Activity Patterns of St. Louis Encephalitis Virus (Flaviviridae) in Mosquito Communities in a Subtropical Region of Argentina

ABSTRACT.

Our aim with this study is to analyze the temporal and spatial dynamics of St. Louis encephalitis (SLEV) in mosquitoes collected from urban and rural environments in the central region of Chaco Province in the subtropical region of Argentina. Mosquitoes were captured between 2012 and 2015. A total of 81 mosquito pools were analyzed (2,105 mosquitoes) for SLEV detection, and 41 of these pools tested positive for SLEV (n = 1,675 individuals; 50%). Thirteen mosquito species were found to be infected with SLEV during this study, nine of which were reported infected for the first time in Argentina. The virus activity was recorded throughout the study period, both in urban and rural environments. The mosquito species in both environments that were infected with SLEV included Culex maxi, Culex brethesi/eduardoi, Culex bidens, and Aedes scapularis. Similar species composition and marked fluctuation in monthly activity were observed for species of the genus Culex between environments. Likewise, a similar seasonal pattern of infected species with a higher frequency of positive SLEV results was recorded during Autumn (19/41) in both environments. Our study confirms the active circulation of SLEV throughout the analyzed period in several mosquito species and its generalist nature in terms of the vectors it can infect in a subtropical region.

INTRODUCTION

Arboviruses are arthropod-borne viruses that represent a worldwide public health concern. Some of them cause emerging and reemerging diseases in America, such as dengue, chikungunya, Zika virus, yellow fever, St. Louis encephalitis (SLEV), and West Nile virus.^1–4 St. Louis encephalitis was identified for the first time during an outbreak in 1933 in St. Louis, Missouri. This virus can cause neuroinvasive diseases in humans, such as meningitis or fatal encephalitis; however, most infections are asymptomatic or present as a febrile condition.⁵ St. Louis encephalitis is present exclusively in America^6–9 and has been active in several countries throughout North and South America.² In North America, the urban transmission network of SLEV involves passerine and columbiform birds and mosquitoes of the species Culex pipiens pipiens and Culex quinquefasciatus (Cx. quinquefasciatus). However, in the rural environment, a wide variety of passerine birds and mosquitoes of the species Culex tarsalis and Culex nigripalpus are involved in transmission.⁷ In South America, SLEV has been detected in mosquitoes in several countries, such as Colombia, Panama, Argentina, and Brazil.^10–17

This virus emerged as a human pathogen in Córdoba Province in the central area of Argentina in 2003 and 2005, causing the first human encephalitis outbreak in the country.^18,19 Recently, the virus reemerged in the western United States, likely via transmission from Argentina.² The transmission network described for SLEV in the center area of Argentina involves Cx. quinquefasciatus, Culex interfor (Cx. interfor), and Culex saltanensis (Cx. saltanensis) mosquito species as urban vectors and Zenaida auriculata and Columbina picui doves as vertebrate hosts.^2,20–22 Since the reemergence of the virus in 2002, Argentina has reported multiple outbreaks of SLEV between 2005 and 2012.^19,23,24 Despite its endemicity in Argentina, little is known about its activity in the subtropical region. Our aim with this study is to describe the temporal and spatial activity of SLEV in urban and rural environments. We also aim to characterize mosquito communities in subtropical environments with different degrees of human intervention in Pampa del Indio City, Chaco Province, in northeastern Argentina.

MATERIALS AND METHODS

Study area.

This study was conducted in Pampa del Indio (26°02’ 54″S, 59°56′ 41″W, 96; Figure 1), a locality that belongs to the Department of Libertador General San Martín, in the northeastern part of Chaco Province, Argentina. The climate of this region is humid subtropical (with a dry season) and belongs to the Parque Chaqueño Oriental.²⁵ Rainfall mainly occurs during the warm climatic season, between November and May, with an average rainfall of 1,280 mm, whereas, during the cold season (between June and September), the average rainfall is 650 mm.^26,27 The average temperature of the warmest month (January) is 27.1°C, with an absolute maximum of 44.4°C, and the average temperature of the coldest month (July) is 16.1°C, with an absolute minimum of –5°C. Daily thermal oscillation is significant, and 10°C differences are common.²⁸ The vegetation has a polymorphic physiognomy, with forests, shrubby and herbaceous steppes, savannahs, palm forests, and grasslands. Soil permeability is moderate or moderately slow, and surface runoff originates from flooded areas, where soil characteristics hinder runoff during periods of excess rain because of a lack of soil infiltration capacity and infrastructure works (such as the construction of roads) with insufficient drainage.²⁵

Figure 1.

