Daytime Resting Activity of Aedes aegypti and Culex quinquefasciatus Populations in Northern Mexico

Mosquito-borne diseases are a major public health concern causing millions of deaths annually (Chaiphongpachara et al. 2018). Aedes aegypti (L.) is the most important global vector of arboviruses, such as dengue, chikungunya, Zika, and yellow fever (Campos et al. 2020). Although Ae. albopictus (Skuse) has been reported in Mexico (Dávalos-Becerril et al. 2019), its role as an arbovirus vector is limited (Lambrechts et al. 2010, Vasilakis et al. 2011, Mousson et al. 2012). Members of the Culex quinquefasciatus Say complex are the primary vectors of West Nile virus (WNV) throughout much of the tropical and temperate regions of the world (del Carpio-Orantes et al. 2018). In addition, they are vectors of St Louis encephalitis virus (SLEV) (Diaz et al. 2013), Usutu virus (Clé et al. 2019), and the filarial nematode Wuchereria bancrofti (Cobbold) (Ant et al. 2020).

The control of mosquito-borne viruses in Mexico and other countries receives considerable resources but returning limited success as severe outbreaks persist in tropical endemic countries worldwide including nonendemic countries such as chikungunya in Italy during 2007 and dengue in France and Croatia during 2010 (Gould et al. 2010, Gjenero-Margan et al. 2011, Poletti et al. 2011, Lwande et al. 2020). In 2020 the Mexican Ministry of Health reported 24,313 cases of dengue, 2,281 of which occurred in Tamaulipas. From 2015 to November 2022, 13,003 confirmed cases of Zika were reported in Mexico, of which 802 cases occurred in Tamaulipas (Mexico Ministry of Health 2022). The persistent transmission is in part, due to a lack of effective vaccines and development of insecticide resistance in vectors populations. Thus, effective vector control needs to be built on detailed knowledge of mosquito population ecology and behavior that has spatio-temporal heterogeneity (Wilke et al. 2019).

Many studies have addressed the resting behavior of Ae. aegypti and Cx. quinquefasciatus. However, most of them were focused on the indoor or outdoor resting sites (Hecht and Hernandez-Corzo 1963, Schoof 1967, Kay 1983, Kittayapong et al. 1997, Perich et al. 2000, Chadee 2013, Dzul-Manzanilla et al. 2017, Diallo and Diallo 2020, Machani et al. 2020, Diouf et al. 2021, Dalpadado et al. 2022, Janaki et al. 2022), resting behavior associated with residual insecticides (Bøgh et al. 1998, Cooperband and Allan 2009, Manda et al. 2011), and insecticide treatment (Pates and Curtis 2005, Tainchum et al. 2013). Very few studies have looked at the peak daytime resting activity of Ae. aegypti and Cx. quinquefasciatus male and female populations. To our knowledge, the only study about the diel resting activity of mosquitoes was by Gjullin et al. (1963) in California. There mosquitoes moved out of resting boxes around sunset and returned around sunrise.

Mosquito traps baited with color and odor are widely used to collect or control mosquito populations. The red box trap for resting mosquitoes has been used for Ae. aegypti (Edman et al. 1997, Kittayapong et al. 1997), which are known to select resting sites based on visual cues such as color (Manda et al. 2011). Aedes aegypti is primarily anthropophilic in host utilization, while Cx. quinquefasciatus is primarily ornithophilic, respectively; however, in Reynosa, Mexico, and the Rio Grande valley of Texas, both species were reported to feed more frequently on dogs than on humans (Olson et al. 2020). This is probably due to a high availability of domestic dogs in many low- to medium-income communities in this region.

Here odor-baited wooden box traps painted red were used to study the daytime resting activity of outdoor of Ae. aegypti and Cx. quinquefasciatus male and female populations. A model was also developed to relate relative humidity (RH) and temperature to each vector resting population. This protocol can help guide the monitoring of Ae. aegypti and Cx. quinquefasciatus resting populations for further studies such as those reported elsewhere (Estrada-Franco et al. 2020) and can greatly improve the effectiveness of mosquito control programs.

This study was carried out from April to December 2021 on the Instituto Politecnico Nacional–Centro de Biotecnologia Genomica campus (IPN-CBG) located in Reynosa, Tamaulipas, northern Mexico (26°4′10.28″N, 98°18′48.53″W; Fig. 1A–C). Reynosa is an industrialized city, which is endemic to dengue in a semiarid climate with 2 marked annual seasons. Winters are short, mild (temperature between 11 and 22°C), and dry, lasting 3 months, from December to February. Summers extend from May to August and very hot with temperatures often exceeding 40°C. Reynosa is 33 m above sea level and has an annual rainfall of 452 mm. Resting traps were placed under trees where many birds roosted and people rarely traversed. The east and south of the campus are adjacent to residential areas, and most of them have dogs or cats at home (Fig. 1D).

