Air pollution monitoring with Tradescantia hybrid and optical sensors in Curitiba and Araucária, Brazil

Complex mixtures of organic and inorganic substances (such as sulfates, halogenated and nitrated organic compounds, polycyclic aromatic hydrocarbons (PAHs) and strong acids) are apparent in the urban atmosphere due to high levels of pollution, which can be potentially mutagenic and carcinogenic, bringing damage to the environment and human health (Luiz et al., 2005; Gábelová et al., 2004; de Brito et al., 2013). Therefore, the monitoring of these substances becomes important, as it provides quantitative information on the compounds present in the atmosphere. The clastogenic or mutagenic effects that such mixtures may have on biota and human health cannot be measured by methods based on pollutant concentration. Thus, biomonitoring becomes necessary to assess the genotoxic potential of the components present in the air.

Biomonitoring uses animal, plant or fungal species continuously exposed in environments to assess the effects that pollutants can have on living beings. The Tradescantia plant is an example of a bioindicator of genotoxicity and is sensitive to mutagenic chemical compounds present in the air, therefore it has been widely used to detect such agents and their possible effects (Ma et al., 1994; Guimarães et al., 2000; De Luccia, 2012). Schairer et al. (1978) performed exposures of Tradescantia clones 4430 to the air pollutants SO₂, NO₂, and O₃ and to vapors of mutagens such as 1,2-dibromoethane (DBE) and ethyl methanesulfonate (EMS). Their results demonstrated the usefulness of the system as a detector of chemical mutagens. Following exposure to either chemical or physical mutagens, the flowers were analyzed in the laboratory each day as they bloom for approximately three weeks after treatment. Pink mutation rates in Tradescantia stamen hairs following a 6-hour exposure to gaseous EMS reach peak values 7–12 days after exposure (Schairer, 1978). In the literature, this value ranged from 10 to 21 days for chronic exposure to radiation. After this period, the highest values of mutations were found in the analyzed flowers matching with values found in the work of Takahashi & Ichikawa (1976), Ichikawa (1981) and Ichikawa et al. (1996). In their dose-effect experiments on exposure to different radiation rates, it was observed that mutations were more frequent when exposed to environmental factors about two or three weeks before flowering. Patussi & Bündchen (2013) used specimens of Tradescantia sp. clone 4430 to evaluate the genotoxicity of herbicides in a corn crop, where pots with the plant were placed between the plantation during herbicide application and analyzed later to verify if mutations occurred. Their final result is that more alterations were observed mutagenically in the stamen hair of exposed plants compared to the control, thus demonstrating the sensitivity of the biomonitor to harmful agents.

The negative effects that pollutants can cause to plants depend on the concentration and time of exposure and cause alterations in the plants’ gas exchange process, leaf necrosis and chlorosis, hamper plant growth and in some cases the suppression of a species in the local flora (Rascio & Navari-Izzo, 2011). Gaseous fractions of air pollution can be absorbed by the stomata of the leaves, particulate matter can be retained on the surface of the plant and can be resuspended by wind or further deposit from the leaves to the ground. The fraction of healthy leaf area, concentration of pollutants and meteorological variables are some of the factors that affect the self-removal of pollutants by plants (Nowak & Dwyer, 2007). The current global trend in urbanization (UN, 2019) increases the pressure on natural and urbanized ecosystems globally.

Long exposure to particulate matter causes adverse effects on human health and can lead to the premature death of millions of people yearly (WHO, 2021). Some diseases caused by this exposure are: cardiovascular diseases, lung cancer, respiratory infections, cerebrovascular accident (CVA) and depression (Vodonos et al., 2018; Alexeeff et al., 2021). The World Health Organization estimated that air pollution caused the premature death of 4.2 million people worldwide in 2016 and a study by Southerland et al. (2022) estimated that 4.14 million people died as a result of air pollution caused by PM_2.5 particulate matter in 2019. It is noted that mortality attributable to air pollutants remained relatively unchanged between 2000 and 2019 in urban areas globally (Southerland et al., 2022).

Particulate Matter (PM) can be considered as a proxy for air pollution and genotoxic effects on plants and negative effects on human health. These particles are subdivided into PM_2.5, particles with an aerodynamic diameter of ≤ 2.5 μm, and PM₁₀, particles with an aerodynamic diameter of ≤ 10 μm. In the work of Carreras et al. (2006) PM was considered as a proxy for the total load of pollutants in the air and through the bioassay Trad-MCN it was found that air pollution caused spontaneous changes in the micronucleus of the pollen mother cells. In the study carried out by Puett et al. (2014), a positive relationship was found between exposure to particulate matter and the increased incidence of lung cancer in women, especially among non-smokers. There is strong evidence of causal relationships between PM_2.5 air pollution exposure and human mortality, as well as acute lower respiratory infections, ischemic heart disease, lung cancer, chronic obstructive pulmonary disease and stroke (Cohen et al., 2017; WHO, 2021).