(A) Map of Argentina showing the Province of Chaco. (B) Map of the Province of Chaco showing the location of Pampa del Indio. (C) Urban plan of Pampa del Indio showing the study sites (triangles: rural sites; circles: urban sites). (D) Photographs of the studied environments.

Sampling sites.

Ten sampling sites were selected: four rural and six urban sites. The rural sites represented largely undisturbed natural environments (lagoon edges and patches of natural forest) located up to 150 m away from houses that have pens for cows, goats, and pigs, and located ∼5 km and 15 km away from the urban area. Human settlements at these rural sites must store water for domestic, health, agricultural, and other uses because there are no potable water pipes. The urban sites corresponded to the dwellings located within the urban area. Only 25% of the streets in the urban area are paved, the population has drinking water and trash collection system, and there is a hospital (Figure 1).

Adult mosquito collection.

Adult mosquitoes were collected monthly during periods of higher temperatures and precipitation (October 2012 to May 2013, October 2013 to May 2014, and October 2014 to May 2015). Mosquitoes were collected using 10 CDC light traps (one for each collection site) supplemented with dry ice. The traps remained active between 6 pm and 9 am. The specimens were live-transferred to the laboratory in refrigerated containers, which were then stored at –70°C until processing. Specimen identification was conducted under a stereomicroscope on a chill table, according to the characteristics outlined in Darsie and Mitchell,²⁹ Consoli and Lourenço de Oliveira,³⁰ Forattini,³¹ and Faran and Linthicum.³² Mosquitoes were pooled by species, sex, sample date, and sample site in lots of 1 to 50 specimens and stored at –70°C.

Mosquito pool homogenization and SLEV molecular detection.

Mosquito pools were homogenized in sterile mortars using 1,000 µL of minimal essential medium (MEM) for pools of up to 25 mosquitoes and 2000 µL of MEM for pools of up to 50 mosquitoes. Later, centrifugation was performed at 10,000 revolutions per minute for 30 minutes to obtain supernatants. Viral RNA was extracted from mosquito homogenized supernatants by using commercial kits (Axygen Bioscience, Union City, CA; Fermentas, Waltham, MA; and Roche, Basel, Switzerland), according to the manufacturers’ instructions. Copy DNA was obtained via retrotranscription using an Improm II kit (Promega, Madison, WI) and random hexamer primer (Promega).

The specific detection of SLEV in mosquito pools was performed by using nested reverse transcription polymerase chain reaction (RT-PCR) which amplifies a 234-base pair (bp) fragment from the envelope protein region.³³

In the nested RT-PCR used for the detection of specific SLEV, the degenerate primers used included Encafilitis de Saint Louis (SLE 1497) ((+):5′1497 rryatgggygagtatggracag 1518 3′) and Encafilitis de Saint Louis (SLE 2517) ((–): 5′2496 ctcctccacayttyarttcacg 2517 3′), which amplified 999 bp of a fragment corresponding to the NSI and E gen (E) genes. For the nested RT-PCR, the primers used were Encafilitis de Saint Louis (SLE 2002) ((+): 5′2002 tggaytggacrccggttggaag 2003 3′) and Encafilitis de Saint Louis (SLE 2257) ((–): 5′2236 ccaatrgatccraartcccacg 2257 3′), which amplify 234 pb of the E gene.³³ A final volume of 25 µL of the mixture was subjected to the following cycling program: denaturation at 94°C for 2 minutes; annealing at 94°C for 30 minutes, –55°C for 30 minutes, and –72°C for 1 hour and 30 minutes (for 40 cycles); and extension at 72° for 7 minutes. For the second amplification, 1 µL of the previous mixture (first round of polymerase chain reaction [PCR]) was added to a new mixture with a total volume of 24 µL, maintaining equal concentrations and volumes of its reagents, except for water, which was calculated at 14.35 µL, and the enzyme, which was calculated at 1.25 µL. This resulted in a a final volume of 25 µL that was subjected to the following cycling program: denaturation at 94°C for 2 minutes; annealing at 94° for 30 minutes, –63° for 30 minutes, and –72° for 1 minute (for 40 cycles); and extension at 72° for 7 minutes. The PCR products were revealed in a 2% agarose gel. The positive amplified fragments were sequenced using the commercial sequencing service at Macrogen Inc. in Seoul, South Korea, for the identification and molecular characterization of the viral agent. The obtained sequences were edited using the Molecular Evolutionary Genetics Analysis version 7.0 (MEGA7; Dortmund, Germany) program.³⁴ They were subjected to BLASTn (Basic Local Nucleotide Alignment Search Tool) analysis to corroborate their identity.³⁵ Part of the positive controls used for molecular detection were provided by the Arbovirus Laboratory of the Dr. J. M. Vanella Virology Institute at the National University of Córdoba.

