Evaluation of Mansonia spp. Infestation on Aquatic Plants in Lentic and Lotic Environments of the Madeira River Basin in Porto Velho, Rondônia, Brazil

Mansonia Blanchard comprises 2 subgenera: Mansonioides Theobald, which has 10 species distributed in Asia and 2 in Africa (Ronderos and Bachmann 1963), and the nominotypical subgenus (Mansonia), which includes 15 species with primarily Neotropical distribution (Guimarães 1997, Forattini 2002, Barbosa 2007, Harbach 2020), with 12 species recorded in Brazil.

Females of Mansonia spp. are voraciously hematophagous. The spiracular apparatus of the active immature forms is adapted to perforate submerged vegetation to obtain oxygen from plant tissue (Harbach 2020). Immature individuals attach themselves to aquatic plants, from whose aeriferous aerenchyma they get the oxygen necessary for breathing. Van den Assem (1958) and Wharton (1959) hypothesized that this behavior among immature stages of Mansonia spp. of piercing submerged plant tissues, especially the root tissue of floating plants (Eichhornia, Pistia, Salvinia, Typha, and grasses), emerged as an evolutionary survival strategy against attacks by fish, aquatic insects, and other aquatic animals.

The proliferation of aquatic plants is sometimes linked to a reduction in water flow and/or increase in organic matter content (eutrophication), often due to human activities, which may contribute to the spread of these mosquitoes (Forattini 2002). Although they are essentially forest species, they can become established on the outskirts of towns and inhabited centers, as long as favorable conditions exist in the form of suitable breeding sites (Forattini 2002).

Breeding habitats are typically rich in nutrients and consist of medium or large accumulations of either still or slow-moving water. These include river and lake backwaters in which the water surface is partially or totally covered with plants (Forattini 2002). These processes provide insight into population dynamics, community structure, and ecosystem function, allowing us to explore the effects of environmental factors and interspecific interactions, making it possible to determine, for example, whether the population will persist in a certain habitat (Tibério 2021).

This study aims to evaluate the infestations of aquatic plants by Mansonia spp. lentic environments of the Madeira River basin near the Jirau Hydroelectric Power Plant (Jirau HPP), in Rondônia, Brazil. It also seeks to establish parameters for the analysis of environmental changes that influence the ecology of this important group of insects, which includes numerous taxa that use lentic environments (here the dam reservoir) as developmental habitats and that transmit pathogens.

Mansonia species are not known to be vectors of endemic pathogens in Brazil. However, some species have been found to be naturally infected with different viruses, and they are real plagues, making human life and livestock difficult in certain regions. For example, Ma. pseudotitillans (Theobald) was found to be infected with the Saint Louis encephalitis virus (Segura and Castro 2007), and Ma. titillans (Walker) with Venezuelan encephalitis virus and febrile disease arbovirus (Hervé et al. 1986, Consoli and Oliveira 1994).

The study was conducted on a 105.7 km stretch of the Madeira River (79 km as the crow flies) in the state of Rondônia that is affected by the reservoir of the Jirau HPP, with 8 sampling points upstream of the dam and 2 downstream. The predominant vegetation in the sample area is classified as lowland and submontane open rainforest and is also categorized as terra firme forest (IBGE 2012). The region's climate is classified as tropical humid hyperthermal (Cochrane and Cochrane 2010), with minimum annual average temperatures ranging between 20°C and 22°C and maximum temperatures between 31°C and 33°C, RH between 85% and 90%, and average rainfall between 2,000 mm and 2,200 mm (Sombroek 2001). The relief of the region is flat, with altitudes ranging from 100 m to 300 m above sea level.

Amazonian rivers are commonly divided into clear, white, and dark waters, depending on the type of sediment and/or organic matter they carry. Whitewater rivers drain geologically young regions (e.g., the Andes) and are rich in sediment. Clearwater rivers originate in older geological regions (central Brazil and the Guyanas).

