Characterization of Indoor-Air Bioaerosols in Southern Taiwan

This study investigated indoor air quality (IAQ) at 39 public sites in southern Taiwan including hospitals, schools, office buildings, hypermarkets, libraries, railway stations, theaters, etc. Indoor air quality was preliminarily assessed using handy digital apparatus. Items detected include carbon dioxide (CO2), carbon monoxide (CO), formaldehyde (HCHO), total volatile organic compounds (TVOCs), total bacteria counts (TBC), total fungi counts (TFC), PM10, PM2.5, ozone (O3) and temperature. Based on the results of walk-through detection, the spatial distribution of indoor air contaminants was further measured over a 24 hour period using the EPA standard method. Major indoor air pollutants were found to include CO2, TBC, and TFC. The measured CO2 concentrations ranged between 438 and 1527 ppm, and only 38.9% of them met the Taiwan EPA suggested threshold of 600 ppm. In the schools and hospitals (Category 1), the measured TFC and TBC concentrations ranged from 62 to TNTC CFU/m and from 196 to 4875 CFU/m, respectively. 33% TFC and 83% TBC concentrations exceeded the suggested threshold, and CO2 concentrations were moderately correlated with TBC levels. In a case study of hospital bioaerosols, high TBC and TFC levels were effectively lowered through disinfectant housekeeping as well as ClO2 spray. Three filamentous fungus genera were identified as Cladosporium perangustum, Cladosporium tenuissimum, and Fusarium incarnatum from outdoor samples with high TFC concentrations.


INTRODUCTION
Life style, dietary habit and customer behaviour may directly influence in indoor air quality (IAQ) and the characteristics of indoor air pollutants.In general, people spend at least 80% of their time in various indoor spaces.Even though indoor air pollutants are present at trace concentrations, the potential threat to human health cannot be overlooked after long-term exposure in indoor environments.Poor IAQ has been shown to have adverse effects on human health (Fabian et al., 2005;Huboyo et al., 2011).To improve IAQ and protect human health, the Taiwan EPA listed the suggested threshold levels of common indoor air pollutants in 2005.The suggested IAQ measurements include carbon dioxide (CO 2 ), carbon monoxide (CO), formaldehyde (HCHO), total volatile organic compounds (TVOCs), total bacteria counts (TBC), total fungi counts (TFC), PM 10 , PM 2.5 , ozone (O 3 ) and temperature.Regulated public sites including schools and hospitals comprise category 1, and stores, markets, office buildings, and transport stations comprise category 2 (Taiwan EPA, 2005).
Sources of indoor air pollutants may be brought about by contaminated outdoor air, air conditioning or office equipment, human activity, building components and furnishings, and other accidental events (USEPA, 1991;Lai et al., 2010;McNamara et al., 2011).Particulate matter frequently couples with compounds of biological origins creating bioaerosol that may ranges in aerodynamic diameter from 0.01 and 100 m (Cox and Wathes, 1995;Macher, 1999).Indoor air contains a complex mixture of bioaerosols, such as bacteria and fungi, and non-biological particles such as dust, tobacco smoke, and cooking-generated particles (Kalogerakis et al., 2005;Cao et al., 2011).The prevailing microorganisms in the air of closed spaces are bacteria, and the majority of bacteria sources are come from the presence of humans in non-industrial indoor environments (Stezenbach, 1997;Kalogerakis et al., 2005;Singh et al., 2011).The predominant microorganism in ambient air is fungi, and the majority of fungal sources are soil, plants, and water bodies outdoors (Karbowska-Berent et al., 2011).Fungal spores have tiny aerodynamic diameters, high environmental durability, and are carried by air currents (Macher, 1999).The spore transport diffusion mechanism was controlled by meteorological conditions such as dew point, relative humidity (RH), temperature, and airflow turbidity.The level of airborne fungi is closely associated with indoor RH, and the optimal condition for fungal growth may be above 70% RH (Burge et al., 1995).The presence of bioaerosols with higher concentrations was observed in the optimum temperature range of 30 to 40°C (Deguchi and Yoshizawa, 1996).
Because Taiwan is a warm humid island, bioaerosol concentrations are relatively high (Tseng et al., 2011).An average indoor bacteria concentration of 1502 CFU/m 3 and indoor fungi concentration of 195 CFU/m 3 were reported for Taipei (Li, 1998).The predominant allergens in Taiwan may be caused by indoor airborne fungi that usually originate outdoors (Wu et al., 2000).A case study in Singapore illustrated that bioaerosols consist of 50.5% bacteria and 49.5% fungi indoors, and 20.6% bacteria and 79.4% fungi outdoors (Rajasekar and Balasubramanian, 2011).According to the hygroscopic effects of airborne fungi in Taiwan, outdoor airborne fungi (2.58 m in summer, 2.79 m in winter) have greater geometric mean diameters than indoor airborne fungi (1.91 m in summer, 1.73 m in winter).Indoor/outdoor (I/O) ratios of airborne fungal concentrations are between 0.29 and 0.58 without indoor sources during the summer, while I/O concentrations drop to 0.12 and 0.16 during the winter (Liao et al., 2004).Airborne bioaerosols including bacteria and fungi can be toxigenic, allergenic, and infectious (Fabian et al., 2005).Human exposure to airborne fungal spores might cause adverse health effects, especially respiratory symptoms (Su et al., 2001).More than 80 fungal genera were reported to be linked to symptoms of respiratory tract allergies (Horner et al., 1995).Cladosporium spp., Alternaria spp., Aspergillus spp., and Fusarium spp.are amongst the most common allergenic genera, and metabolites of fungi are also believed to irritate the respiratory system (Kalogerakis et al., 2005).Accordingly, bioaerosol characterization has become an important issue due to its related health effects.This study was aimed at discovering common indoor air pollutants, examining the characteristics of indoor bioaerosols, and evaluating the relationship between indoor air pollutants and bioaerosols based on IAQ investigation of public sites from the southern Taiwan.