Analysis of mosquito community data.

The mosquito community was analyzed by calculating the nonparametric richness estimator; the abundance-based coverage estimator,³⁶ assuming that the inventories may be incomplete;³⁷ and the completeness of the samples for each environment and community. The Kruskal–Wallis test was used to compare the abundances of mosquitoes and the climatic seasons studied (Infostat Software, Infostat Systems, Inc., Sacramento, CA).³⁸ Alpha diversity was estimated by using the species richness and the method proposed by Jost,³⁹ who introduced the term true diversity. In this study, the true diversity used was of the order 2” (qD = 2), where q is the order of diversity and defines the sensitivity of the index to the relative abundances of species. Its value determines the degree to which common species or rare species influence the measure of diversity, and it can take any value that the user deems appropriate (Hill 1973). The D value represents true diversity, and the exponent determines the sensitivity of the index to the relative abundances of the species, specifically, the influence that common or rare species can have on the measure of diversity.^37,39–42 The difference between the Culicidae communities was also expressed as the ratio of the most diverse to the least diverse, as well as the percentage of diversity represented by the community with the lowest diversity compared to that with the highest diversity.³⁷ Beta diversity between environments was estimated using the complementarity index (CAB).⁴³

The data on the climatic variables analyzed in this study were obtained from the National Meteorological Service and recorded by the meteorological station located in Presidencia Roque Sáenz Peña (Chaco), the closest city to the study area.

To determine the association between the abundances of mosquito species and meteorological variables, we considered a time interval of daily records and the monthly averages for each season. We used a regression model with a negative binomial response, including climate–species interactions as regressors. Because of the potential dependence among the errors of the model, we estimated standard errors considering the cluster structure of observations for each species. This model allowed for the estimation of the effects of each meteorological variable on each species, expressed as the incidence rate ratio (IRR), along with its P-value (significance value). It also allowed us to model the association of the variables with respect to a reference species selected for having the lowest coefficient of variation (standard deviation over the mean) and for not exhibiting atypical observations using Stata software, version 14 (StataCorp., College Station, TX).⁴⁴ Incidence rate ratio values equal to 1 indicate no influence of the climatic variables on the mosquito species abundances; values lower than 1 indicate a negative association between the variables and the abundances; and values greater than 1 indicate a positive association between the variables and the abundances.

All the species exhibited positive or negative statistically significant associations (P <0.01) with the meteorological variables analyzed (Table 1). Aedes aegypti (Ae. aegypti), Culex bidens (Cx. bidens), Culex chidesteri, Culex eduardoi, Culex maxi, Mansonia titillans, Psorophora paulli, and Anopheles triannulatus were correlated positively and significantly with the average temperature. The incidence rate ratio (IRR) revealed that a 1°C increase in temperature resulted in a 20% increase in Ae. aegypti abundance (IRR = 1.207), an increase of 41% in Cx. bidens abundance (IRR = 1.413), and a decrease of 87% in Culex quinquefasciatus (Cx. quinquefasciatus) abundance (IRR = 0.879; Table 1). Precipitation positively and significantly influenced (IRR >1) the abundances of species of the Aedes and Psorophora genera, as well as Cx. quinquefasciatus abundance (Table 1).

Table 1.

Relationship between the meteorological variables and the abundances of mosquitoes captured in Pampa del Indio, Chaco Province, between October 2012 and May 2015

Species	IRR Values
	Temperature (°C)	Relative Humidity	Precipitation (mm)
Aedes aegypti	1.207	1.139	1
Ae. scapularis	0.871	1.026	1.011
An. (Ano.) spp.	0.576	0.758	0.977
An. (Nys.) spp.	1.438	1.226	0.963
Cx. (Cux.) bidens	1.413	1.079	0.99
Cx. (Cux.) chidesteri	1.292	1.134	0.987
Cx. (Cux.) eduardoi	1.552	1.357	0.991
Cx. (Cux.) maxi	1.262	1.132	0.992
Cx. (Cux.) quinquefasciatus	0.878	1	1.004
Cx. (Cux.) spp	1.228	1.119	0.996
Ma. spp.	1.061	0.933	0.995
Ma. titillans	1.11	0.936	0.985
Ps. (Jan.) cyanescens	0.909	0.768	1.051
Ps. (Grab.) paulli	1.169	0.863	1.054
An. (Nys.) triannulatus	1.369	0.968	0.991

Ae. scapularis = Aedes scapularis; Ae. aegypti = Aedes aegypti; An. triannulatus = Anopheles triannulatus; Cx. bidens = Culex bidens; Cx. chidesteri = Culex chidesteri; Cx. eduardoi = Culex eduardoi; Cx. maxi = Culex maxi; Cx. quinquefasciatus = Culex quinquefasciatus; IRR = incidence rate ratio; Ma. titillans = Mansoni titillans; Ps. paulli = Psorophora paulli; Ps. cyanescens = Psorophora cyanescens. Reference = IRR relative risk value index. The three climatic variables exhibited P-values <0.01, indicating that their associations with all the analyzed species were significant.