One wooden box trap painted red on the interior walls (Fig. 2) with a size of 120 cm × 90 cm × 90 cm was placed in 3 collection points around the IPN-CBG campus (Fig. 1D). Three traps were placed in the shade of trees on the institute campus. One Centers for Disease Control and Prevention autocidal gravid ovitrap (AGO) with hay infusion made by steeping 30 g of hay packet in 10 l of tap water for 7 days (Barrera et al. 2014) was placed inside the red wooden box to use as bait. On the top, AGO traps have a screen barrier to prevent any mosquitoes from emerging inside (Barrera et al. 2014). Mosquitoes were collected from April 21 to December 15, 2021, 11 days per month on average, every 2 hours from 0700 h to 1900 h, using Improved Prockopack Aspirators (John W. Hock Co., Gainesville, Florida). Time intervals are chosen to allow for collection and identification. Upon collection, mosquitoes were anesthetized in a −20°C freezer for 5 min, then transferred to a Petri dish and put on a chill table (1429 Chill Table, BioQuip, Rancho Dominguez, CA, USA). Identification was carried out using morphological characters as described in the taxonomic keys by Burkett-Cadena (2013). Upon identification, the number of males, females, and blood-fed females of each species was recorded. After recording, mosquitoes were recovered at room temperature for 30 min before releasing them in a central location on the campus, at an average of 88 mosquitoes from 3 resting boxes (Fig. 1D). Temperature and RH reading were recorded, using an Xiaomi sensor (Mi temperature and humidity monitor 2, Xiaomi, Beijing, China) placed at one of the 3 red wooden box traps.

Descriptive statistics based on raw counts for the 107 days of data collection of Ae. aegypti and Cx. quinquefasciatus males, females, and blood-fed females were carried out using the R with ggplot2 package. Data were inspected graphically through boxplots that revealed dispersion by means of interquartile range (IQR).

The total counts of Ae. aegypti and Cx. quinquefasciatus using 3 red wooden box traps at 7 daytime intervals (0700, 0900, 1100, 1300, 1500, 1700, and 1900) are shown in Table 1. A total of 5,327 Ae. aegypti, 6859 Cx. quinquefasciatus, 108 Ae. albopictus, and fewer than 10 unidentified mosquito specimens were collected. Of all the Ae. aegypti sampled, 84.3% were males; of all the female Ae. aegypti sampled, 12.3% were blood-fed females. Of the 6859 Cx. quinquefasciatus sampled, 50.0% were males; and of all the female Cx. quinquefasciatus sampled, 49.2% were blood-fed.

The daytime resting activity of Ae. aegypti and Cx. quinquefasciatus males, females, and blood-fed females are shown in Fig. 3. Aedes aegypti males' resting activity peaked at 0900 h, with a secondary peak at 1100 h; the lowest collection hour was at 1900 (Fig. 3A). For Ae. aegypti females and blood-fed females, almost equal numbers of mosquitoes were obtained between 0900 h and 1100 h. The lowest collection hour of Ae. aegypti females was at 1900 h, and the lowest collection hour of Ae. aegypti blood-fed females was at 1500 h.

For Cx. quinquefasciatus (Fig. 3B), the highest peak collection hour for males, females, and blood-fed females was 0700 h, and the lowest peak collection hour was at 1700 h. At peak resting (0900 h) for Ae. aegypti males, females, and blood-fed females, the average temperature and RH were 25.4 ± 0.34°C and 84.1 ± 0.61%, respectively. At peak resting (0700 h) for Cx. quinquefasciatus males, females, and blood-fed females, the average temperature and RH were 24.2 ± 0.31°C and 83.3 ± 0.63%, respectively.

The association between temperature and RH and Ae. aegypti male, female, and blood-fed female mosquitoes analyzed with GAM is shown in Table 2 and Fig. 4. Aedes aegypti male, female, and blood-fed female mosquitoes had weak positive correlation with humidity (adjusted R² = 0.0712, P < 0.001; adjusted R² = 0.0301, P < 0.001; adjusted R² = 0.0162, P < 0.001, respectively) and temperature (adjusted R² = 0.122, P < 0.001; adjusted R² = 0.0389, P < 0.001; adjusted R² = 0.0186, P < 0.05, respectively).

