Comparing Light—Emitting—Diodes Light Traps for Catching Anopheles Mosquitoes in a Forest Setting, Western Thailand

Simple Summary A field study was conducted in a forest to compare the effectiveness of light traps fitted with different bulbs across the wavelength spectrum. Ultraviolet (UV) fluorescent light was found to be most effective to collect adult Anopheles mosquitoes from 21:00 h to the pre-dawn hours in the dry season. These findings have important implications for monitoring vector density in the endemic malaria areas where other methods cannot be executed. A more comprehensive and systematic study of how mosquitoes respond to light would benefit Thailand’s national control program. Their potential for more precisely sampling vectors holds promise as a tool for mosquito monitoring endemic malaria areas and outbreak hotspots. Abstract Light traps are a common method for attracting and collecting arthropods, including disease vectors such as mosquitoes. Various types of traps have been used to monitor mosquitoes in a forest in Western Thailand. In this study, four Light Emitting Diodes (LED) light sources (UV, blue, green, and red) and two fluorescent lights (white and UV) were used to trap nocturnal adult mosquitoes. These traps were used with light alone and not any additional attractant. The experiment was conducted from 18:00 to 06:00 h. on six consecutive nights, every two months, across dry, wet, and cold seasons. All specimens were first identified by morphological features and subsequently confirmed by using PCR. We collected a total of 873 specimens of 31 species in four genera, Anopheles, Aedes, Culex, and Armigeres. Anopheles harrisoni was the predominant species, followed by Aedes albopictus, Culex brevipalpis, Culex nitropunctatus, and Armigeres (Leicesteria) longipalpis. UV fluorescent light was the most effective light source for capturing forest mosquitoes, followed by UV LED, blue LED, green LED, white fluorescent, and red LED. The optimal times for collection were from 21:00 to 03:00 h in the dry season. Our results demonstrate that appropriate sampling times and light sources should be selected for optimal efficiency in vector surveillance programs.


Introduction
Mosquitoes are well-recognized as important arthropod vectors that are responsible for transmitting many medically important pathogens and parasites, including viruses, bacteria, protozoans, and nematodes, which cause serious diseases, such as malaria, dengue, chikungunya, encephalitis, and filariasis [1]. Of these, malaria is a serious and sometimes fatal disease caused by a parasite transmitted to humans via Anopheles mosquitoes. Interventions can be implemented to mitigate and reduce the risk of infection and prevent disease. Long-lasting insecticidal nets and indoor residual spraying have long been used as the main interventions to combat malaria indoors [2]. Despite the efficacy of these

Mosquito Collection
Six trapping locations were established in the study area, approximately 200 m apart. Each light trap was hung approximately 150 cm above the ground level [29]. The experiment was conducted on six consecutive nights in each of the following months: February and April (dry season), June and August (wet season), and October and December (cold season), in 2020. During each of six consecutive trapping nights, the light traps were rotated among the six locations, using a Latin square design. Mosquito collection was conducted over 12 h, from 18:00 to 06:00 h. Each night, mosquitoes were collected from each trap every 3 h (at 21:00 24:00, 03:00, and 06:00 h). The collected mosquitoes were placed in a −20 °C freezer for 60 s and then morphologically identified.

Morphological Species Identification
Mosquitoes were separated from other insects, carefully examined by using a stereomicroscope, and identified according to sex. Species were morphologically identified based on the external features of proboscis, maxillary palpus, scutum, wing vein, spiracular-setae, legs, abdomen, and scales, using a standard taxonomic key [30][31][32][33][34][35]. All primary Anopheles species, including the Anopheles minimus complex, Anopheles maculatus group, and Anopheles dirus complex, were then stored at −20 °C for molecular identification.

DNA Extraction
All laboratory work was performed at the Department of Entomology, Faculty of Agriculture, Kasetsart University, Bangkok, Thailand. DNA from individual An. minimus complex, An. maculatus group, and An. dirus complex mosquitoes was extracted by using an EZNA ® tissue DNA kit (Omega Bio-Tek, Norcross, GA, USA). The final elution volume for DNA extractions was 50 μL. Distilled water was used as a negative extraction control. DNA solutions were stored at −20 °C until further use [36].

Mosquito Collection
Six trapping locations were established in the study area, approximately 200 m apart. Each light trap was hung approximately 150 cm above the ground level [29]. The experiment was conducted on six consecutive nights in each of the following months: February and April (dry season), June and August (wet season), and October and December (cold season), in 2020. During each of six consecutive trapping nights, the light traps were rotated among the six locations, using a Latin square design. Mosquito collection was conducted over 12 h, from 18:00 to 06:00 h. Each night, mosquitoes were collected from each trap every 3 h (at 21:00 24:00, 03:00, and 06:00 h). The collected mosquitoes were placed in a −20 • C freezer for 60 s and then morphologically identified.