Analysis of SLEV activity.

To evaluate the associations between the prevalence of SLEV and the type of environment (rural or urban) and climatic seasons, bivariate association tests were conducted. This test consisted of a comparison of proportions between two samples (two-sample test of proportions = Z-test), which was used to evaluate the association between positivity for SLEV and the two environments and arcross two seasons (spring and autumn).^44,45

The positivity analysis for the SLEV infection, which was used to evaluate the effect of the species on SLEV positivity, was performed using logistic regression and the standard hypothesis tests for the significance of the regressors.

For the test, groups of pools of different species were formed and compared with a base group, which for this study consisted of up to 14 species with fewer positive pools (fewer than 7).

The probability of greater or lesser infection relative to the base group was evaluated using the odds ratio, with an odds ratio >1 indicating a protective factor or positive effect (greater probability of infection with SLEV) and an odds ratio <1 indicating a negative effect (lesser probability of infection with SLEV).

RESULTS

Mosquito community.

After 24 collections were made in the field, a total of 14,034 mosquitoes belonging to 47 species distributed in 10 genera, were identified. Culex was the most abundant genus (50.93%), followed by Mansonia (23.70%), Psorophora (10.13%), Anopheles (7.31%), and Aedes (7.13%; Table 2). The genus that made the greatest contribution to the species richness was Culex, with 17 species, with the subgenus of Culex contributing nine species, Melanoconion contributing six species, and Microculex contributing two species. The genus Psorophora was represented by nine species, Anopheles and Aedes contributed six species each, Mansonia and Uranotaenia contributed three species each, and the remaining genera included one or two species each (Table 2). Numerous female specimens of the genus Culex could not be clearly identified at the species level because of a lack of diagnostic characteristics; thus, they were named as Culex spp.

Table 2.

Total abundance of mosquitoes collected in Pampa del Indio, Chaco, Argentina, between October 2012 and May 2015

Mosquito Species	Environment
	Urban	Rural	Total	% Total
Aedeomyia (Aed.) squamipennis	24	23	47	0.3
Aedes (Och.) spp.	3	18	21	0.2
Aedes (Stegomyia) aegypti	61	19	80	0.6
Aedes (Och.) albifasciatus	3	–	3	0.0
Aedes (Och.) fulvus		2	2	0.0
Aedes (Och.) hastatus	3	9	12	0.1
Aedes (Och.) scapularis	246	629	875	6.2
Aedes (Och.) stigmaticus	1	2	3	0.0
Anopheles (Ano.) evandroi	–	3	3	0.0
Anopheles (Ano.) spp.	–	325	325	2.3
Anopheles (Nys.) spp.	20	536	556	4.0
Anopheles (Nys.) albitarsis	8	19	27	0.2
Anopheles (Nys.) benarochi	–	5	5	0.0
Anopheles (Nys.) deaneorum	3	4	7	0.1
Anopheles (Nys.) evansae	1	3	4	0.0
Anopheles (Nys.) triannulatus	9	91	100	0.7
Coquilletttidia (Rhyn.) nigricans	–	9	9	0.1
Culex (Mel.) ocosa	11	13	24	0.2
Culex (Cx.) spp.	1764	2001	3765	26.8
Culex (Mel.) spp.	1		1	0.0
Culex (Cx.) ameliae	17	18	35	0.3
Culex (Cx.) bidens	342	505	847	6.0
Culex (Cx.) brethesi/eduardoi	248	576	824	5.9
Culex (Cx.) chidesteri	105	662	767	5.5
Culex (Mel.) clarki	–	4	4	0.0
Culex (Cx.) coronator	4		4	0.0
Culex (Mcx.) davisi	–	2	2	0.0
Culex (Mel.) delpontei	–	7	7	0.1
Culex (Mel.) educator	1	6	7	0.1
Culex (Cx.) heperi		1	1	0.0
Culex (Mcx.) imitator	3	8	11	0.1
Culex (Mel.) intrincatus	2	7	9	0.1
Culex (Cx.) maxi	291	194	485	3.5
Culex (Cx.) mollis	3	–	3	0.0
Culex (Mel.) pilosus	26	–	26	0.2
Culex (Cx.) quinquefaciatus	281	41	322	2.3
Hemagogus spp.	1	–	1	0.0
Mansonia (Ma.) indubitans	–	2	2	0.0
Mansonia (Ma.) humeralis	2	12	14	0.1
Mansonia (Ma.) spp.	30	386	416	3.0
Mansonia (Ma.) titillans	68	2827	2895	20.6
Psorophora (Grab.) spp.	7	17	24	0.2
Psorophora (Jan.) spp.	38	14	52	0.4
Psorophora (Ps.) spp.	–	3	3	0.0
Psorophora (Ps.) ciliata	1	–	1	0.0
Psorophora (Ps.) cingulata	1	–	1	0.0
Psorophora (Grab.) confinnis	8	39	47	0.3
Psorophora (Jan.) cyanescens	160	497	657	4.7
Psorophora (Grab.) dimidiata	9	13	22	0.2
Psorophora (Jan.) ferox	–	3	3	0.0
Psorophora (Ps.) pallescens	9	1	10	0.1
Psorophora (Grab.) paulli	127	401	528	3.8
Psorophora (Grab.) varinervis	67	9	76	0.5
Uranotaenia (Ur.) apicalis	4	–	4	0.0
Uranotaenia (Ur.) davisi		1	1	0.0
Uranotaenia (Ur.) lowii	5	25	30	0.2
Uranotaenia (Ur.) spp.	–	5	5	0.0
Wyeomyia (Pho.) spp.	–	1	1	0.0
Wyeomyia (Pho.) muehlensi	–	8	8	0.1
Total	4358	9678	14,034	100