This study aims to understand the air pollution and mutagenic dynamics of the bioindicator Tradescantia sp. clone 4430 throughout 2020 and the beginning of 2021 in Curitiba and Araucária, cities located in southern Brazil. Thus, this work seeks to evaluate the relationship between mutations in stamen hair from samples of flowers of Tradescantia sp. clone 4430 and PM₁₀ and PM_2.5 measurements made with an optical sensor distributed at eight locations in the study region. We propose a method for estimating the lag time and the exposure window before inflorescences sampling which might help to understand the response of the bioindicator plant to air pollution for the appearance of mutations in the stamen hairs. Few studies were found in the literature to estimate the response time to mutagenic effects in Tradescantia sp., therefore, this research brings novelty and contribution to the area of plant ecophysiology and environmental toxicology.

The study was carried out in two municipalities – Curitiba and Araucária – which are located in the eastern region of the State of Paraná, in Brazil (Figure 1). The estimated population for Curitiba is 1,948,626 and for Araucária 146,214 inhabitants. (IBGE, 2020). The main economic activities of these two municipalities are commerce and industry, in addition to having together a fleet of approximately 1,539,491 vehicles, which contributes to the release of particulate matter and gases into the environment, thus interfering with air quality, public health and the local biome (IPARDES, 2019). According to the Köppen climate classification, the municipalities have a Cfb climate (temperate climate proper), with well-distributed rainfall, mild summer, occurrence of frost (autumn and winter), average temperature below 20°C (except summer) and no defined dry season (Nitsche et al., 2019).

Figure 1.

Inflorescence sampling and laboratory analysis

The Trad-SHM assay procedures adopted in this research followed an adaptation of the procedures of Ma et al. (1994) and Rodrigues (1999), where the plants followed a chronic exposure, i.e. the pots remained in the study sites throughout the time of the research and not for an acute exposure as suggested by Rodrigues (1999). Figure 2 illustrates the biomonitor used and its main parts. The inflorescences of Tradescantia sp. clone 4430 were scheduled to be collected weekly from January 2020 until February 2021 and in each collection all inflorescences were removed by cutting their peduncle. After cutting, the branches were identified according to their sampling point and taken to the laboratory. The flowers were collected in the morning, because during the day they close and wither. If the evaluation was not possible on the same day of collection, two procedures were adopted: the first was to collect the flowers and keep them cooled in pots with lids and moistened filter paper and analyze them up to 24 hours after harvesting, and the second was to collect the flowers and remove the stamens with the help of tweezers and mount the slides, taking care to leave the stamens aligned on the slide; an adhesive tape was used to make the adherence of the lamina on the slide; these finished slides were also kept in refrigeration until the analysis, preferably within 24h (Rodrigues, 1999).

Figure 2.

To count the cells, all the flowers collected from each point were used, six stamens were removed with the help of tweezers and arranged on glass slides. A 1:1 solution of 70% alcohol and glycerin was used to fix the biological material. With the help of tweezers and a needle, the hairs were aligned on the slide and observed under a magnifying glass. Then, a coverslip was superimposed and the material was analyzed with the aid of an optical microscope. All the pink cells present in the stamen hair and also the total number of stamen hairs contained in the stamen were counted. If there was no pink cell in the observed stamen hair, only the total number of filaments (stamen hair) was counted. The mutation frequency per flower was then expressed as the number of events/1000 stamen hair.

Backward Sliding Window (BSW) Method

The Backward Sliding Window (BSW) Method was developed by the authors and applied for detecting two parameters related to the chronic exposure of Tradescantia sp. clone 4430 to atmospheric pollution. The parameters are: i) size of the window of exposure, in days; and ii) lag time for mutation appearance after the exposure, counted as days before inflorescence collection, as shown in Figure 3. The method consists of varying the number of days of exposure and the position of the window of exposure to particulate matter, our proxy for air pollution. For different exposure periods the correlation between the mean PM concentration and number of mutations/1000 cells is calculated. The period with the highest Pearson correlation coefficient was chosen for the analysis.

Figure 3.

