring along the Atlantic and Gulf coasts and to a lesser extent near the Great Lakes (Lindsey et al. 2018). The majority of EEEV infections remain asymptomatic or present as a self-resolving febrile illness, with less than 5% of infections progressing to encephalitis (CDC 2021a). From 2010 to 2019, the CDC has recorded 107 human cases of EEEV resulting in 48 fatalities, or a mortality rate of 45%. All reported cases were neuroinvasive (CDC 2021a).

Similar to WNV, EEEV is maintained through an enzootic cycle between mosquitoes and birds. However, the mosquito responsible for the maintenance of the enzootic cycle, Culiseta melanura (Coq.), is not generally considered an important bridge vector due to its ornithophilic feeding preference, rarely feeding on mammals or reptiles (Edman et al. 1972, Magnarelli 1977, Molaei et al. 2006, Burkett-Cadena et al. 2015). Bridge vectors of EEEV, including Aedes spp., Coquillettidia spp., and Culex spp., are believed to be responsible for the infrequent human and horse disease outbreaks by feeding on both mammals and birds (Chamberlain et al. 1956, Vaidyanathan et al. 1997, Cupp et al. 2003, Bingham et al. 2016).

In Florida, mosquito surveillance is the responsibility of mosquito control programs (MCPs) managed at the city, county, or special taxing district level. Sixty-three state approved programs exist within Florida with total expenditures exceeding $150 million annually (Tabachnick 2008). The MCPs use a variety of surveillance systems, including mosquito traps, larval dipping, and serosurveillance to monitor mosquito presence and characterize health risks in their respective regions. The arboviral sentinel chicken surveillance program in Florida, implemented in 1978, was a result of outbreaks of St. Louis encephalitis virus (SLEV) that occurred between 1959 and 1977 (Shroyer and Rey 1990). Initial sampling locations were selected near known human SLEV cases (Shroyer and Rey 1990), with subsequent locations selected based upon recommendations of the Florida Interagency Arbovirus Task Force (FDOH 2019), taking into consideration maintenance and sampling convenience to determine their final placement. Since the program's inception, WNV and EEEV have emerged as the primary arboviruses within the state. Revising the existing coop placement protocol to serve areas determined as high probability for WNV and EEEV would improve surveillance efficacy.

Here we report the results of a spatial analysis for use by MCPs in the state of Florida that should improve the existing sentinel chicken surveillance program by focusing on areas identified as high probability habitats for both WNV and EEEV. This analysis makes use of existing WNV and EEEV models that take into consideration Florida's climate, ecology, land cover, and geophysical traits. These results should enable MCPs to increase the probability of detection for WNV and EEEV, reduce operational costs by directing resources to areas identified as high probability for both EEEV and WNV activity, and allow for earlier and more focused mosquito control responses to viral activity.

Study area

The state of Florida is the most southeastern state of the USA with a geographic area of 170,300 km2 (USCB 2010). The majority of the state is bordered by the Atlantic Ocean, Gulf of Mexico, and the Straits of Florida on its eastern, western, and southern borders, respectively. Elevation ranges from sea level to 345 ft (University of Florida Geoplan Center 2013). The Florida climate ranges from subtropical to tropical with mild winters and abundant rainfall (Main and Allen 2007, Collins et al. 2017, Miley et al. 2020). These characteristics contribute to Florida's unique ecosystem and, as a result, allow for year-round transmission of EEEV and WNV (Bigler et al. 1976, CDC 2002, Shaman et al. 2004, Day and Shaman 2011).

Existing models

The WNV model was developed using sentinel chicken serosurveillance data from 2014 to 2018 provided by the Florida Department of Health in conjunction with all Florida MCPs. These data in combination with publicly available land cover and remote sensing environmental and geophysical raster data were used to create a weighted average ensemble of 3 machine learning models—maximum entropy, random forest, and boosted regression trees—to predict high probability habitats in Florida (Beeman et al. 2021).

The EEEV model is a risk index model developed using veterinary horse case data from 2005 to 2010, in combination with land cover variables to predict EEEV risk to horses in Florida (Kelen et al. 2014). Model inputs were derived from a previously published study designed to quantify habitats associated with EEEV equine cases (Kelen et al. 2012). Both model rasters characterize the viral habitat with values ranging from 0 to 1, representing the presence probability of the target virus.