Study Area
According to the Taiwan EPA's suggested threshold levels for common indoor air pollutants, regulated public sites are classified into two categories: category 1 includes hospitals, health care centres, schools, tutoring centres, kindergartens, and nursing centres, and category 2 includes office buildings, banks, restaurants, supermarkets, theatres, stores, and transport stations.This study selected 39 public sites in Tainan city including 8 hospitals, 7 tutoring centres, and 3 kindergartens to represent Category 1, and 5 supermarkets, 2 transport stations, 10 office buildings and 4 theatres to represent Category 2, and investigated their IAQ.

IAQ Investigation
An IAQ survey was implemented in 2 stages, and the reported investigation data was collected during 2009 and 2010.In the first stage, various handy digital apparatuses were utilized to preliminarily screen IAQ hot-spot locations in each selected public site through walk-through detection.Levels of CO 2 , CO, and O 3 were quantified using a KD Engineering AirBoxx detector.TVOC concentration was detected using a ToxiRAE PGM-30 PID detector, and HCHO concentration was detected using a PPM Technology htV-M formaldehyde meter.Levels of PM 10 and PM 2.5 were quantified using a Met one Instruments AEROCET 531 detector.Based on the space arrangement, 8 to 20 locations were selected for IAQ walk-through detection.A single detection was conducted for 2 min at a height of 1.2 to 1.5 m.According to the first stage IAQ survey, the concentration contours were used to examine the distributions of indoor air pollutants at each site and to identify the IAQ hot-spot locations among the investigated public sites (Wang et al., 2011).
The identified IAQ hot-spot location was subsequently subjected to the second stage IAQ survey.Following Taiwan EPA standard methods, the indoor air pollutants were continuously monitored for 24 hours or sampled for lab analysis.Concentrations of CO 2 and CO were measured using the non-dispersive infrared method, and O 3 was quantified using the ultra-violet absorption method.TVOCs were sampled using a canister and analysed using a HP 6890 gas chromatography equipped with a FID detector (GC/FID).The quantitative analysis of HCHO applied 2,4dinitrophenylhydrazine (DNPH) absorption and was analysed using HPLC.Concentrations of PM 10 and PM 2.5 were determined using the -ray decay method.The measurements of indoor air quality by using handy digital apparatuses and the Taiwan EPA standard methods have been compared by Kung et al. (2011).

