Portal de Periódicos da CAPES

Impacts of sampling rhythm and exposition on the effectiveness of artificial resting shelters for mosquito collection in northern Germany

2020; Wiley; Volume: 45; Issue: 1 Linguagem: Inglês

10.1111/jvec.12383

1948-7134

Felix Gregor Sauer, Linda Jaworski, Renke Lühken, Ellen Kiel,

Viral Infections and Vectors

Mosquitoes are among the most important arthropod vectors of pathogens and are of growing medical importance in European countries (Semenza and Suk 2017). Hence, knowledge of the spatial-temporal distribution of mosquitoes and the monitoring of mosquito-borne pathogens is crucial to assess the risk for human and animal health. For that purpose, mosquitoes are commonly collected by conventional active traps (Versteirt et al. 2013, Zittra et al. 2017, Wagner et al. 2018). Those traps attract adult mosquitoes by olfactory lures or visual cues. They are often able to collect large numbers of mosquitoes and mosquito species but are significantly biased towards host-seeking females as main targets for pathogen surveillance (Lühken et al. 2014). In contrast, methods for the collection of resting mosquitoes are generally expected to provide rather unbiased results of the population, including higher amounts of males and blood-fed females compared to conventional traps. The collection of a higher amount of blood-fed females is beneficial for the analyses of the host-feeding patterns of mosquitoes (Brugman et al. 2017) or pathogen surveillance programs (Burkett-Cadena et al. 2016). Artificial resting sites are thus an efficient tool to sample certain mosquito taxa (Burkett-Cadena et al. 2008). Several types of artificial resting sites have been developed and used in North America, including wooden boxes (Edman et al. 1968), wire-frame shelters (Burkett-Cadena 2011) or fiber pots (Komar et al. 1995). However, their deployment in large numbers would need considerable effort for transport and construction. In a study from Florida, Burkett-Cadena et al. (2019) developed a promising artificial resting site with a pop-up design, which can sustainably simplify set-up and transport. The trapping performance of pop-up bags for European mosquito fauna is unknown. In general, the number of European studies with usage of artificial resting sites is quite low, compared to the number of studies using conventional traps. Therefore, the aim of our study was to evaluate the efficacy of garden pop-up bags as artificial resting sites for European mosquito studies. This included an analysis of the influence of 1) sampling rhythm, 2) hour of sampling (morning vs evening), 3) height level (ground level vs approximately 1.8 m) and 4) orientation of the bags relative to the tree (sideways vs downwards) on the sampling rates of mosquitoes in order to develop an effective sampling strategy. We conducted two sub-studies to analyze the effects of hour of sampling, height level and sampling rhythm on trapping efficacy in the summer of 2016, and to evaluate the effects of different orientation of garden pop-up bags (Relaxdays GmbH, Halle, Germany; 76 l), i.e., with the opening exposed downwards vs sideways relative to the tree (summer of 2017). These studies were conducted in a forest close to Oldenburg (Lower Saxony, Germany) (N53.158540°, E8.124390°). The forest is interspersed with temporary waters such as ditches and ponds that provide larval habitats for mosquitoes. The study area of approximately 0.5 ha was situated in the inner parts of the forest, which is characterized by homogeneous vegetation dominated by oaks and hornbeams. The garden pop-up bags we used were comparable to those tested by Burkett-Cadena et al. (2019), but we did not supply the bags with a manual-powered capture chamber. Instead, resting mosquitoes were sampled with a hand-made aspirator, which was built with a fan blower (Seaflo, Freehold, U.S.A.). Construction and suction capacity of our aspirator is based upon the description by Vazquez-Prokopec et al. (2009). In sub-study 1 from mid-June to late August, 2016, 24 garden pop-up bags were attached to six trees within a minimum distance of 15 m. At each tree, the bags were positioned at two heights: two pop-up bags at the ground and two at a height of approximately 1.8 m measured from the lower edge of the bags (Figure 1A). We expected there would be fewer resting mosquitoes at even greater heights (unpublished data). These bags, colored dark green outside and black inside, were directed sideways, so that the circular opening (diameter ≈ 0.46 m) was exposed vertical to the ground (Figure 1A). Compass direction of the bags was chosen randomly. In order to analyze the impact of exposure time on sampling rate, resting mosquitoes were sampled eight times in the afternoons (14:00 to 16:00) by choosing a randomized exposure time between one and thirteen days. One CO2-baited BG-Sentinel trap (BioGents, Regensburg, Germany) was placed in the center of our study site and served as a reference. The trap was used weekly for 24 h only during the study period (total of seven times). For five additional days during the study period, resting mosquitoes were sampled twice a day in the morning (06:00 – 08:00) and late afternoon (17:00 – 19:00) to determine the effects of hour of sampling on trapping efficacy. In sub-study 2 from August to September, 2017, an additional experiment was conducted to study the influence of the orientation of the bags on the mosquito samples. Then, 14 garden pop-up bags (same as described above) were attached to seven trees, all of them at the same height (approximately 1.8 m). At each tree, one bag was directed downwards (n = 7) and one bag was directed sideways (n = 7) on the opposite tree side (Figure 1B). Compass direction of the set-up was chosen randomly for each tree. Resting mosquitoes in the garden pop-up bags were sampled twice a week from August to September, 2017 (in total nine sampling days) with a minimum exposure time of three days. Aspirated mosquitoes were killed by freezing and stored at -18° C. All specimens were identified by morphological characters (Mohrig 1969, Lechthaler 2005, Becker et al. 2010). Physiological status of the females was differentiated between non-blood-fed and blood-fed, whereby the term “blood-fed” denotes all specimens from freshly engorged to gravid females. Non-parametric Wilcoxon rank-sum tests were used to test the influence of hour of sampling, height (sub-study 1, data from 2016) and orientation of bags (sub-study 2, data from 2017) on the mosquitoes per resting site and collection. We separately analyzed the mean numbers of blood-fed females of the three most frequently collected taxa, Culiseta morsitans/fumipennis, Culex pipiens s.l., and Aedes spp. mosquitoes, and a Spearman's rank order correlation test was applied to analyze the influence of exposure time on the total number of mosquitoes per resting site and the number of blood-fed females (sub-study 1, data from 2016). As a potential confounding factor, the influence of the compass direction (expressed as four levels: north, east, south, and west) on the mosquito sampling rate was checked by a Kruskal-Wallis rank sum test. Statistical analyses and visualization were conducted with the R program version 3.5.1. In sub-study 1 (2016), the CO2-baited reference trap collected 1,516 females, comprising nine mosquito taxa (Table 1). None of these females was blood-fed. During the resting site studies, 1,774 mosquitoes of nine taxa were collected, including 433 males and 351 blood-fed females (Table 1). The number of specimens per taxa differed strongly between both methods. Culiseta morsitans/fumipennis represented 85% of mosquitoes sampled in the artificial resting sites but only 1% in the CO2-baited trap. In contrast, Cx. pipiens s.l. dominated in CO2-baited traps (83% of all mosquitoes). Aedes geniculatus (n = 1), Aedes vexans (n = 3) and Coquillettidia richiardii (n= 2) were trapped only by means of the CO2-baited trap, whereas Culex territans (n = 21) and Culiseta annulata/subochrea (n = 17) were collected only with garden pop-up bags (Table 1). Exposure time had no significant correlation with the total number of specimens per resting site (r = −0.068, p = 0.352), but the number of blood-fed females per resting site significantly increased with exposure time (Spearman correlation, r = 0.146, p = 0.046). The number of mosquito specimens per resting site and collection was significantly higher in the evening (W = 5882, p = 0.046). This diurnal difference was not observed for the collected blood-fed females (W = 6138, p = 0.166). No significant effects of hour of sampling on the mean number of collected Cs. morsitans/fumipennis (W = 7791.5, p = 0.077) (Figure 2B) and Aedes spp. (W = 6645, p = 0.429) were evident, but more Cx. pipiens s.l. were sampled in the evening (W = 8031, p < 0.001) (Figure 2E). The mean number of mosquito specimens (W = 2382, p < 0.001) and of the number of blood-fed females (W = 3050, p < 0.001) per resting site and collection differed significantly between artificial resting sites exposed at different heights, with higher values in the elevated resting sites (Figure 2A and 2D). This statistical difference, however, results from the high number of collected Cs. morsitans/fumipennis (Table 1). Almost three times more specimens of Cs. morsitans/fumipennis rested in garden pop-up bags deployed at a height of 1.8 m compared to the ground (W = 2410, p < 0.001) (Figure 2A). Similarly, significantly more Cx. pipiens s.l. were sampled in the elevated pop-up bags compared to the pop-up bags on the ground (W = 3646, p = 0.004) (83 vs 40 specimens) (Figure 2D). On the other hand, significantly more mosquitoes of the genus Aedes were sampled in the pop-up bags on the ground (66 specimens) compared to resting sites at 1.8 m (17 specimens) (W = 5412, p < 0.001). The compass direction of the pop-up bags had no significant influence on the number of collected mosquitoes per resting site (Kruskal-Wallis test, p = 0.114). In sub-study 2 (2017), we collected significantly more mosquitoes per