Overlay model

ArcGIS Pro version 2.8.1 was used for spatial analysis (Environmental Systems Research Institute 2021). The EEEV model was projected to match the USA contiguous Albers equal area conic USGS of the WNV model. Both models were used at their native resolution of 30 × 30 m cells. A quintile equal area analysis for each model was conducted along with generation of histograms representing their frequencies. The cell values for each model were categorized as very low (0–0.2), low (0.2–0.4), moderate (0.4–0.6), high (0.6–0.8), and very high (0.8–1) probability for each virus specific model.

New raster layers for each model were created representing a binary probability habitat for each virus using the reclassify tool. High and very high probability value cells (0.6–1) were aggregated and defined as high probability habitat, with all remaining cells (0–0.6) aggregated as defined as low probability habitat. The raster calculator was then used to create an overlay of the binary habitat rasters. The resulting raster presents a cross-tabulation of 4 classes: WNV and EEEV high probability, WNV high and EEEV low probability, WNV low and EEEV high probability, and WNV and EEEV low probability. This raster was filtered to indicate only the cells representing WNV and EEEV high probability habitat congruence. This overlay raster was mapped with a state shape file (USCB 2019) and primary and county roads shapefiles (USCB 2013) to improve its operational value to Florida MCPs. Risk probability values for existing sentinel coop locations were extracted and summarized in order to evaluate how well those locations may capture WNV and EEEV transmission.

The WNV model distributions, by quintile probability category, were 46% very low, 17% low, 10% moderate, 17% high, and 10% very high. The WNV model mean pixel value was 0.351 (SD 0.287). As seen in Fig. 1, the majority of very high and high probability habitat is found in the peninsular region of the state along the coast and extending inland. Another region of very high and high probability habitat is seen in the western panhandle region.

Fig. 1.

WNV risk model: Colors indicate relative risk from low (blue) to high (red) as indicated in the inset.

Fig. 1.

WNV risk model: Colors indicate relative risk from low (blue) to high (red) as indicated in the inset.

Close modal

The EEEV model distributions by probability category were 52% very low, 11% low, 22% moderate, 7% high, and 8% very high. The EEEV model mean pixel value was 0.301 (SD 0.282). As seen in Fig. 2, the very high and high probability habitat for EEEV is more widely dispersed than that of WNV, and is more common inland.

Fig. 2.

EEEV risk model: Colors indicate relative risk from low (blue) to high (red) as indicated in the inset.

Fig. 2.

EEEV risk model: Colors indicate relative risk from low (blue) to high (red) as indicated in the inset.

Close modal

The overlay model indicated that 5% of the state (7,800 km2) met the study criteria of high probability for both WNV and EEEV, 11% (18,052 km2) met the criteria of low probability for WNV and high for EEEV, 23% (39,918 km2) met the criteria of high probability for WNV and low for EEEV, and 61% (104,530 km2) met the criteria of low probability for both WNV and EEEV. As seen in Fig. 3, the peninsular region of the state had the greatest area of high probability for both WNV and EEEV along the coastal region and extending inland. The panhandle region contained areas of high probability for both viruses, but these were far more dispersed. The southern portion of the peninsular region indicated little to no areas of high probability.

Fig. 3.

Combined risk map of Florida for WNV and EEEV: Red color indicates areas at high risk for both viruses.

Fig. 3.

Combined risk map of Florida for WNV and EEEV: Red color indicates areas at high risk for both viruses.

Close modal

From 2014 to 2018, 307 sentinel chicken coops were in operation across the state of Florida. Of these, 67 (22%) are located within an area identified as high probability for both WNV and EEEV; 186 (61%) are located within an area that is high probability for 1 virus, but low for the other; and 54 (18%) are located within an area that is low probability for both WNV and EEEV (Fig. 4).

Fig. 4.

Existing locations of sentinel chicken coops relative to calculated risk for EEEV and WNV: Coop locations are indicated by black circles. Colors indicate relative risk for both viruses from low (blue) to high (red) as indicated in the inset.