Morphological Species Identification
Mosquitoes were separated from other insects, carefully examined by using a stereomicroscope, and identified according to sex. Species were morphologically identified based on the external features of proboscis, maxillary palpus, scutum, wing vein, spiracularsetae, legs, abdomen, and scales, using a standard taxonomic key [30][31][32][33][34][35]. All primary Anopheles species, including the Anopheles minimus complex, Anopheles maculatus group, and Anopheles dirus complex, were then stored at −20 • C for molecular identification.

DNA Extraction
All laboratory work was performed at the Department of Entomology, Faculty of Agriculture, Kasetsart University, Bangkok, Thailand. DNA from individual An. minimus complex, An. maculatus group, and An. dirus complex mosquitoes was extracted by using an EZNA ® tissue DNA kit (Omega Bio-Tek, Norcross, GA, USA). The final elution volume for DNA extractions was 50 µL. Distilled water was used as a negative extraction control. DNA solutions were stored at −20 • C until further use [36].

Data Analysis
The numbers of each mosquito species captured by six different light traps were compared using the Kruskal-Wallis test by mean ranks. The efficacy of the traps to collect mosquitoes was evaluated by a generalized linear model (GLM). The total number of collected Anopheles mosquitoes per trap-night was treated as the response variable, and the light sources, seasons, and collection periods were defined as key factors. The goodnessof-fit model was validated by considering the deviance value; the optimal model was selected by mean deviance closest to one. The results from testing the model showed statistical significance for all tests with p < 0.05. To finalize, the parameter (key factor) that was a statistically significant predictor of the number of mosquitoes caught was used to determine the regression coefficients (B), standard errors, p-values, and 95% confidence intervals for the coefficients, using the Wald Chi-square test. Mosquito-trap efficacy was analyzed based on incidence rate ratio (IRR), which provided a standard incidence rate (IRR = 1) for comparison of variables. All data were analyzed by using the SPSS program (version 11.0; SPSS Inc., Chicago, IL, USA).
In the dry season, the trap equipped with UV fluorescent light was the most effective for attracting Anopheles mosquitoes; the highest number was collected from 21:00 to 24:00 h, followed by 24:00-03:00 h and 03:00-06:00 h (Table 4; Figure 2). In the wet season, the UV fluorescent light trap was also the most effective for attracting Anopheles mosquitoes from 21:00 to 24:00 h (Table 4, Figure 3). In the cold season, the UV florescent trap captured most Anopheles mosquitoes from 18:00 to 21:00 h (Table 4; Figure 4).      We used a GLM with negative binomial regression to evaluate the factors (light sources, seasons, and time periods) that influenced the efficacy of light traps to capture Anopheles species. Deviance from the goodness-of-fit test at 0.751 and Pearson Chi-square at 1886.706 indicated that the negative binomial regression was suitably obtained (Omnibus test; p < 0.001). Light sources, seasons, and collection time periods were all significant predictors that influenced the number of Anopheles mosquitoes collected per trap (p < 0.05; Table 5). Based on the IRR values, the best efficiency was achieved by using the UV fluorescent light (treated as the standard; IRR = 1), followed by UV LED (IRR = 0.437), blue LED (IRR = 0.202), white fluorescent (IRR = 0.127), green LED (IRR = 0.063), and red LED (IRR = 0.045). The predicted count for mosquito captures in the dry season was 21.649, compared to the cold season as the standard (IRR = 1). In addition, the best trapping time period was 21:00-24:00 h (IRR = 2.54), followed by 24:00-03:00 h (IRR = 2.25) and 06:00-09:00 h (IRR = 1.93), as compared to 03:00-06:00 h as the standard (IRR = 1; Table 5). We used a GLM with negative binomial regression to evaluate the factors (light sources, seasons, and time periods) that influenced the efficacy of light traps to capture Anopheles species. Deviance from the goodness-of-fit test at 0.751 and Pearson Chi-square at 1886.706 indicated that the negative binomial regression was suitably obtained (Omnibus test; p < 0.001). Light sources, seasons, and collection time periods were all significant predictors that influenced the number of Anopheles mosquitoes collected per trap (p < 0.05; Table 5). Based on the IRR values, the best efficiency was achieved by using the UV fluorescent light (treated as the standard; IRR = 1), followed by UV LED (IRR = 0.437), blue Insects 2021, 12, 1076 9 of 13 LED (IRR = 0.202), white fluorescent (IRR = 0.127), green LED (IRR = 0.063), and red LED (IRR = 0.045). The predicted count for mosquito captures in the dry season was 21.649, compared to the cold season as the standard (IRR = 1). In addition, the best trapping time period was 21:00-24:00 h (IRR = 2.54), followed by 24:00-03:00 h (IRR = 2.25) and 06:00-09:00 h (IRR = 1.93), as compared to 03:00-06:00 h as the standard (IRR = 1; Table 5).