Aed. = Aedeomyia; Ano. = Anopheles; Cx. = Culex; Grab. = Grabhamia; Jan. = Janthinosoma; Ma. = Mansonia; Mcx. = Mircroculex; Mel. = Melanoconion; Nys. = Nyssorhynchus; Och. = Ochlerotatus; Pho. = Phoniomya; Ps. = Psorophora; Rhyn. = Rhynchotaenia; Ur. = Uranotaenia.

For the calculation of diversity, only the taxonomically determined species were considered (N = 47). The global estimate of the number of species for the entire study revealed an average completeness of 90.4% (mean = 52). Regarding the estimation of species richness by environment, in the rural area, a completeness of 93% (mean = 44.2) was revealed, and in the urban area, a completeness of 93% (mean = 36.5) was revealed. Species richness and the abundance of individuals were greater in the rural environment (S = 41; 6,566 individuals) compared with the urban environment (S = 34; 2,449 individuals).

On the other hand, regarding true diversity, it was observed that in the rural environment, the common species exhibit less equality than the common species at the urban site. This finding is reflected in the results of the true diversity analysis, which revealed that the urban environment was twice as diverse (²D  =  8.9 effective species) as the rural environment (²D  =  4.4 effective species). Beta diversity (CAB) exhibited 42% complementarity (0.42 index) between the environments, indicating that the composition of species is dissimilar, with 28 shared species out of a total of 47. On the other hand, the urban environment has eight exclusive species, whereas the rural environment has 17.

Temporal distribution of Culicidae.

The Kruskal–Wallis test revealed statistically significant differences (H = 22.34; P = 0.001) in species abundances across climatic seasons, with Autumn and Spring 2014 showing the greatest abundances.

Culex chidesteri (Cx. chidesteri), Culex brethesi/eduardoi (Cx. brethesi/eduardoi), Culex bidens (Cx. bidens), and Culex maxi (Cx. maxi) were captured throughout the sampling period, with marked fluctuations in their abundances, reaching peaks in 1 or 2 months (Cx. bidens: Sumer and Spring 2013, Cx. chidesteri and Cx. brethesi/eduardoi: Spring 2013 and Autumn 2014, Cx. maxi: Spring 2013) and then reaching low or nearly zero levels for 1 or 2 months. Culex quinquefasciatus was also captured over the sampling period but did not exhibit marked fluctuations. The greatest abundances of this species were detected in Spring 2013 and Autumn 2014 (Figure 2).

Figure 2.

Seasonal mosquito abundance between October 2012 and May 2015.

Mansonia titillans was also found throughout the sampling period, with one peak of abundance in October 2014 (Spring). Aedes scapularis and Ae. aegypti were captured throughout the study period, showing their highest abundances between March and April 2014 (Autumn) (Figure 3). Psorophora paulli showed greater abundances in February, March and November 2014, whereas Ps. cyanescens presented a single peak of abundance in November 2014 (Figure 3).

