Anomalous scales of Tillandsia usneoides ( L . ) L . ( Bromeliaceae ) exposed in the Metropolitan Region of Campinas , SP , Brazil as air pollution markers

(Anomalous scales of Tillandsia usneoides (L.) L. (Bromeliaceae) exposed in the Metropolitan Region of Campinas, São Paulo State, Brazil, as air pollution markers). Tillandsia usneoides is an epiphytic bromeliad that has been used as a universal bioindicator. The species accumulates metals and presents foliar scale variations when exposed to air pollutants. This study aimed to use the variations in foliar scales as microscopic markers of pollutant effects in the Metropolitan Region of Campinas (MRC), São Paulo State, Brazil. T. usneoides plants were exposed for 12 weeks during dry and wet seasons, totaling four exposures, at five sites in the MRC. Samples were selected before each exposure for initial evaluation of the plants (T0). Leaf fragments were fixed in glutaraldehyde and total scale density and anomalous scale percentage were evaluated. Plants exposed in the MRC showed anomalies and changes in number of subperipheral cells of leaf scales. When compared to T0, T. usneoides presented higher total scale density and anomalous scale percentage at disturbed sites, thus the species can be used as a pollution bioindicator for MRC.


Introduction
The Metropolitan Region of Campinas (MRC) is composed of 20 cities and accounts for approximately 15% of São Paulo State total population.The region economic growth initiated in the sixties with the implementation of large urban industrial complexes (Gutjahr 2004), as the city of Paulínia, which has a significant amount of mainly chemical and petrochemical industries (Tresmondi & Tomaz 2004).The MRC has been affected by high levels of primary air pollutants originated from road traffic and, is responsible for over 80% of CO, NOx and hydrocarbon emissions, and develops industrial activities that emit about 70% of the particulate matter present in the region (Ueda & Tomaz 2011).
Air pollution levels are determined by quantifying polluting substances in the air.The main air pollutants are carbon monoxide (CO), sulphur dioxide (SO 2 ), nitrogen oxides (NO x ), particulate matter (PM) and ozone (O 3 ) (CETESB 2013).Pollutants generated in the MRC and those carried by the south-southwest wind from the Metropolitan Region of São Paulo (MRSP) affect air quality in the region (Boian & Andrade 2012) and impact the existing vegetation of fragments of the seasonal semideciduous forest (Nascimento et al. 2006, Cielo-Filho & Santin 2002).
Studies have shown that structural markers allow evaluating the effects of pollutants in plants (Vollenweider et al. 2003).Changes in the stellate trichome morphology due to environmental agents have been described in the nickel-accumulating species Alyssum corsicum Rob.ex Gren.& Godr., when exposed to high concentrations of the metal (Babaoğlu Aydas et al. 2013).Trichomes of Ligustrum japonicum Thunb.and Olea europaea L. lost the original form when exposed to high concentrations of CO and SO 2 from automobile emissions (Abdallah & Mejnum 2013).
The genus Tillandsia is found throughout Latin America and has epiphytic habits, living independently of ground in trees or inert substrates and presenting absorbing trichomes on the entire leaf surface (Brighigna et al. 1997).Trichomes, also called scales (Benzing et al. 1992), protect the plant from UV radiation by reflecting most of the incoming light, absorb moisture from the air, retain dust and prevent the stomata from losing excessive water (Laboy-Nieves et al. 2010).Several studies have confirmed that Tillandsia genus plants are great bioindicators of gaseous pollutants (Benzing et al. 1992, Figueiredo et al. 2001, 2004, 2007) and are highly capable of accumulating many metals (Calasans & Malm 1997, Amado-Filho et al. 2002).
T. usneoides is an atmospheric epiphyte that depends exclusively on the atmosphere for survival (Brighigna et al. 1997).It extracts mineral nutrients and water from the rain and moisture in the air (Benzing et al. 1976), thus the species is considered a universal accumulator (Arndt & Schweizer 1991) and is used for metal biomonitoring.However, Alves et al. (2008) reported the presence of anomalies in T. usneoides scales exposed to polluted areas of São Paulo city, suggesting that anomalous scales are an alternative bioindicator parameter.
The study aimed to detect changes in the scales of Tillandsia usneoides exposed to disturbed environments, providing additional information about its use as air quality bioindicator, and generate information about the impact of air pollutants on the atmosphere of the Metropolitan Region of Campinas (MRC).

Study area and plant material
The MRC is located in an intertropical zone with predominantly humid subtropical climate, characterized by two distinct seasons: wet period with high temperatures and dry period with mild temperatures and less rainfall (Rolim et al. 2007).Between seasons, exist significant differences on the intensity and direction of the winds, which are predominantly from Southeast to Northwest (Domingos et al. 2015).The local climate is defined as prevalent Cfa (humid subtropical zone with dry winter and hot summer) according to Köppen classification (Alvarez et al. 2014) or B1rB'4a type according to Thornthwaite's classification, which considers topography and altitude as direct influence on climatic elements (Rolim et al. 2007).
Plants of Tillandsia usneoides (L.) L. were obtained from a commercial producer located in Cordeirópolis, an agricultural area in the region, and held for six months in a greenhouse at the Botanical Institute of São Paulo, for acclimatization before exposures started.
Five sites were selected in the 10-km radius from the industrial complex of Paulínia (figure 1 .At the other study sites and in the greenhouse, passive SO 2 , NO 2 , and O 3 (except in the greenhouse) samplers were used (Ogawa 2001, Machado et al. 2008, Allen et al. 2010), and evaluated weekly (table 1).