Fig. 4.

Existing locations of sentinel chicken coops relative to calculated risk for EEEV and WNV: Coop locations are indicated by black circles. Colors indicate relative risk for both viruses from low (blue) to high (red) as indicated in the inset.

Close modal

The individual WNV and EEEV models generally predict different regions of the state as being at high risk for their targeted viruses. These differences are the result of several factors, including viral vectors/bridge vectors and viral presence data. Thus, a relatively small proportion of the state was identified in the overlay analysis as high risk for both viruses. While both viruses are maintained through an enzootic cycle, there are a number of differences in their ecology and life cycle. The enzootic cycle of WNV occurs primarily between birds and mosquitoes, with Alligator mississipiensis (Daudin) also identified as an amplifying host (Klenk et al. 2004). Culex quinquefaciatus and Cx. nigripalpus serve as both the primary enzootic vectors and the primary bridge vectors in Florida. This differs from the enzootic cycle of EEEV where, in addition to birds, studies have implicated rodents and a variety of reptiles as potential EEEV hosts (Arrigo et al. 2010, White et al. 2011, Bingham et al. 2012, Miley et al. 2021). The primary enzootic vector for EEEV is Cs. melanura, and while it has been identified as a potential bridge vector, additional Aedes spp., Coquillettidia spp., and Culex spp. have been implicated as bridge vectors. The ecological niche of Cx. quinquefaciatus and Cx. nigripalpus overlaps as indicated by their involvement in the enzootic cycle of WNV. However, this niche differs from the optimal habitat for Cs. melanura. The ecological niche of these mosquito species is unique, which impacts the geospatial distribution of the viruses for which they act as a vector or bridge vector.

Both the WNV and EEEV models used serosurveillance data for model development. However, the WNV model used sentinel chicken surveillance coop locations, which provides a specific location for infection origin. The EEEV model used veterinary equine data, which is georeferenced to the address of the owner of the infected horse. While this provides a location for modeling purposes, it is generalized, indicating a location in which the infected horse is located the majority of the time, but does not necessary represent the specific location that the infection occurred. Another limitation of the EEEV model is that while it was designed to correct for differences in horse populations in different areas of the state, it did not correct for differences in horse vaccination rates throughout the state, since these data are not recorded by the state. The EEEV model thus implicitly assumes a homogeneous vaccination rate throughout the state, which may not be accurate.

Development of a unified model to identify probable habitats for more than 1 virus is key to creating an effective sentinel surveillance program. While both WNV and EEEV have been identified by the existing program, over 75% of the coops in operation were not placed to have a high probability of detecting both viruses. As a result, the true incidence of each virus may be underestimated. However, one disadvantage of the combined model is that it includes some sites that are at high risk for transmission of one virus, but low risk for the other. In this case, if one of the viruses is considered to be of greater public health importance in a given district, programs may wish to consider running the underlying virus specific models individually and placing some coops in places where the risk is high for only the virus that is the district's priority.

A future study should be considered to conduct sentinel surveillance within the Southern Coast Plain ecoregion of Florida. Currently no sentinel surveillance is conducted within this ecoregion, forcing the existing models to extrapolate beyond the bounds of the existing data. As a result, predictions within this region should be considered with caution. Sentinel surveillance of the Southern Coastal Plain would be useful to determine which, if any, arboviruses are present and if or to what degree this ecoregion may contribute to the year-round transmission of WNV and EEEV.

The rasters developed are hosted by the University of South Florida Geographic Information Systems Library at URL https://arcg.is/1bG1SL2, allowing MCPs to access and use the model in an operational manner. The model provides MCPs with selectable layers for high and low probability habitats for both WNV and EEEV, the ability to input existing and potential future sentinel chicken surveillance coop locations, and overlays the World Street Map basemap. This will allow MCPs to select locations for sentinel chicken coop placement that provide a high probability of detecting both WNV and EEEV while also considering site access and maintenance concerns. Furthermore, the 30-m resolution of the model allows for greater ease of coop placement within a selected location while still providing sufficient spatial specificity for probability prediction. However, ground truthing in conjunction with Florida MCPs should be conducted to verify the model in an operational setting. This should be done selectively with a few MCPs to determine model efficacy before widespread implementation of the model.