Figure 3.

Seasonal mosquito abundance between October 2012 and May 2015.

St. Louis encephalitis activity in mosquito communities.

The mosquitoes (N = 2,105) were grouped into 81 pools and tested for SLEV using RT-PCR. Of these pools, 41 tested positive (n = 1,675 individuals; 50%); 22 positive pools (53.65%) corresponded to mosquitoes captured in the urban environment, and 19 pools (46.34%) corresponded to mosquitoes from the rural environment (Table 3). Thirteen mosquito species were found to be infected with SLEV during this study, nine of which were reported as infected for the first time in Argentina: Cx. bidens, Cx. brethesi/eduardoi, Cx. chidesteri, Culex imitator (Cx. imitator), Cx. maxi, Aedes stigmaticus (Ae. stigmaticus), Psorophora confinnis (Ps. confinnis), Psorophora ciliata (Ps. ciliata), and Mansonia humeralis (Ma. humeralis; Table 3). Infected Aedes aegypti (Ae. aegypti), Cx. imitator, and Cx. quinquefasciatus species were only detected in the urban environment, whereas infected Ae. stigmaticus, Ps. confinnis, and Ps. ciliata were only detected in the rural environment.

Table 3.

Positive mosquito pools for St. Louis encephalitis sorted by species, collection date, climatic season, and environment studied in Pampa del Indio, Chaco Province, Argentina, between 2012 and 2015

Species	Number of Mosquitoes	Month/Year	Climatic Season	Environment
Aedes aegypti	2	November 2014	Spring	Urban
Aedes scapularis	2	April 2014	Autumn	Urban
Aedes scapularis	8	October 2014	Spring	Rural
Aedes stigmaticus *	1	December 2013	Spring	Urban
Anopheles spp.	4	April 2014	Autumn	Rural
Culex bidens *	6	April 2013	Autumn	Rural
Culex bidens	53	December 2013	Spring	Urban
Culex bidens	23	April 2014	Autumn	Rural
Culex bidens	52	October 2014	Spring	Rural
Culex chidesteri *	3	April 2013	Autumn	Rural
Culex chidesteri	10	February 2014	Summer	Rural
Culex chidesteri	1	April 2014	Autumn	Rural
Culex chidesteri	48	October 2014	Spring	Rural
Culex eduardoi/brethesi *	29	April 2013	Autumn	Rural
Culex eduardoi/brethesi	12	December 2013	Spring	Urban
Culex eduardoi/brethesi	10	April 2014	Autumn	Urban
Culex imitator *	1	December 2012	Spring	Urban
Culex imitator	1	October 2013	Spring	Urban
Culex maxi *	3	February 2014	Summer	Rural
Culex maxi	27	April 2014	Autumn	Urban
Culex maxi	1	May 2014	Autumn	Urban
Culex maxi	10	October 2014	Spring	Rural
Culex quinquefasciatus	9	October 2013	Spring	Urban
Culex quinquefasciatus	6	May 2014	Autumn	Urban
Culex quinquefasciatus	1	October 2014	Spring	Urban
Culex spp.	2	May 2013	Autumn	Urban
Culex spp.	67	December 2013	Spring	Urban
Culex spp.	19	January 2014	Summer	Urban
Culex spp.	28	February 2014	Summer	Rural
Culex spp.	11	April 2014	Autumn	Rural
Culex spp.	10	April 2014	Autumn	Urban
Culex spp.	5	April 2014	Autumn	Urban
Culex spp.	54	October 2014	Spring	Rural
Mansonia humeralis *	1	November 2014	Spring	Rural
Mansonia titillans	2	April 14	Autumn	Urban
Mansonia titillans	50	October 2014	Spring	Rural
Psorophora ciliata *	1	April 2014	Autumn	Rural
Psorophora confinnis *	4	April 2014	Autumn	Rural
Total	577	–	–	–

^*Species reported infected for St. Louis encephalitis for the first time in Argentina.

The highest number of positive pools belonged to Cx. bidens (five pools; n = 223 individuals), followed by Cx. chidesteri, Cx. maxi, and Cx. quinquefasciatus (four pools for each species), corresponding to 261, 73, and 47 tested individuals, respectively. Aedes aegypti also tested positive for SLEV (one pool = 21 individuals; Table 3). In the rural environment, the highest number of positive pools (9/19) was recorded during the spring (October 2014), and the mosquito species that tested positive for SLEV included Cx. bidens, Cx. chidesteri, Cx. brethesi/eduardoi, Ps. ciliata, and Ps. confinnis (Table 3). In the same year, the other positive pools (7/19) were recorded in the autumn (April 2014), corresponding to the Cx. bidens, Cx. chidesteri, Cx. maxi, Ae. scapularis, Mansonia titillans (Ma. titillans), and Ma. humeralis species, followed by Autumn 2013 (3/19), corresponding to the Cx. bidens, Cx. chidesteri, Cx. brethesi/eduardoi, and Cx spp. species. Finally, in the summer (February 2014; 3/19), pools corresponding to Cx. chidesteri and Cx. maxi tested positive for SLEV (Table 3).