The association between temperature and RH and Cx. quinquefasciatus male, female, and blood-fed female mosquitoes analyzed with GAM is depicted in Table 2 and Fig. 5. Culex quinquefasciatus males, female, and blood-fed females had a positive association with RH (adjusted R² = 0.406, P < 0.001; adjusted R² = 0.441, P < 0.001; adjusted R² = 0.347, P < 0.001, respectively), and a negative association with temperature (adjusted R² = 0.363, P < 0.001; adjusted R² = 0.325, P < 0.001; adjusted R² = 0.247, P < 0.001, respectively). Increasing number of Cx. quinquefasciatus were observed when RH ranged from 25% to 84%; however, the mosquito numbers started to decrease as RH exceeded 84%. Therefore, RH recorded during our mosquito sampling had an effect on the number of Cx. quinquefasciatus caught daily in the red wooden box traps in Reynosa, Mexico. We also observed increasing numbers of Cx. quinquefasciatus when the temperature ranged from 22 to 26°C; however, mosquito numbers dropped sharply when the temperature went over 26°C.

We present the daytime resting activity of Ae. aegypti and Cx. quinquefasciatus males, females, and blood-fed females and perform a GAM to relate RH and temperature to these 2 species of mosquitoes in Reynosa, Mexico. Temporally, males, females, and blood-fed females of both species, Ae. agypti and Cx. quinquefasciatus, preferred resting in the red wooden box traps at 0900 h and 0700 h, respectively. However, no large numbers of blood-fed Ae. aegypti females were collected at any hour point of the day. We suspect 1) the odor in traps was not attractive for Ae. aegypti females, 2) the vertebrate host utilization and flight range of Ae. aegypti were different as compared with Cx. quinquefasciatus, and 3) the location of traps was farther from Ae. aegypti hosts compared to Cx. quinquefasciatus hosts. In this study we used hay infusions as odor attractants. Hay infusion has been shown to attract gravid Ae. aegypti females (Ponnusamy et al. 2010), which did not agree with our result. It is possible that the ratio of hay to water and age of the infusion water could vary in composition and volatiles among the studies. Dormont et al. (2021) reviewed mosquito attractants and noted various plant odors attractive to Ae. aegypti, which could be tested with our resting box trap in the future. Aedes aegypti females generally fly 100–500 m (McDonald 1977, Trpis and Hausermann 1986, Juarez et al. 2020). The mean distance traveled per day by female Ae. aegypti was only 16.8 m and 24.7 m for indoor and outdoor release, respectively, and the maximum overall distance per day was 160 m (Muir and Kay 1998). For Cx. quinquefasciatus, we found a large number of males, females, and blood-fed females resting inside the traps. This can be explained as Cx. quinquefasciatus, a member of the Cx. pipiens complex, has a maximum flight range from <1.0 to 2.1 km, with a mean distance between 1.27 km and 1.64 km in 36 h (Fussell 1964, Lindquist et al. 1967, Schreiber et al. 1988, Lapointe 2008, Medeiros et al. 2017). Studies investigating the feeding patterns of Cx. quinquefasciatus found that they fed primarily on birds, while in many areas of the world, while Ae. aegypti is considered anthropophilic, feeding mostly on humans (Sallam et al. 2017, Telang and Skinner 2019, Estrada-Franco et al. 2020, Fikrig and Harrington 2021). Our resting boxes were placed near the trees, which could have been locations of roosting birds. After feeding, Ae. aegypti are endophilic and mainly rest inside human dwellings (Diallo and Diallo 2020). A study in Panama found that a low percentage of resting Ae. aegypti (24.7%) was found outdoors, while a higher percentage of resting Ae. aegypti (75.1%) was found indoors. Janaki et al. (2022) found a higher proportion of resting Ae. aegypti females was caught indoors (88%) than outdoors (12%).

Casas Martinez et al. (2013) monitored the daily activity of Ae. aegypti males and females using their newly designed BioDiVector tent traps in Tapachula, Chiapas, Mexico. They found that Ae. aegypti showed 2 peaks of activity (morning and afternoon) at extra-domicile sites. Aedes aegypti female activity was higher between 0600 and 1000 h in the morning and 1600 to 1800 h in the afternoon. The activity of Ae. aegypti males was higher between 0800 and 1000 h in the morning and 1600 to 1800 h in the afternoon. Captain-Esoah et al. (2020) reported host-seeking Ae. aegypti in northern Ghana with bimodal peaks between 0600 to 0800 and 1500 to 1600 h. These results do not completely correspond to our resting peak results, which may be due to the different type of traps used, hour of collection or geographic features, and environmental variations. Mutebi et al. (2022) conducted a study about the diel activity patterns of host-seeking Ae. aegypti females in Brownsville, Texas. They found Ae. aegypti females had 2 peaks of diel biting activities, between 0700 and 0800 h and between 1900 and 2000 h. In our study the resting period of Ae. aegypti male, female, and blood-fed females was between 0900 and 1300 h. These corroborative results of Ae. aegypti populations on both sides of the USA-Mexico border suggest that individuals seek resting habitat later in the morning following a peak in host-seeking activity.