Sampling was conducted monthly from February 2016 to June 2018, based on a survey of potential breeding sites for Mansonia spp. in the Jirau HPP reservoir. Sample points for aquatic macrophytes and Mansonia mosquitoes located in reservoir of the Jirau HPP, in 10 sampling sites were selected, as follows: Foz do Rio Jaci, site 1 (9°12′49.9″S, 64°24′06.0″W); Igarapé Flórida, site 2 (9°07′55.9″S, 64°31′46.5″W); Igarapé do Raul, site 3 (9°15′14.4″S, 64°42′16.1″W); Casa da Colina, site 4 (9°15′55.3″S, 64°40′16.6″W); Base da Barragem, site 5 (9°16′58.1″S, 64°38′09.4″W); Foz do Igarapé Jirau, site 6 (9°22′47.2″S, 64°45′30.0″W); Caiçara/Primavera, site 7 (9°24′11.8″S, 64°49′14.6″W); Igarapé Caiçara, site 8 (9°32′52.4″S, 64°50′25.8″W); Bolsão Mutum, site 9 (9°35′51.5″S, 64°55′55.1″W); and Igarapé São Lourenço, site 10 (9°33′52.50″S, 65°1′23.10″W) (Fig. 1).

Each site had a high concentration of aquatic vegetation belonging to the following plant species: Eichhornia crassipes (Mart.) Solms-Laubach, E. azurea (Sw.) Kunth, Ceratopteris pteridoides (Hook.) Hieron, Pistia sp., Salvinia sp., and Paspalum sp.

Sampling was performed between 9:00 a.m. and 5:00 p.m. with a total of 29 collections per site. There were 7 washing points for macrophytes, with 15 min spent on sampling per point, for a total of 105 min (1.45 h) per site. The same methodology was followed at all 10 sampling sites during a 14.5-h effort for each collection, totaling 420.5 h. Larval collections were carried out according to the methodology used by Ferreira (1999), where the locations were selected at random, following which they were removed manually and quantified, separated according to the species of the plant, and placed in a 25-liter white polypropylene tray (54 cm × 33.5 cm × 19 cm) to be washed with water. Subsequently, the water was filtered using a BioQuip® cone-shaped net to collect larvae detached from the roots. The specimens were then taken to the laboratory for further identification. Mosquito collections were authorized by the Chico Mendes Institute for Biodiversity Conservation (ICMBio), no. 26805-2.

The identification of the specimens to the genus level was performed from the direct observation of the morphological characters observed under the stereomicroscope (Zeiss®) and consultation of the respective descriptions/diagnoses of the genus using dichotomous keys prepared by Lane (1953), Consoli and Lourenço-de-Oliveira (1994), Forattini (2002), and Barbosa (2007). Later the specimens were incorporated into the Entomological Collection of the Federal University of Rondônia (UNIR).

The Kruskal-Wallis test, with a 5% significance level, was used to test for significant differences between the mosquito larvae captured and the different macrophyte species at the 10 capture sites. Dunn's post hoc Multiple Comparisons Test was then used. Both analyses were performed using SPSS® statistics software (IBM 2015) Version 23 and Past (Hammer et al. 2001) Version 4.5.

The X² (chi-square) test was used to verify if there was an association between Mansonia larvae and plant species. In addition, the standard error of the association relationship was calculated.

The graphs of the mosquito population curves were established using Williams (geometric means) means (Williams 1937, Haddow 1960) and their 95% confidence intervals. We were able to observe the monthly frequency of immature individuals in each of the analyzed locations and compare periods of lower population abundances. To check for possible differences in water quality metrics, the t-test was used. The influence of climatic factors (water pH, dissolved O₂, conductivity, and temperature) on the population density of mosquitoes was analyzed with the Spearman's rank correlation coefficient, with a 95% and 99% confidence level, using SPSS (IBM 2015) Version 23.