In the urban environment, the highest number of positive pools for SLEV (10/22) was recorded during the spring (December 2013) and autumn (April 2014), with Cx. maxi, Cx. brethesi/eduardoi, Ae. scapularis, and Ma. titillans testing positive in the autum and Cx. bidens, Cx. brethesi/eduardoi, and Ae. stigmaticus testing positive in the spring. Finally, in the summer, we recorded two positive pools corresponding to Ae. aegypti (January 2013 and March 2015) and one positive pool corresponding to an undetermined Culex species (Table 3). A bivariate test used to analyze the association between the prevalence of SLEV infection and the type of environment (z = 0.36; P-value 0.719, <0.05), as well as the association between SLEV infection and the spring and summer seasons (z = 1.01; P-value 0.31, <0.05) did not reveal significant differences.

In addition, the odds ratio test revealed that the presence of Cx. bidens (99%) and Cx. chidesteri (90%) increased their probability of infection with SLEV, regardless of the environment or climatic season.

DISCUSSION

The ecology of SLEV has been poorly studied in South America, especially in subtropical and tropical areas. In Argentina, most studies have been performed to examine its activity in temperate areas.^{2,14–16, 21,45,46} In our region, SLEV activity was recorded throughout the study (over three periods: October–May 2012/2013; October–May 2013/2014, October–May 2014/2015), both in urban and rural environments, indicating its endemicity in the study area. The percent positivity found in the processed pools was high (50% = 41/81) for a significant number of mosquito species from a small geographical region, considering both types of environments.

As expected in subtropical areas, no statistically significant difference was detected in the spatial and seasonal activity, indicating similar levels of viral activity between rural and urban sites and across climatic seasons. These results could indicate the presence of equally vector-competent mosquito species in both types of environments (urban and rural) and stable mosquito population abundance, particularly during summer and autumn. A different activity pattern was observed in temperate areas (the city of Cordoba), where SLEV activity in the avian host community varied among years and climatic seasons.⁴⁷ The same pattern was observed in the mosquito communities, probably indicating some mechanism for the annual introduction of SLEV in the city.¹⁴ In contrast to this temperate scenario, our data suggest that subtropical urban and rural areas of Chaco province provide all biological requirements that SLEV needs to remain active throughout the year. The temporal abundance observed for several potential SLEV mosquito vector species (Cx. quinquefasciatus, Cx. bidens, and Cx. maxi) indicates that mosquito abundance is maintained during autumn and possibly during the winter season, as well. This mosquito population dynamic would allow for the remnant vector transmission of SLEV during winter, thereby enabling its endemicity in subtropical areas. This hypothesis should be tested by applying molecular characterization of circulating SLEV strains to look for genetic identification and seasonal stability of circulating genotypes. In this study, 27.65% of the total species were found to be infected with SLEV; because there is no record of disease in vertebrate hosts, it is possible that mosquitoes harbor the virus temporarily after a blood meal at the time of capture.