Bioaerosol Analysis
The numeration of TBC and TFC were sampled and analysed based on the Taiwan EPA's standard methods, NIEA E301.11C and NIEA E401.11C, respectively.Air samples were suctioned with a SKC Quick TAKE 30 impacter sampler with 400 holes 0.25 mm in diameter.The drawing air-stream had a flow rate of 28.3 L/min and was directed onto an agar surface in a standard Petri dish.Tryptic soy agar (TSA) consisting of 15 g/L tryptone, 5 g/L soytone, 5 g/L NaCl, and 15 g/L agar was used for the bacterial collection.Malt extract agar (MEA) consisting of 12.75 g/L maltose, 2.75 g/L dextrin, 2.35 g/L glycerol, 0.78 g/L peptone, and 15 g/L agar was used for fungal collection.The TSA and MEA mediums were sterilized before they were placed in an autoclave operating at 121°C and 1.2 kg/cm 2 for 20 min.The incubation periods were 48 hours at 30°C for bacterial growth and 4 days at 25°C for fungal growth.Bioaerosol concentrations as CFU/m 3 were calculated by dividing CFU counts by the sampling flow rate and sampling time using positive-hole conversion tables.
According to the appearance, colour, and distribution of fungal growth on the agar plate, flourishing colonies were selected for the identification of fungal genera by the Bioresource Collection and Research Centre.The identification procedures were followed by genus selection, separation, purification, microscopic observation, and identification using the rDNA ITS1-5.8S-ITS2sequence.

Measurements of Indoor Air Quality
The results of the IAQ investigation at 39 public sites are summarized in Table 1.Levels of HCHO and PM 2.5 in Category 1 spaces and levels of CO, O 3 , HCHO, PM 10 , and PM 2.5 in Category 2 spaces met the Taiwan EPA's suggested threshold.As illustrated in Fig. 1, the common IAQ measurements that exceeded the Taiwan EPA's suggested threshold included CO 2 , TVOC, TFC, and TBC in both Category 1 and Category 2. CO concentrations at hospitals were found to exceed the Category 1 threshold of 2 ppm.The cause of the high CO concentration may be ascribed to emissions from vehicle exhaust resulting on hospitals' central high-traffic locations.Tutoring centers and kindergartens were found to exceed the Category 1 O 3 threshold of 0.03 ppm and PM 10 threshold of 60 ppm.The highest O 3 concentration occurred around noon (AM 11:00 to PM 2:00), and the O 3 source was most likely from outdoor air due to the absence of potentially O 3 -generating electric facility indoors.The high PM 10 levels at tutoring centers and kindergartens were also caused by outdoor air, while the high PM 10 levels at hospitals may be associated with air conditioning, therapeutic action, and patient motion.
In general, CO 2 concentrations above 1000 ppm indicate poor ventilation of the indoor environment and can be remedied by improving ventilation (Lee and Chang, 2000).In Fig. 1, CO 2 exceeded the Taiwan EPA's suggested threshold by 61.1% in Category 1 spaces and 9.5% in Category 2 spaces.A Category 1 CO 2 threshold of 600 ppm may be too strict for most public sites.Air conditioning usually employs constant recycling based to save electricity, thus ventilation may be insufficient to lower indoor CO 2 levels when crowds gather indoors.
TVOC concentrations exceeding the suggested threshold of 3 ppm occurred in tutoring centers, office buildings, and fitness centers.The fitness centre is located in the basements of buildings, and the heating boiler is regarded as potential sources of indoor TVOCs.In tutoring centers and office buildings, high levels of TVOCs may be caused by new furniture and vehicle exhaust from outdoor air (Panagopoulos et al., 2011).
TBC and TFC levels that exceed the suggested threshold are a common IAQ problem among the investigated public sites.The differences of TBC as well as TFC between Category 1 and Category 2 were examined with the aid of statistical tools, F-test and T-test.The theory test of TBC showed that the p-values are 0.193 (> 0.05) for F-test and 0.063 (> 0.05) for T-test.This result implicated that there is no significant differences on variance and mean of TBC between Category 1 and Category 2. For theory test of TFC, both p-values of F-test (9.7 × 10 -6 ) and T-test (0.0016) are smaller than the significant level 0.05.Thus, there exists the significant difference of TFC between Category 1 and Category 2. High TBC levels may be caused by outdoor pollutants and indoor crowds.Other factors that can also affect indoor TBC concentrations include the cleaning of air conditioning systems, disinfectant housekeeping, and control of the ventilation rate.In addition to indoor green plants, the majority of fungal sources that contribute to high indoor TFC levels may be in the surrounding outdoor area.