Preparation of anatomical slides
Entire fragments of T. usneoides were submerged in 2.5% glutaraldehyde and, after remaining 12 hours in vacuum desiccator, were refrigerated at approximately 5 °C for further anatomical study.The fragments were washed in phosphate buffer Sorensen 0.067 M, pH 7, and transferred to 70% alcohol.The third apical parts of the leaves were cut lengthwise freehand with steel blades and immersed in a maceration solution of hydrogen peroxide 30 vol and glacial acetic acid 1:1, remained in an oven at 60 °C for 12 hours and subsequently washed with distilled water.The fragments were stained with 0.05% toluidine blue (Feder & O'Brien 1968) and assembled in 50% glycerin.Three blades were produced per plant sample.
Images of the scales were captured by Olympus microscope (BX53) equipped with Olympus camera Q color 5 and analyzed with the aid of Image Pro Express 6.3 software.For each exposure site and exposure period, 15 fields were assessed per blade (scales mm -2 ) and the total scale density and anomalous scale percentage (number of anomalous scales in relation to the total number of scales) were estimated in 1,080 fields.

Statistical analysis
The data were submitted to analysis of variance (ANOVA on ranks) and the averages were compared according to the Kruskal-Wallis method to establish differences between sites, and the Man-Whitman method was used to establish differences between dry and wet exposure periods.For multiple comparisons between the periods and the sites, the Dunm test (p < 0.05) was applied.The Spearman correlation test (p < 0.05) determined possible associations between pollutant concentration variations and the analyzed parameters (total density of scales and percentage of anomalous scales).