Culex quinquefasciatus has been studied as vector of SLEV in Argentina^{11,12,20–22} and proposed to be a vector in Guatemala.⁴⁸ Several authors have mentioned this species as the most abundant and infected in studies conducted in temperate regions of Argentina.^11,12,14,15 Almirón and Brewer⁴⁹ recorded a pattern for this species that indicates peak abundance in summer (February–March) and in some periods in spring (November–December) in the city of Córdoba (temperate region). Diaz et al.¹⁴ detected six pools of this species infected with SLEV (n = 20 positive results out of 315 pools analyzed) in three sites with different degrees of urbanization in the city of Córdoba. In the present study, Cx. quinquefasciatus was found in both the urban and rural environments, but SLEV infection was only detected in pools from the urban environment. Culex bidens was the second most abundant species in the urban environment and the third most abundant in the rural environment, making it the species with the second-highest number of positive pools. Additionally, unlike Cx. quinquefasciatus, Cx. bidens exhibited a positive and significant effect on the probability of being infected with SLEV. Based on experimental studies, Beranek et al.²⁰ concluded that both Cx. interfor and Cx. saltanensis are susceptible to SLEV infection and competent for transmission. Beranek et al.²⁰ and Díaz et al.²² found Cx. interfor infected with SLEV in Córdoba city during enzootic and epidemic periods between 2004 and 2016. Batallán et al.⁵⁰ observed that this mosquito species is more abundant in peri-urban areas with vegetation. The biology of Cx. bidens is very similar to that of Cx. interfor.^50,51 In previous studies, Stein et al.^52,53 found Cx. bidens in urban, semi-urban, and wild environments and captured this species using chicken-, rabbit-, and human-baited traps. These findings, as well as those of the present study, including this species’ high probability of infection with SLEV (99%), could be of particular interest for the region. They could also encourage further research into the characterization of the biology of Cx. bidens regarding its role as an SLEV vector in the study area. Culex chidesteri, an abundant species in the local culicid fauna, ranked fifth among the most abundant species in this study. It also tested positive for SLEV in both environments and exhibited a positive and significant effect (90%) on the probability of being infected with SLEV. This species exhibited similar activity patterns as Cx. quinquefasciatus, Cx. maxi, Cx. brethesi, and Cx. bidens, with marked fluctuation in abundances. Moreover, Cx. chidesteri has a feeding preference for avian hosts and is considered a eusynanthropic species that is adapted to urban, peri-urban, and rural environments.^51,53–55 It is necessary to increase the number of studies conducted to determine the role of this species as a vector of SLEV in this subtropical region of the country. Mansonia titillans was found to be infected with SLEV in the present study, and specimens were collected throughout the study period, in both rural and urban environments. Beranek et al.¹⁶ reported the detection of SLEV in this species in the Córdoba Province (temperate area) during the fall season, when this mosquito species is most abundant. Moreover, Ma. titillans was also abundant in mosquito collections conducted in urban, peri-urban, and forested areas in the wet Chaco region.⁵³ These findings confirm the significant adaptation of this species to different environments. Moreover, an important factor for consideration is this species’ capacity to feed on a wide variety of vertebrates (including rabbits and chickens).⁵² Aedes scapularis was found to be infected in Autumn 2014, when it was abundant. Mitchell et al.¹¹ identified a wide variety of blood sources (including humans, horses, birds [Momotidae], marsupials [Didelphide], Felidae, Bovidae, and Cervidae) in engorged Ae. scapularis mosquitoes collected in the provinces of Chaco and Corrientes, indicating this species’ potential role as bridge vector. Given that the eggs of this species are able to survive over winter periods, it could also play a role as a hibernating host of SLEV. We consider it important to perform vector competence studies involving this species. Psorophora confinnis has been found to be infected with SLEV in Pampa del Indio (in this study), although it seems to exhibit little competence for SLEV transmission.⁵⁶

CONCLUSION

In this study, we describe the activity pattern of SLEV and the mosquito community in Pampa del Indio, Chaco, in the subtropical region of Argentina. The estimated mosquito species richness (S = 47) was intermediate between that previously found by our group in the wet Chaco (S = 74) in a semi-urban environment⁵³ and in the dry Chaco (S = 21) in a rural environment.^57,58 It is noteworthy that in the urban environment of this study, we detected species that were considered exclusive to semi-urban or wild environments, such as Cx. imitator, Aedes hastatus/oligopistus, and Ae. stigmaticus in the wet Chaco.⁵³ This is a possible indicator of the scarce anthropic modification of the urban area in the studied city. Nonetheless, in the present research, both environments exhibited significant dissimilarity, with 50% shared species and a greater number of exclusive species (n = 17) in the rural environment.

The Culex species found in the present study were collected in all studied climatic seasons and were more abundant in the spring and autumn, when they were found to be infected with SLEV. Assuming that the sampling may be incomplete,⁴³ the calculation of the abundance-based richness estimator³⁷ revealed high integrity and representativeness that were the same for both environments.

The absence of clinical cases in humans and disease in animals could be due to the circulation of nonpathogenic strains of the virus. It could also indicate that the mosquito species are inefficient as vectors. Therefore, studying the food preferences of the species that were found to be positive, examining the periods of vector activity that align with those of virus occurrence, and conducting vector competition tests in the laboratory could complete this line of research.

It is also necessary to consider the possible identification of different strains of SLEV that could be circulating in the study area, as well as changes in avian host, in further studies. The results obtained in this study highlight the circulation of SLEV in Chaco Province, confirming the endemicity in the study area and indicating the possible existence of a complex transmission network involving more than one species of mosquito vector. It is also possible that these species are not the same ones involved in the circulation of the virus in temperate areas. All this information will be useful for the development of different strategies for the prevention and control of SLEV by health authorities.