During the sampling period, 31,013 larvae of Mansonia spp. were collected at 10 sample sites, 4 of which were in lotic and 6 in lentic environments. Though fewer in number, the lotic environments accounted for 58.9% of collected individuals, with the highest abundance of larvae found at the Igarapé Florida site (28.33%) and the lowest at Casa da Colina (1.32%). The X² test showed no equality of proportions based on given probabilities, as follows: X² = 144,450, df = 5, P < 2.2^e-16.

Mansonia larvae showed the highest abundance during March, April, and May 2016 and January, February, and December 2017, indicating a greater abundance during the rainy season. In contrast, the number of individuals recorded dropped sharply during the dry season, with an especially sharp drop in June 2016 and the lowest levels in September 2016; we also observed lower density in the dry months of the years 2017 and 2018 (Fig. 2).

Of the 12 species of macrophytes collected, 70.1% of all plants belonged to Eichhornia crassipes, which contained, on average, 96.2% of all Mansonia spp. larvae collected at each site, increasing to 100% at certain sites and never lower than 83.3% of all larvae. Ceratopteris pteridoides had the second-highest number of larvae, with 2.3% of the total larvae. The other 11 species of macrophytes accounted for the remaining 29.9% of the plants collected and just 3.8% of the larvae (Table 1). The highest density of Mansonia larvae was found on E. azurea, though the latter species was one of the least common (Fig. 3). This pattern could indicate a possible element of selection for this plant species. No Mansonia spp. larvae were found in any of the other aquatic plant species present, including Hymenachne sp., Oxicarium cubensis (Poepp. and Kunth) Lye, Ludwigia helminthorriza (Mart.) H. Hara, Phyllantus fluitans Benth. Ex Mull. Arg., Eleocharis sp., and grasses (Table 1).

The Foz do Igarapé Jirau site was the only one in which larvae were more evenly distributed between macrophyte species, with 83.3% of larvae found in E. crassipes, 9.2% in C. pteridoides, 3.6% in E. azurea, 2.0% in Salvinia sp., and 1.9% in Pistia sp. (Tables 1 and 2; Fig. 4).

The null hypothesis of the Kruskal-Wallis test was rejected (H = 33.767, df = 5, P = 0.000), and the analysis indicated significant differences between the number of larvae caught and the different macrophyte species across the 10 sites; aquatic plant species with zero larvae for all locations were not included. The test was performed for larvae found in E. crassipes, E. azurea, C. pteridoides, Pistia sp., Salvinia sp., and Paspalum sp.

Dunn's post hoc Multiple Comparisons Test indicated which species showed differences. The number of larvae caught in E. crassipes was significantly different from the number of E. azurea, C. pteridoides, Salvinia sp., and Paspalum sp. (P < 0.001). Meanwhile, larval populations did not differ in E. azurea, C. pteridoides, Pistia sp., Salvinia sp., and Paspalum sp. (P > 0.001).

There were some differences between sites in terms of the number of Mansonia spp. larvae collected each month. The total number of larvae of these mosquitoes tended to increase between August and December at the Igarapé Florida site. Meanwhile, at the Caiçara/Primavera site, January and the period between July and September had the main increases. In Igarapé Caiçara, the months of March and May were the most favorable. However, in the sample areas of Foz do Igarapé Jirau, Casa da Colina, Igarapé Raul, and São Lourenço, the highest total number of larvae was observed from January to May. The site at the Jaci River's mouth saw the highest numbers of larvae during December, January and April, while April and June saw the highest figures of larvae at the Dam Base site (Fig. 5).

The Spearman's correlation analysis indicated a significant positive correlation between the number of larvae collected in and pH at the Igarapé Florida and Casa da Colina sites. The number of larvae collected in E. crassipes was significantly and positively correlated with conductivity at Caiçara/Primavera, Dam Base, Igarapé Raul, and Bolsão Mutum. In addition, the number of larvae collected was negatively correlated with temperature at Caiçara/Primavera and Foz do Rio Jaci, meaning that there was a decrease in the population density of these larvae when temperature increased. However, at the Foz do Igarapé Jirau site, there was a positive correlation between the number of larvae found in C. pteridoides and temperature (Table 3).