Results and Discussion
The environmental data obtained in PC, when compared with those provided by other monitoring stations, showed little variation in the concentrations of pollutants and climatic parameters, with the exception of precipitation.Therefore, the monitoring station at PC appropriately characterizes the study region and can be used as a reference to the correlation analysis (table 1).Studies conducted in the region (Tresmondi & Tomaz 2004, Boian & Andrade 2012, Bulbovas et al. 2015) also showed little spatial variation in air pollution levels in the MRC, confirming the representativeness of the monitoring station in Paulínia Center for the study area.
The surface of Tillandsia usneoides leaves is covered with scales that, in frontal view, consist of a central disc formed by four cells surrounded by two sets of cells: the first consists of eight slender cells, called pericentral cells, and the second has sixteen cells and is called subperipheral.Externally, there are empty, elongated, thin-walled and radially arranged peripheral cells, forming a shield that is called wing (figure 2a-f).The scale wings (figure 2a) exhibit a confluent layering, forming a space between them and the epidermal surface (Benzing et al. 1976, Souza & Neves 1996, Proença & Sajo 2007).
The anomalies observed in the scales of plants exposed in the MRC were: change in the number of cells in the subperipheral layer, less (figure 2c) or more (figure 2d) than 16 cells and addition of cells between the subperipheral layer of the disc and the wing (figure 2e).Changes were also observed in cell shapes in different layers, resulting in fully deformed scales (figure 2f).According to Cardoso-Gustavson et al. (2015), changes in the number of cells of T. usneoides scales exposed to pollutants are related to occurrence of mutagenic processes.The authors found that, during the development of scales, endoreduplication cycles could occur in the nucleus without cell division, resulting in reduced number of subperipheral cells with large nucleus.
Plants exposed in PC and PA showed a significant increase in the density of total scales in the dry period and showed no significant variation in the wet period, when compared to T0 plants.The highest percentage of anomalous scales was observed in plants exposed at PC in the wet period, compared with the T0 plants and did not vary significantly in the dry period.
Total scale density and anomalous scale percentage were significantly different comparing the dry and wet exposure periods at each site.Plants exposed in PA presented higher scale density and anomalous scale percentage in the dry period.At the other sites, plants exposed in the wet period had higher scale density and anomalous scale percentage than those exposed in the dry period.The positive association of the percentage of anomalous scales with moisture is related to greater interaction of scales and pollutants in the presence of water, which facilitates diffusion of pollutants, increasing acidity in intercellular spaces (Larcher 2000) and that is intensified in T. usneoides since the species is an atmospheric bromeliad (Benzing et al. 1976).Alves et al. (2008) observed a significant increase in the total density of scales in T. usneoides exposed at sites in São Paulo city after 24 weeks, compared with the control material.The studied sites were a park with high concentrations of O 3 and an area of intense traffic, with high concentrations of SO 2 .In plants of Pueraria lobata Will, Sharma et al. (1980)  also observed higher frequency of trichomes due to increased amount of pollutants.Quercus ilex L. exposed to high CO 2 concentrations presented higher trichome density (Paoletti et al. 2007); although some authors do not consider CO 2 as a pollutant (Larcher 2000).JA, HO and CA are under greater agricultural influence, while PA and PC are highly influenced by urban and industrial pollution sources.Higher exposure to pollutants from sources such as SO 2 and PM 10 can justify the increase of total scale density at the sites, as well as increased anomalous scales in PC.
In this study, the total scale density and anomalous scale percentage were larger than those observed by Alves et al. (2008), and the plants exposed in plants at PC presented twice as many anomalous scales as the authors observed in the same species.
The statistical analysis showed higher percentage of anomalous scales in plants exposed at Paulínia (PC); therefore, with the aid of the Spearman correlation test, we established the relationship between the anomalous scale percentage, total scale density of plants exposed in Paulínia and environmental parameters (table 2).The percentage of anomalous scales in plants exposed in PC correlated negatively with SO 2 and PM 10 concentrations and positively with relative humidity (RH), whereas, total scale density showed no significant correlation with any of the environmental parameters.These results showed that the increasing concentrations of PM 10 and SO 2 in the region did not influence anomalies of the scales.Cardoso-Gustavson et al. (2015) did not observe significant correlation between the formation of anomalous scales in plants of T. usneoides and the presence of trace metals, such as Ba, Fe and Cu, normally present in PM 10 , suggesting that anomalies would be related to other air pollutants from vehicular source, as NO 2 and SO 2 .
The plants exposed in Paulínia presented the highest percentages of anomalous scales (table 3), differing significantly from T0, but showed no significant correlation with pollutants from urban sources.However, Alves et al. (2008) observed relationship between anomalous scale increase in T. usneoides plants and the presence of pollutants in places with high levels of O 3 in the MRSP, when compared to the rural environment.O 3 and SO 2 promote little visible damage in the epiphytic bromeliads with CAM photosynthesis, when compared to C3 plants, due to the presence of thick cell walls and the crown of dead cells (wings of scale) which covers the leaf surface, protects the stomata, promotes greater tolerance to pollutants and functions as primary physical barrier (Benzing et al. 1992).
T. usneoides is a species that accumulates metals in particulate matter, (Calasans & Malm 1997, Figueiredo et al. 2007), due to its adsorption on the scales (Amado-Filho et al. 2002), which makes it more tolerant to PM 10 .This assumption has been confirmed by the results obtained in this study, considering that there was negative correlation between anomalous scales and PM 10 .
The largest emissions of pollutants in the MRC, especially particulate matter, NOx, CO and SO 2 , come from industrial and automotive traffic sources (Ueda & Tomaz 2011), which is in accordance with the largest percentages of anomalies in PA and PC (table 3), the  Studies carried out with T. usneoides in the same cities of the MRC showed low or no impact of air pollutants on physiological (as chlorophyll contents and plant mass loss) or biochemical (antioxidant) activities (Giampaoli et al. unpublished data), indicating that the tolerance of the species is related to the primary barrier represented by trichomes, as observed in the density variation and/or structure changes in other species (Moura 2013).
T. usneoides responded to environmental stress with high presence of anomalous scales and total scale density in plants exposed in disturbed regions, mainly at PC and PA.
Figure 1.Map distribution of exposure sites of Tillandsia usneoides plants in Metropolitan Region of Campinas (MRC), São Paulo State, Brazil.

Figure 2 .
Figure 2. Paradermic surface of Tillandsia usneoides leaf, highlighting scales with normal pattern (a-b) and anomalous scales (c-f).a. Epidermis with scales.b.Highlighting scales four central cells (CC) forming a central disc, eight pericentral cells (PC), 16 subperipheral cells (SP) and peripheral layer forming the wings (WI).c.Reduction of three cells in the subperipheral disk (arrows); d.Addition of one cell in subperipheral disk (arrow); e. Addition of six cells between the subperipheral layer and the wing; f.Shape and number of cells of scale modified in all layers, including the central disk.Bars = 20 µm.

Table 1 .
Environmental data of Metropolitan Region of Campinas (MRC).Average data from CETESB monitoring stations (Paulínia Downtown and Campinas) combined with values of local passive samplers.EA and EC: dry periods; EB and ED: wet periods; PC (data from Paulínia-Downtown CETESB monitoring station).

Table 2 .
Alves et al. (2008)ollutant concentrations and environmental data provided by CETESB station-Paulínia Downtown and anomalous scales in leaves of Tillandsia usneoides at PC (Paulínia Downtown).-correlationcoeficient;p-statisticalsignificanceclosestpoints to the industrial complex region.In the study ofAlves et al. (2008), T. usneoides plants were exposed continuously for 8, 16 and 24 weeks; changes in the scales became visible after 8 weeks of exposure.In the present study, the plants were exposed for 12 weeks and no direct relationship between O 3 concentrations and percentage of anomalies was observed.The studied region encompasses different sources of air pollutants and climatic regime comparing with the sites studied byAlves et al. (2008)in the city of São Paulo.The difference among dry and wet periods is more pronounced in the MRC than in São Paulo(Alvarez et al. 2014), suggesting that the variations in density of the scales and occurrence of anomalies can be induced not only by pollutants, but by weather conditions as well.