After choosing a reasonable window size and the lag time, a dataset was constructed relating the number of mutations in stamen cells to the mean PM₁₀ and PM_2.5 concentration in the backward exposure window. Only sites with more than 16 inflorescence sampling days were chosen for this analysis, as it was considered that few collection days (≤ 16) would not be statistically significant. The Pearson correlation coefficient between PM and Tradescantia mutations were calculated for the sampling points.

Daily profiles with hourly average data of particulate matter were calculated to compare the dynamics throughout the day and night between the sampling points. It is known that in the Araucária sampling point are several industries that contribute to the emission of particles and in Curitiba the main pollution comes from motor vehicles.

The collection of flowers of Tradescantia sp. clone 4430 for the monitoring points occurred at different dates, quantities and frequency. In the laboratory analysis, all pink cells were counted and some were photographed. Figure 4 shows some examples of pink mutation in stamen hair of Tradescantia sp. clone 4430 found during the analysis of the Trad-SHM bioassay, where the stamen hair cells that present a pink coloration can be seen.

Figure 4.

Figure 5 shows the boxplot of PM₁₀, PM_2.5 daily concentrations and mutation frequencies in stamen hair of the sample points, including the control point.

Figure 5.

Figure 5(c) shows the frequency of mutations in stamen hair for all points studied, different behaviors were observed between them, which might be associated with the frequency and quantity of flowers collected not being similar between sampling points (the number of sampling days “n” are indicated at the top of each boxplot). Figures 5(a) and 5(b) indicate more similar n values for particulate matter measurements, but some stations had less data because the sampling started later, like Guaíra and the control point. In addition, these graphs show the outliers as dark diamonds and the averages are indicated as white circles. Unlike the control point, the medians of the other points were closer to the first quartile, thus presenting a positive asymmetry, since the averages were all above the median, therefore leaving the values more concentrated at the bottom of the boxplot. Also in Figure 5(c), Mercês and Jardim das Américas sampling points presented greater dispersion in relation to the others, due to the difference between the third and the first quartile being greater in these two points, in the Orleans point, it was where the smallest variation of the data occurred.

The particulate matter concentrations were collected by the SDS011 sensor from January 2020 to February 2021. Some stations had few inflorescence sampling days, which would affect the correlation calculation against particulate matter measurements. To circumvent this limitation, a criterion of at least 17 days of inflorescence sampling was adopted. The selected stations for analysis were: Jardim Botânico, Jardim das Américas, Mercês and Orleans. The data was processed, daily averages were calculated for all stations and the data overview given in Figure 6. During 2020 and beginning of 2021 our data shows an increase of mutations in the winter period, between May and November of 2020, coincidently the air quality was also worse between May and October of 2020. It is known that during the coldest months of the year normally there is less dispersion of pollutants in the lower atmosphere. This process is related to the planetary boundary layer height variation, changes in incident solar radiation and meteorological variables, such as rain and wind patterns.

Figure 6.

In 2021 the WHO presented guidelines for the guide values for particulate matter and other pollutants that are harmful to health. This update is due to growing evidence of the adverse health effects of air pollution, based on advances in the measurement of pollutants and assessment of human exposure to them (PAHO, 2021). The goal of the guideline updates is to provide quantitative recommendations and guidance for minimizing air pollutant levels in order to reduce the damage caused by exposure to pollutants (PAHO, 2021). In comparison with the last report from WHO, the value for PM_2.5 has become more restrictive because it is a fine particulate that is easily inhaled and can penetrate deeply into the lower airways, reaching the lung alveoli and bronchioles (SISAM, 2019).

The red and purple dashed horizontal lines in Figure 6 refer to the limits imposed by the World Health Organization (WHO, 2021), where the daily average level of PM₁₀ and PM_2.5 must not exceed concentrations of 45 μg/m³ and 15 μg/m³, respectively. During the data collection period and considering all the sampling points, 25 days exceeded the limit for PM₁₀ and 202 days exceeded the limit for PM_2.5, not necessarily being the same days. The station that most often exceeded the WHO limits was Araucária (14 days for PM₁₀ and 65 days for PM_2.5), probably due to its location close to the city’s industries and also because there are roads with greater vehicle traffic in this region. This fact was observed by Castelhano (2019) who noticed a pattern of PM concentration being higher in Araucária compared to its surroundings. In Curitiba metropolitan area, the sampling points Jardim das Américas, Batel, Jardim Botânico, Mercês exceeded PM_2.5 levels during 28, 27, 23, 20 days, respectively. The control sampling point did not exceed any of the limits.