While not a complete arboviral model, the model presented in this study allows for a substantial improvement to the existing sentinel chicken surveillance program in Florida. This is achieved through the integration of GIS with modern modeling algorithms, which evaluate vector ecological interactions to predict high potential arboviral habitat sites for sentinel coop placement resulting in improved surveillance for both WNV and EEEV. In addition, this model will likely provide effective surveillance for SLEV due to the overlap of WNV and SLEV vectors. Optimal locations for a statewide network of surveillance sites could be determined using risk models and spatial optimization modeling (Downs et al. 2020), so future work might explore optimal configurations of sites to cover both WNV and EEEV in Florida based on the results presented here. Improvements in the sentinel chicken surveillance program in Florida will provide public health benefits through earlier detection of circulating arboviruses in the environment allowing for a more rapid, targeted response by MCPs. The MCPs will further benefit by reducing costs through the reallocation of limited resources from low to high probability habitats for arboviruses.

This publication was supported by Cooperative Agreement Number U01CK000510, funded by the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.

Arrigo
NC,
Adams
AP,
Watts
DM,
Newman
PC,
Weaver
SC.
2010
.
Cotton rats and house sparrows as hosts for North and South American strains of eastern equine encephalitis virus
.
Emerg Infect Dis
16
:
1373
1380
.
Beeman
SP,
Morrison
AM,
Unnasch
TR,
Unnasch
RS.
2021
.
Ensemble ecological niche modeling of West Nile virus probability in Florida
.
PLoS One
16
:
e0256868
.
Bigler
WJ,
Lassing
EB,
Buff
EE,
Prather
EC,
Beck
EC,
Hoff
GL.
1976
.
Endemic eastern equine encephalomyelitis in Florida: a twenty-year analysis, 1955-1974
.
Am J Trop Med Hyg
25
:
884
890
.
Bingham
AM,
Burkett-Cadena
ND,
Hassan
HK,
Unnasch
TR.
2016
.
Vector competence and capacity of Culex erraticus (Diptera: Culicidae) for eastern equine encephalitis virus in the Southeastern United States
.
J Med Entomol
53
:
473
476
.
Bingham
AM,
Graham
SP,
Burkett-Cadena
ND,
White
GS,
Hassan
HK,
Unnasch
TR.
2012
.
Detection of eastern equine encephalomyelitis virus RNA in North American snakes
.
Am J Trop Med Hyg
87
:
1140
1144
.
Burkett-Cadena
ND.
2013
.
Mosquitoes of the Southeastern United States
.
Tuscaloosa, AL
:
The University of Alabama Press
.
Burkett-Cadena
ND,
Bingham
AM,
Hunt
B,
Morse
G,
Unnasch
TR.
2015
.
Ecology of Culiseta melanura and other mosquitoes (Diptera: Culicidae) from Walton County, FL, during winter period 2013-2014
.
J Med Entomol
52
:
1074
1082
.
CDC [Centers for Disease Control and Prevention].
2002
.
West Nile virus activity—United States, 2001
.
MMWR Morb Mortal Wkly Rep
51
:
497
501
.
CDC [Centers for Disease Control and Prevention].
2021a. Eastern equine encephalitis—statistics and maps [Internet]
.
Atlanta, GA
:
Centers for Disease Control and Prevention [accessed September 5
,
2021]
.
CDC [Centers for Disease Control and Prevention].
2021b. Final cumulative maps & data for 1999-2019 [Internet]
.
Atlanta, GA
:
Centers for Disease Control and Prevention [accessed September 5
,
2021]
.