Gjullin et al. (1963) conducted a study about the daily resting cycles of Culex species using red box shelter units and found that the daily resting period of Cx. pipiens and Culex tarsalis extended from 0800 h to sunset, leaving the box after sunset and returning before sunrise, as Culex species are often crepuscular in host-seeking behavior. This result is in accordance with our findings as we have noted a unimodal clear peak at 0700 h; it is at that time in the morning when we captured Culex species that were done host-seeking and then ready to rest for the rest of the day. To investigate whether mosquitoes rested at night in our red wooden box traps, we inspected the traps during 3 nights from 1900 to 0500 h (data not shown), and no resting mosquitoes were found as reported elsewhere (Zuharah and Sumayyah 2019). As previously mentioned, Cx. quinquefasciatus may rest close to hosts at night given their crepuscular feeding behavior. Karlekar and Andrew (2016) found that Cx. quinquefasciatus were feeding at dusk (1800 to 2000 h) and resting on the roof of cattle sheds and grain storage, while during the daytime, they preferred resting near water tanks and water puddles.

This GAM analysis helps to explain daytime resting behavior of Ae. aegypti and Cx. quinquefasciatus. Some discrepancies in temperature and RH were observed between Ae. aegypti and Cx. quinquefasciatus. For Cx. quinquefasciatus, temperature and RH were associated with their resting behavior but not with the resting pattern of Ae. aegypti. The GAMs have been used to indicate the distribution and abundance of mosquito larvae in residential areas (Heersink et al. 2016), to evaluate different mosquito collection methods (Câmara et al. 2022), and to analyze the relationship between climate variables with mosquito density (Xu et al. 2017). To our knowledge, no studies have used GAMs to explain the relationship between environmental factors and mosquito daytime resting patterns. However, due to the univariate model construction, this model cannot fully explain the relationship between the environmental factors and the resting behavior of those mosquito species. Adding more variables could better refine the model for interpretation.

Here the resting pattern for the 2 species of mosquitoes peaked at high RH recorded during the day. However, high temperature was less related to mosquito captures. More Cx. pipiens and Ae. aegypti preferred to stay in the cooler and more humid zone of the cage during the night and day, respectively (Kessler and Guerin 2008), and Cx. quinquefasciatus preferred to rest in cool, humid, shady areas such as water tanks, shade near puddles, and trees (Rios et al. 2006, Karlekar and Andrew 2016). In correspondence, we did not observe more mosquitoes at 1300, 1500, and 1700 h, where the temperature was higher than that of 0700 and 1100 h. We suspect mosquitoes were resting more during the highest RH, and for Culex, the highest resting counts were first thing in the morning, which was when the humidity was the highest. The time of the day (Fig. 2) shows that male and female Ae. aegypti arrived to resting habitat over the course of the morning. However, for Culex it is more complicated to interpret since they were already utilizing resting habitat during the first sampling period, and we are not sure how many actually arrived at the first sampling period (0700 h) versus how many arrived before (1900 to 0500 h). We also noted strong day-to-day mosquito variations indicating that other environmental factors influence shelter-seeking behavior, such as wind speed, cloud cover, and rain, which we did not consider in our analysis.

A limitation of this current study was that all wooden box traps were placed on a university campus away from residential homes, which means the results could be influenced by different proximity to larval and host-seeking habitat. We also collected very few Ae. albopictus and other mosquito species given that the traps were at the same collection site. Second, all traps are placed on the ground and placing resting habitat in tree canopies could have influenced the results, similar to how traps collecting host-seeking mosquitoes were different at ground level versus forest canopy (Andreadis and Armstrong 2007, Pereira-Silva et al. 2021). Third, we used hay infusion AGO traps to augment the resting box habitat. This odor and the increased humidity from the water might have attracted oviposition seeking females. We did not distinguish gravid from unfed female mosquitoes, which could have been done to distinguish the 2 behaviors. Fourth, additional environmental variables should be recorded to investigate additional abiotic and biotic factors that influence mosquito resting behavior.

The red wooden odor-baited box traps can be used efficiently for monitoring mosquitoes as other traps do for mosquito surveillance such as gravid traps (Reiter et al. 1986), light traps (Sudia and Chamberlain 1962), BG sentinel traps (Maciel-de-Freitas et al. 2006), and oviposition traps (Barbosa et al. 2007). This study may help guide sampling studies aimed at collecting blood-fed females for host-range studies (Estrada-Franco et al. 2020, Olson et al. 2020) and can improve vector surveillance and disease management programs.

We thank the Secretaría de Investigación y Posgrado–Instituto Politécnico Nacional (IPN) (SIP Nos.20220163 and 20201972) for financial support to N.A.F.-S. The article processing charge was funded by Texas A&M University, College Station. The funders did not support the study design, data collection, and their analysis. We thank the administrators of Centro de Biotecnología Geómica–IPN for providing facilities to deploy the red wooden box traps within the campus. Last, but not least, we also appreciate the support from Dr. Adebiyi Adeniran for designing the red wooden box traps.