Figures 7(a) and 7(b) show the daily dynamics of PM₁₀ and PM_2.5 concentration, respectively, for all sampling points, except the control. Both graphs represent the hourly mean and 5 times the standard error (SE) for the period from January 2020 to February 2021. Araucaria sampling point had the highest PM₁₀ and PM_2.5 concentration for most of the hours of the day, with higher concentrations from 15h to 22h. Jardim das Américas, Batel, Botânico and Guaíra stations had a similar behavior with higher concentrations near the busiest vehicle traffic times, from 6h to 9h and from 18h to 21h. Concentration by time of day can give insight into possible sources of pollution.

Figure 7.

Mutation frequency and PM correlation

Figure 8 represents the application of the BSW Method with the correlation results between mutations and mean PM concentration for the selected sampling points. We found that a 6-day exposure window beginning 2 days before inflorescence sampling resulted in the highest correlation between PM₁₀ and PM_2.5 mean concentration and cell mutations, considering the 4 selected stations (Jardim Botânico, Jardim das Américas, Mercês and Orleans).

Figure 8.

Figure 9 shows the window size (6 days) and lag time (2 days) framework considered for Tradescantia sp. clone 4430 analysis of chronic exposure to air pollution.

Figure 9.

In order to verify whether air pollution, represented by the particulate matter indicator, is capable of causing changes in the genetic material in plants, specimens of Tradescantia sp. clone 4430 were subjected to chronic exposure from January 2020 to February 2021, and the mutation frequency in stamen hair was compared with the concentration of particulate matter PM₁₀ and PM_2.5, collected during the same period.

As already mentioned before, a lag time of 2 days before inflorescence sampling was used to calculate the mean value of particulate matter in the 6-day exposure window, so the statistical correlation between PM and mutations was calculated for all the selected stations. The results of the Pearson correlation (r) test between mutations and PM₁₀, PM_2.5 data from the selected monitoring points are shown in Figure 10.

Figure 10.

Figure 10 shows the r values of the Pearson correlation, showing a positive relationship, with the r values ranging from a weak correlation to a strong correlation. In the work by Guimarães et al. (2004) the Trad-SHM bioassay was performed with the KU-2 clone to evaluate the toxicity of PM₁₀ in two cities of São Paulo (capital and country side) and the mutation frequency observed in the flowers of the country side city was lower in relation to the capital, where the mean PM₁₀ in the capital is 64μg/m³ against 14μg/m³ in the interior and the correlation between PM and mutations was positive (r = 0.47).

The present work shows the procedure of data collection and evaluation of particulate matter concentration in Curitiba and Araucária, distributed in eight points plus the control point, as well as the mutation frequency in stamen hair by the Trad-SHM bioassay in Tradescantia sp. clone 4430. Furthermore, this research investigated if the concentrations of PM impact the result of the Trad-SHM bioassay. The particulate matter was considered a proxy for air pollution (Carreras et al., 2006; Puett et al., 2014).

Daily mean concentrations of PM at the monitoring points exceeded the limits specified by WHO (2021). PM_2.5 is more critical than PM₁₀ as there were 16 days during the entire studied period when the mean PM_2.5 daily average was bigger than 15μg/m³, considering all the sampling points. This happened mostly in seasons when atmospheric dispersion is less favorable, thus causing a higher concentration of pollutants in the atmosphere. Therefore, there is an urge to increase monitoring and to detect the most critical points so that political actions can be taken.

There was an occurrence of mutation frequency in stamen hair at all points studied, being more frequent in the periods between autumn and spring. For the same period of the particulate matter, a tendency can be noticed between the results. Comparisons between PM concentration and mutations in stamen hair of Tradescantia sp. clone 4430 showed that the results followed the same trend of increase and decrease in relation to different periods of the year.

The correlations between particulate matter and mutation frequency showed positive values, ranging from weak to strong correlation. The mean correlation between particulate matter and mutation frequency was r = 0.53 for PM₁₀ and r = 0.43 for PM_2.5 (moderate correlation), with a maximum value of 0.61 and a minimum of 0.31. The use of Tradescantia sp. clone 4430 with the Trad-SHM bioassay showed sensitivity to the environments in which it was exposed, and this study reveals that biomonitoring may be an important tool for understanding the effects of pollutants on the ecosystem, especially facing increasing urbanization globally. We suggest for future work the comparison of mutation frequency with the concentration of other polluting gases, such as CO, SO_x, O₃ and NO_x.