Chamberlain
RW,
Kissling
RE,
Stamm
DD,
Sudia
WD.
1956
.
Transmission of eastern equine encephalitis to horses by Aedes sollicitans mosquitoes
.
Am J Trop Med Hyg
5
:
802
808
.
Chancey
C,
Grinev
A,
Volkova
E,
Rios
M.
2015
.
The global ecology and epidemiology of West Nile virus
.
Biomed Res Int
2015
: 376230.
Ciota
A,
Kramer
L.
2013
.
Vector-virus interactions and transmission dynamics of West Nile virus
.
Viruses
5
:
3021
3047
.
Collins
JM,
Paxton
CH,
Wahl
T,
Emrich
CT.
2017
.
Climate and weather extremes
.
In:
Chassignet
EP,
Jones
JW,
Misra
V,
Obeysekera
J,
eds.
Florida's climate: changes, variations, & impacts
.
Gainesville, FL
:
Florida Climate Institute
.
Cupp
EW,
Klingler
K,
Hassan
HK,
Viguers
LM,
Unnasch
TR.
2003
.
Transmission of eastern equine encephalomyelitis virus in central Alabama
.
Am J Trop Med Hyg
68
:
495
500
.
Curren
EJ,
Lehman
J,
Kolsin
J,
Walker
WL,
Martin
SW,
Staples
JE,
Hills
SL,
Gould
CV,
Rabe
IB,
Fischer
M,
Lindsey
NP.
2018
.
West Nile virus and other nationally notifiable arboviral diseases—United States, 2017
.
Morb Mortal Wkly Rep
67
:
1137
1142
.
Day
JF,
Shaman
J.
2011
.
Mosquito-borne arboviral surveillance and the prediction of disease outbreaks
.
In:
Ruzek
D,
ed.
Flavivirus encephalitis
.
Rijeka, Croatia: InTech.
p
105
130
.
Downs
J,
Vaziri
M,
Deskins
G,
Kellner
W,
Miley
K,
Unnasch
TR.
2020
.
Optimizing arbovirus surveillance using risk mapping and coverage modelling
.
Ann GIS
26
:
13
23
.
Edman
JD,
Webber
LA,
Kale
HW.
1972
.
Host-feeding patterns of Florida mosquitoes II. Culiseta
.
J Med Entomol
9
:
429
434
.
Environmental Systems Research Institute.
2021
.
ArcGIS Pro 2.7.0
.
Redlands, CA
:
ESRI Inc
.
FDOH [Florida Department of Health].
2019
.
Non-human mosquito-borne disease monitoring activities. Mosquito-Borne Disease Guidebook
.
Tallahassee, FL
:
Division of Disease Control and Health Protection
.
Feemster
RF.
1938
.
Outbreak of encephalitis in man due to the eastern virus of equine encephalomyelitis
.
Am J Public Health Nations Health
28
:
1403
1410
.
Kelen
PTV,
Downs
JA,
Burkett-Cadena
ND,
Ottendorfer
CL,
Hill
K,
Sickerman
S,
Hernandez
J,
Jinright
J,
Hunt
B,
Lusk
J,
Hoover
V,
Armstrong
K,
Unnasch
R,
Stark
L,
Unnasch
T.
2012
.
Habitat associations of eastern equine encephalitis transmission in Walton County Florida
.
J Med Entomol
49
:
746
756
.
Kelen
PV,
Downs
JA,
Unnasch
T,
Stark
L.
2014
.
A risk index model for predicting eastern equine encephalitis virus transmission to horses in Florida
.
Appl Geogr
48
:
79
86
.
Klenk
K,
Snow
J,
Morgan
K,
Bowen
R,
Stephens
M,
Foster
F,
Gordy
P,
Beckett
S,
Komar
N,
Gubler
D,
Bunning
M.
2004
.
Alligators as West Nile virus amplifiers
.
Emerg Infect Dis
10
:
2150
2155
.
Lindsey
NP,
Staples
JE,
Fischer
M.
2018
.
Eastern equine encephalitis virus in the United States, 2003-2016
.
Am J Trop Med Hyg
98
:
1472
1477
.
Magnarelli
LA.
1977
.
Host feeding patterns of Connecticut mosquitoes (Diptera: Culicidae)
.
Am J Trop Med Hyg
26
:
547
552
.
Main
MB,
Allen
GM.
2007
.
The Florida environment: an overview
.
Gainesville, FL
:
University of Florida
.
No. WEC 229.
McDonald
E,
Martin
SW,
Landry
K,
Gould
CV,
Lehman
J,
Fischer
M,
Lindsey
NP.
2019
.
West Nile virus and other domestic nationally notifiable arboviral diseases—United States, 2018
.
Morb Mortal Wkly Rep