Discussion
Among the four LED wavelength ranges and two fluorescent lights, UV fluorescent was the most effective for mosquito collection, followed by UV LED and blue LED. This is consistent with previous studies that reported the effectiveness of UV fluorescent light for collecting nocturnal mosquitoes [28,40,41]. The UV fluorescent light used in this study had a lower wavelength range (354-468 nm) than those previously evaluated. Breyev et al. [42] reported that more night-biting Anopheles mosquitoes, Anopheles hyrcanus and Anopheles maculipennis, were captured using UV fluorescent light traps (300-400 nm) compared with other light traps of longer wavelengths. This previous study also reported that light traps with spectral beams and 364-400 nm wavelengths increase mosquito attraction [43]. Other previous studies also reported differences in mosquito attraction to LED lights of different wavelengths [18,22,44]. Insect vision responds differently to UV, blue, and green spectra, and responses may vary between species, as well as individuals of the same species inhabiting different areas [21]. Silva et al. [45] reported that light traps with green and blue LEDs attracted more mosquitoes than other LEDs and incandescent lights. In our study, traps with blue LED attracted more mosquitoes than those with green and red LEDs. Although colors of the same brightness are used to evaluate mosquito attraction, physiological light intensities can be affected by differing wavelength absorption in the mosquito eye [46].
In our study, a higher number of Aedes species were collected by using the UV fluorescent light trap than the Culex species. Similarly, Tchouassi et al. [47] sampled Rift Valley fever vectors, using LED CDC light traps (red, green, and blue) and captured more Aedes species than Culex species. The lower response of Culex species to light traps has not been confirmed but could be attributed to their sensitivity to different wavelengths of light or neurophysiological aspects of their visual systems [48]. Kawada et al. [49] documented that nocturnal host-seeking behavior in nonblood-fed females of Aedes aegypti (L.) and Aedes albopictus (Skuse) was positively correlated with increasing light intensity. The study used an automatic recording device equipped with a photoelectric sensor and found that the eye of Ae. aegypti was highly sensitive to dim light, allowing the species to be active at night [49]. Our results also showed that diurnal Aedes mosquitoes were captured during the early evening period. Additionally, Muie et al. [50] reported that the eye of female Ae. aegypti has a broad spectral sensitivity, ranging from UV (323 nm) to orange (621 nm) with peaks in the UV (323−345 nm) and green (523 nm) wavelength ranges. Nocturnal Culex species' attraction to light sources differs from other mosquito species [42,51]. In our study, more Culex mosquitoes were collected by using the UV fluorescent light trap than other light sources. Our results demonstrate that light source has a significant effect on the number of mosquitoes collected and can be tailored to attract specific genera.
Our light-trap mosquito sampling was conducted from February to December 2020 (dry, wet, and cold seasons). Most Anopheles species were collected in the dry season. A previous publication reported that heavy rainfall can flush out larval mosquitoes, resulting in reduced adult densities [52], which is consistent with the lower numbers we observed in the wet season. The best collection times were 21:00-24:00 h and 24:00-03:00 h during the dry season. Harbach et al. [53] reported that mean biting activity of anopheline mosquitoes peaked between 21:00 and 22:00 h Rattanarithikul et al. [54] also reported a prominent peak of Anopheles species blood-feeding between 18:00 and 23:00 h. Differences in mosquito behaviors among individuals of the same species have been reported, and are often related to adaptations to human behaviors [55], as well as geographical, climatic, and environmental conditions [7,56,57].

Conclusions
Trapping Anopheles species was found to be most efficient when using light traps fitted with a UV LED light, with the optimal times for collection from 21:00 to 03:00 h in the dry season. We demonstrated that the standard commercial UV fluorescent traps can be replaced with UV LED light traps in sampling Anopheles mosquitoes. For future surveillance of adult mosquitoes and incorporation of lights with either the "black hole" styled traps or modifications of other styles of traps, such as CDC, the inclusion of additional chemical lures, such as octanol or lactic acid, may change number of mosquitoes collected, even in the absence of carbon dioxide. This study contributes crucial information for monitoring vector density in regions affected by malaria. A more comprehensive and systematic investigation of mosquito responses to light would be beneficial to the national control program, facilitating more precise vector sampling and monitoring.