The formation of large hydroelectric reservoirs completely alters the landscape, eliminating the floodplain and transforming the lotic system into a lentic system, which tends to favor the floating species of aquatic macrophytes (Resende et al. 2019). As a result, aquatic macrophytes form a microclimate that is much richer in oxygen (Raven 1977), and it becomes an excellent place for the development of several species of mosquitoes.

In the present study, it was observed that despite the lotic environments being represented by a lower number of sampling points, 4 and 6 in lentic environments, the collection points located in the lotic environment corresponded to 58.9% of the collected larvae; thus the rainy period in this region demonstrated the greatest abundance.

The aquatic macrophytes contribute with quality food to animals within the aquatic habitat (Gadelha et al. 1990) since they store various nutrients in their biomass (Moore et al. 1994). However, the high rates of infestation by aquatic plants in the reservoirs of hydroelectric power plants lead to the urgent need for control measures due to the plants' high capacity of propagation and biomass production (Martins et al. 2003).

In a study carried out by Ferreira (1999) on Marchantaria Island, in the municipality of Iranduba, Amazonas, the number of larvae and pupae found in E. crassipes was higher than in P. stratiotes. Ferreira et al. (2003) similarly demonstrated a strong association between Mansonia larvae and the macrophytes E. crassipes and Ceratopteris sp. in a study carried out in Lago Camaleão, Amazonas. These observations are consistent with the present findings, in which E. crassipes had a greater association with larvae than other macrophytes, followed by Pistia sp. Ferreira et al. (2003) developed hypotheses to explain these patterns, including the fact that the size of the aerenchyma in E. crassipes is larger than in Ceratopteris sp. Thus, Eichhornia can maintain higher levels of oxygen and consequently a larger number of larvae, while the root tissues of Ceratopteris sp. are more rigid, which makes it difficult for larvae to perforate the plant tissue; factors related to toxic secondary substances also play a role (Tokarnia et al. 1979) since Ceratopteris sp. may contain secondary substances that are harmful or repellent to Mansonia larvae.

In this study, we observed that two sample areas showed a positive correlation between the populations of Mansonia spp. and temperature. Ferreira et al. (2003) reported high water temperatures in Mansonia spp. habitat in their study area (29°C to 32°C) and discussed the possibility that temperature is related to the absence of Trichomycetes fungi in Mansonia spp. larvae.

Martins et al. (2003) evaluated the chemical composition of 4 important species of aquatic plants found in the reservoir of the Salto Grande hydroelectric power plant in Americana, São Paulo, Brazil. They found that E. crassipes and P. stratiotes had the highest average potassium, calcium, and magnesium levels. According to the results found by Martins et al. (2003) on the chemical constitution of macrophytes, the population density of Mansonia spp. in the study area may be influenced by the potassium, calcium, and magnesium content of the macrophages.

In this study, the populations of Mansonia spp. did not show a defined seasonal pattern throughout the year. In contrast, Ferreira et al. (1999) observed that the largest number of immatures of Mansonia spp. were found in August–October), with a peak in August, which was not significantly correlated with abiotic factors.

In conclusion, the larvae of Mansonia spp. were more abundant in sample areas with E. crassipes and Pistia sp. The number of specimens collected was positively correlated with temperature, pH, and conductivity. These correlations showed a marked increase in the rainy season. In parallel with the findings of Martins et al. (2003), it is also possible to consider that the presence of Trichomycetes fungi may influence the larval density of Mansonia spp. in the experimental area.

The findings of the present study are of great relevance since they contribute to the orientation of entomological surveillance and to the conduction of the application of mosquito control measures in the region.

This work was funded by Energia Sustentável do Brasil (ESBR) as an action of the Public Health Program (Convênio Jirau no. 141/2013).