resting site and collection in the garden pop-up bags directed downwards (W = 2446, p = 0.018) compared to bags directed sidewards. This was mainly due to Cs. morsitans/fumipennis (W = 2492, p = 0.007) (Figure 2C), which was again the dominant species (82%). No differences in preference for orientation were observed for the other mosquito taxa, due to low numbers of collected specimens in 2017. The proportion of 18% blood-fed mosquitoes is similar to the results of other artificial resting site studies, in which the number of blood-fed specimens ranged between 10% and 20% (Edman et al. 1968, Komar et al. 1995, Burkett-Cadena 2011, Brugman et al. 2017). In contrast to these studies, the amount of collected males at resting sites was relatively low, particularly in 2017 (Table 1). This is likely due to the hot summer of 2017. We assume that the numbers of rather short-lived males had already declined when our study started in August. The dominance of Cs. morsitans/fumipennis and Cx. pipiens s.l. in garden pop-up bags is in line with studies from North America, demonstrating that artificial resting sites are predominately utilized by members of the genera Anopheles, Culex, and Culiseta. These mosquitoes naturally rest in cavities such as hollow trees or undercut banks. Garden pop-up bags mimic those structures and are, therefore, a useful tool to systematically sample resting mosquitoes, when appropriate natural structures are missing or are not accessible (Burkett-Cadena et al. 2019). In practice, garden pop-up bags provide additional benefits compared to other commonly used artificial resting sites such as wooden boxes (Edman et al. 1968, Morris 1981), wire frame shelters (Burkett-Cadena 2011), or plastic trash cans (Burkett-Cadena et al. 2008). The space-saving and lightweight design of garden pop-up bags facilitate transport and storage in large numbers and they do not require any construction time. In addition, most models of garden pop-up bag are equipped with loops originally used for the transport during gardening. Those loops can simplify the stable exposition on trees or other structures. The sampling rhythm and the hour of sampling are important aspects for the effective planning of a mosquito sampling study. In this study, we observed only a weak relationship between exposure time and the number of collected mosquitoes by resting sites. Likewise, Howard et al. (2011) observed no difference in the trapping efficacy between a resting site sampling twice weekly vs a sampling of four consecutive days. We suppose that mosquitoes frequently change or leave their resting sites, so that an exposure time of 24 h is sufficient to collect resting mosquitoes effectively by means of garden pop-up bags. In addition, the hour of sampling only plays a minor role for the number of collected resting mosquitoes. This is supported by the study of Burkett-Cadena et al. (2008) comparing resting site sampling between 07:00 to 08:30 and 16:00 to 19:00 without significant differences for most mosquito species. The appropriate exposition of artificial resting sites is a crucial aspect for field studies (Edman et al. 1968). In this study, resting site-specific patterns were detected for a height difference of just 1.8 m. Different authors demonstrated vertical host-seeking patterns, but the resting height preferences are largely unknown. Our findings suggest that mosquitoes have specific resting height preferences. However, this aspect needs further investigation, since our observations are based on a single study site and on certain mosquito taxa. Finally, pop-up bag opening directed downwards revealed significantly more mosquitoes in general and Cs. morsitans/fumipennis in particular. The direction of the opening downwards maximizes the shaded area within the pop-up bags, which may thus positively affect resting site selection of Cs. morsitans/fumipennis. This is in agreement with Morris (1981), who concluded that Culiseta prefer dark-colored artificial resting sites in shaded habitats. In summary, garden pop-up bags provide a useful alternative to collect resting mosquitoes. Their practical benefits (lightweight, space-saving, cost-effective, and no construction time) combined with the high proportion of blood-fed females make them a useful complement for surveillance purposes and to monitor certain mosquito taxa. An exposure time of 24 h is sufficient for an efficient sampling, whereby the hour of sampling can be chosen flexibly. The appropriate exposition of the garden pop-up bags (e.g., height level) might play an important role in maximizing the number of species and specimens, which needs further investigation. We greatly acknowledge Jacqueline Grave for her help during fieldwork as well as Esther Timmermann and Melanie Willen by preparing the trips. The study was funded by the German Federal Ministry of Food and Agriculture (BMEL), grant numbers 2819105215, 2819104515, and 2819104315. We thank the forest ranger office (Revierförsterei Oldenburg) for the acquiescence of our fieldwork.