68
:
673
678
.
Miley
KM,
Downs
J,
Beeman
SP,
Unnasch
TR.
2020
.
Impact of the Southern Oscillation Index, temperature, and precipitation on eastern equine encephalitis virus activity in Florida
.
J Med Entomol
57
:
1604
1613
.
Miley
KM,
Downs
J,
Burkett-Cadena
ND,
West
RG,
Hunt
B,
Deskins
G,
Kellner
B,
Fisher-Grainger
S,
Unnasch
RS,
Unnasch
TR.
2021
.
Field analysis of biological factors associated with sites at high and low to moderate risk for eastern equine encephalitis virus winter activity in Florida
.
J Med Entomol
58
:
2385
2397
.
Molaei
G,
Oliver
J,
Andreadis
TG,
Armstrong
PM,
Howard
JJ.
2006
.
Molecular identification of blood-meal sources in Culiseta melanura and Culiseta morsitans from an endemic focus of eastern equine encephalitis virus in New York
.
Am J Trop Med Hyg
75
:
1140
1147
.
Rossi
SL,
Ross
TM,
Evans
JD.
2010
.
West Nile virus
.
Clin Lab Med
30
:
47
65
.
Shaman
J,
Day
JF,
Stieglitz
M,
Zebiak
S,
Cane
M.
2004
.
Seasonal forecast of St. Louis encephalitis virus transmission, Florida
.
Emerg Infect Dis
10
:
802
809
.
Shroyer
DA,
Rey
JR.
1990
.
Saint Louis Encephalitis: A Florida problem
.
Gainesville, FL
:
University of Florida Institute of Food and Agricultural Sciences
.
Smithburn
KC,
Hughes
TP,
Burke
AW,
Paul
JH.
1940
.
A neurotropic virus isolated from the blood of a native of Uganda
.
Am J Trop Med
s1–20:471–492.
Tabachnick
WJ.
Florida's state support budget for mosquito control: tough times may undermine Florida public health [Internet]
.
Gainesville, FL
:
Florida Medical Entomology Laboratory [accessed September 5
,
2021]
.
Turell
MJ,
Dohm
DJ,
Sardelis
MR,
O'Guinn
ML,
Andreadis
TG,
Blow
JA.
2005
.
An update on the potential of North American mosquitoes (Diptera: Culicidae) to transmit West Nile virus
.
J Med Entomol
42
:
57
62
.
USCB [United States Census Bureau].
2013
.
TIGER/line shapefile—Florida, primary and secondary roads state-based shapefile
[Internet].
Washington, DC
:
US Department of Commerce
USCB [United States Census Bureau].
2010
.
State area measurements and internal point coordinates
[Internet].
Washington, DC
:
US Department of Commerce
USCB [United States Census Bureau].
2019
.
TIGER/line shapefile, U.S., current state and equivalent national shapefile
[Internet].
Washington, DC
:
US Department of Commerce
University of Florida Geoplan Center.
Florida digital elevation model (DEM) mosaic—5-meter cell size [Internet]
.
Gainesville, FL
:
University of Florida [accessed January 23
,
2021]
.
Vaidyanathan
R,
Edman
JD,
Cooper
LA,
Scott
TW.
1997
.
Vector competence of mosquitoes (Diptera:Culicidae) from Massachusetts for a sympatric isolate of eastern equine encephalomyelitis virus
.
J Med Entomol
34
:
346
352
.
Watson
JT,
Pertel
PE,
Jones
RC,
Siston
AM,
Paul
WS,
Austin
CC,
Gerber
SI.
2004
.
Clinical characteristics and functional outcomes of West Nile fever
.
Ann Intern Med
141
:
360
365
.
White
G,
Ottendorfer
C,
Graham
S,
Unnasch
TR.
2011
.
Competency of reptiles and amphibians for eastern equine encephalitis virus
.
Am J Trop Med Hyg
85
:
421
425
.

Author notes

1

Center for Global Health Infectious Disease Research, University of South Florida, 3720 Spectrum Boulevard, Tampa, FL 33612.

2

School of Geosciences, University of South Florida, 4202 E Fowler Avenue, Tampa, FL 33620.