Correlationship between Indoor Bioaerosols and Particulate Matters
The TBC, TFC, PM 2.5 and PM 10 concentrations were subjected to statistical linear regression analysis to examine the correlationships among them.Fig. 2 illustrated that regression coefficients were 0.12 for TBC and PM 2.5 , 0.08 for TBC and PM 10 , 0.19 for TFC and PM 2.5 , and 0.27 for TFC and PM 10 .All regression coefficients were below 0.4, indicating that indoor bioaerosols and particulate matters may have only low correlationships.Indoor air may contain bioaerosols and non-biological particles, and thus the low correlationship between bioaerosols and particulate matters implies that the percentage of bioaerosols could be low in indoor air.A comparison of the regression coefficients of TBC and particulate matters shows that TFC had higher regression coefficients and slopes of regression line for both PM 2.5 and PM 10 .This suggests that particulate matter might contribute more to TFC than TBC.In light of the reported indoor/outdoor (I/O) PM 2.5 ratio of 0.93, indoor PM 2.5 could reasonably ascribed to outdoor sources in roadside areas (Huang et al., 2007).This finding is also consistent with the fact that most TFC originates outdoors (Wu et al., 2000;Karbowska-Berent et al., 2011).In certain residences close to high traffic roads, outdoor pollutant penetration should be noted (Lai et al., 2010).

Correlationship between Indoor Bioaerosols and CO 2
In order to depict the correlationships between TBC and CO 2 as well as TFC and CO 2 , the results of the statistical linear regression analysis are illustrated in Fig. 3.A moderate correlationship between TBC and CO 2 was observed based on its regression coefficient of 0.44, while TFC and CO 2 had a low correlationship with a regression coefficient of 0.16.Kim et al. (2009) pointed out that microbial activity can be affected by CO 2 and RH, and Kalogerakis et al. (2005) claimed humans are the major source of bacteria in closed spaces.The literature also agrees with our result that TBC concentrations are moderately linked to CO 2 levels (Kim et al., 2009).Accordingly, TBC concentrations can be potentially affected by the intensity of indoor human activity and control of the ventilation rate.

Control of Indoor TBC and TFC
One hospital with an excessively high level of bioaerosols was selected for observation of the temporal transience of bioaerosols.Baseline IAQ concentrations were measured on September 2, 2009, and the concentrations of CO 2 and TBC exceeded the Taiwan EPA's suggested threshold.The ventilation period was increased and alcohol and NaOCl were used for disinfection to improve the IAQ.As shown in Table 2, IAQ measurements on January 7, 2010 showed that the CO 2 concentration dropped from 944 to 620 ppm and the TBC level dropped from 2935 to 823 CFU/m 3 , which were both still higher than their respective Category 1 thresholds.Also, the measured TFC level was 1674 CFU/m 3 higher than the Category 1 threshold.After thorough disinfectant housekeeping, the TBC level decreased to 277 CFU/m 3 , while the TFC level increased to 6701 CFU/m 3 based on IAQ measurements on January 25, 2010.In order to evaluate the cause of the high TFC level, indoor TFC and outdoor TFC were measured on March 5, 2010.The measured TFC concentrations were 569 CFU/m 3 in the lobby, 1422 CFU/m 3 outdoors, 1713 CFU/m 3 in the ventilation intake, and 1901 CFU/m 3 in the  TNTC (lobby) TNTC (outdoor air) TNTC (intake of ventilation) TNTC (ventilation exhaust) ventilation exhaust.Thus, it was confirmed that the TFC came from outdoors.The TFC measurement was checked again on March 25, 2010, and all TFC measurements were too numerous to count (TNTC).
Increasing the ventilation period and disinfectant housekeeping can effectively control the TBC concentration, but not TFC concentration.In Table 2, the measured TFC levels in January and March were much higher than the measurement in September.The extremely high TFC levels may be ascribed to dust storms and furrowing activity on nearby farmland that increased the amount of fungal spores in the outdoor air, which then moved indoors.Consequently, the temporal transience of TFC concentration is mainly affected by the outdoor environment and seasonal climate change.
Several aerobiological technologies have been proposed to control indoor bioaerosols such as ultraviolet germicidal irradiation, photocatalytic oxidation, ionization, ozone, negative ion, etc. (Huang et al., 2012).In this study, ClO 2 was used as a disinfectant to control the indoor TFC level.With ClO 2 spraying, the concentration of TFC effectively decreased from TNTC to 3071 CFU/m 3 after 0.5 hours, but jumped back to TNTC after 1 hour.The result of this disinfection illustrates that ClO 2 spraying can effectively control indoor TFC levels when done properly.The dose of disinfectant, the span of spraying, and the frequency of disinfection are all important to maintaining good IAQ conditions

Identification of Fungal Genera
MEA agar plates sampled on March 25, 2010 were selected according to their appearance, color, and the distribution of fungal growth and were sent to Bioresource Collection and Research Centre for identification of fungal genera.Three identified filamentous fungus genera included Cladosporium perangustum (Fig. 4), Cladosporium tenuissimum (Fig. 5), and Fusarium incarnatum (Fig. 6).Kendrick (1990) listed the big eight genera of molds as Alternaria spp., Cladosporium spp., Curvularia spp., Drechslera spp., Epicoccum spp., Fusarium spp., Nigrospora spp.and Stemphylium spp., because of their allergenicity and frequency of appearance in air.In Beijing, the dominant fungi were Cladosporium spp., Alternaria spp., Penicillium spp.and Aspergillus spp. in the outdoors (Fang et al., 2005); Fusarium spp., Aspergillus spp., Penicillium spp.and Basipetospora spp.were found in west Korea during the Asian dust periods (Yeo and Kim, 2002).In Turkey, the prevailing mold genera were Cladosporium spp., Aspergillus spp., Penicilliu, spp.and Rhizopus spp.(Mentese et al., 2009).In this study, Cladosporium and Fusarium were the most common allergens in Taiwan, which were usually present in soils, withered fallings, leaves' surfaces, and moldy plants.The Taiwan EPA (2012) reported the dominant fungi to be Cladosporium spp., Aspergillus spp., Penicillium spp., and Curvularia spp. in an IAQ investigation of schools and kindergartens.An I/O TFC ratio of 1.09 and I/O dominant fungal ratio of 1.0 imply that indoor fungi come from the outdoor sources (Taiwan EPA, 2012).Based on our fungal observation, the numeration of Cladosporium is estimated to comprise 70-80% of all fungal genera, which is higher than the reported Cladosporium indoor dominance of 48.8% by the Taiwan EPA (2012).The generic classifications of Cladosporium are Cladosporium perangustum and Cladosporium tenuissimum, which represent the dominant fungal genera in southern Taiwan.Fusarium is estimated to comprise 5-6% of all fungal genera that can cause plant disease and harm to 70 to 80 economic plants.Fusarium mostly originates outdoors, although there are some potential sources indoors as well (Taiwan EPA, 2012).In this case study, furrowing activity may have caused adsorbed microorganisms in soil and plants to be released into the outdoor air, which is our explanation of high TFC levels and identified dominant fungal genera.

CONCLUSIONS
This study investigated 39 public sites in southern Taiwan.IAQ measurements commonly exceeding the Taiwan EPA's suggested threshold included CO 2 , TBC, and TFC.TBC and TFC concentrations were weakly linked to particulate matters due to the low percentages of bioaerosols in indoor air.Particulate matters might contribute more to TFC than TBC.This is ascribed to the fact that most TFC originates outdoors.TBC concentration was moderately correlated with CO 2 level, and thus TBC concentrations can be potentially affected by the intensity of indoor human activity and ventilation rate.A case study of bioaerosols in a hospital found extremely high TFC levels that may be ascribed to the occurrence of dust storms and furrowing activities on the nearby farmland.The temporal transience of TFC concentrations is mainly affected by the outdoor environment and seasonal climate change, and ClO 2 spraying can effectively control the indoor TFC level when executed properly.The high-TFC outdoor air samples contained three filamentous fungus genera: Cladosporium perangustum, Cladosporium tenuissimum and Fusarium incarnatum, which are the most common allergens in Taiwan.The dominant fungal genus was Cladosporium (estimated 70-80%) in indoor air, which is a similar finding in other countries.Comparatively, indoor Fusarium (estimated 5-6%) is not so frequently found in other

Fig. 1 .
Fig. 1.Ratio excess rates of IAQ measurements at public sites of Category 1 and Category 2.

Table 2 .
Exceeding IAQ measurements and improving actions in a hospital.