j ourna l homepage: www.e lsev ie r .com/ locate /env int
The impact of flood and post-flood cleaning on airborne microbiologicaland particle contamination in residential houses
Congrong He a, Heidi Salonen a,b, Xuan Ling a, Leigh Crilley a, Nadeesha Jayasundara a, Hing Cho Cheung a,Megan Hargreaves a, Flavia Huygens a, Luke D. Knibbs a,c, Godwin A. Ayoko a, Lidia Morawska a,⁎a International Laboratory for Air Quality and Health, Queensland University of Technology (QUT), GPO Box 2434, Brisbane, Queensland 4001, Australiab Developing Indoor Environments, Finnish Institute of Occupational Health (FIOH), Topeliuksenkatu 41 aA, FI-00250 Helsinki, Finlandc School of Population Health, The University of Queensland, Herston, Queensland 4006, Australia
Article history:Received 22 January 2014Accepted 2 April 2014Available online xxxx
Keywords:Indoor airParticle numberPM10
FungiBacteriaIndoor dust
In January 2011, Brisbane, Australia, experienced amajor river flooding event. We aimed to investigate its effectson air quality and assess the role of prompt cleaning activities in reducing the airborne exposure risk. A compre-hensive,multi-parameter indoor and outdoormeasurement campaignwas conducted in 41 residential houses, 2and 6 months after the flood. The median indoor air concentrations of supermicrometer particle number (PN),PM10, fungi and bacteria 2 months after the flood were comparable to those previously measured in Brisbane.These were 2.88 p cm−3, 15 μg m−3, 804 cfu m−3 and 177 cfu m−3 for flood-affected houses (AFH), and2.74 p cm−3, 15 μg m−3, 547 cfu m−3 and 167 cfum−3 for non-affected houses (NFH), respectively. The I/O (in-door/outdoor) ratios of these pollutants were 1.08, 1.38, 0.74 and 1.76 for AFH and 1.03, 1.32, 0.83 and 2.17 forNFH, respectively. The average of total elements (together with transition metals) in indoor dust was 2296 ±1328 μg m−2 for AFH and 1454 ± 678 μg m−2 for NFH, respectively. In general, the differences between AFHand NFH were not statistically significant, implying the absence of a measureable effect on air quality from theflood.We postulate that this was due to the very swift and effective cleaning of the floodedhouses by 60,000 vol-unteers. Among the various cleaning methods, the use of both detergent and bleach was the most efficient atcontrolling indoor bacteria. All cleaning methods were equally effective for indoor fungi. This study providesquantitative evidence of the significant impact of immediate post-flood cleaning on mitigating the effects offlooding on indoor bioaerosol contamination and other pollutants.
Floods are one of themost common natural disasters and significantflooding events have often resulted in increased morbidity and mortal-ity throughout the world (Ahern et al., 2005; Alderman et al., 2012; Duet al., 2010). Based on the Fifth Assessment Report (AR5) of the UnitedNations Intergovernmental Panel on Climate Change (2013), extremeprecipitation events over most of the mid-latitude land masses andover wet tropical regions will very likely become more intense andmore frequent by the end of this century. It implies that major floodingis becoming more frequent and greater in magnitude as the global cli-mate continues to change (Taylor et al., 2011). Therefore, the environ-mental and public health risks associated with major flooding eventsare projected to increase in the future.
Flooded areas can become a source and reservoir for pathogenswhichcan impact the health of the residents through various transmission
f Technology, GPO Box 2434,616; fax: +61 7 3138 9079.).
pathways (Taylor et al., 2011). Damp andfloodeddwellings can supportmicrobial growth, including mold, bacteria, and protozoa, as well aspersistence of flood-borne microorganisms (Taylor et al., 2013), one ofwhich is aerosolization of part or all of themicro-organisms into the in-door air. Exposure to fungal contamination can lead to infectious dis-ease and other health effects which can impact on the respiratorysystem, skin and eyes. Adverse health effects can be categorized as in-fections, allergic or hypersensitivity reactions, or toxic irritant reactions(Metts, 2008). However, the role of floods in this process is still not wellquantified (Hsu et al., 2011).
Laboratory-based examination of the aerosolization of culturableand total fungi, (1-3)-b-D glucan and endotoxins from eight flood-affected floors and bedding material samples collected from NewOrleans homes following Hurricane Katrina was conducted byAdhikari et al. (2009). Their results indicated that significantly higherlevels of these contaminants were observed in the flood-affected mate-rials compared to other studies conducted in urban homes. At the sametime, the levels of culturable and total fungi found in these materialswere slightly lower than those previously reported for moldy buildings.Molds and mycotoxins in indoor dust samples after the same event
10 C. He et al. / Environment International 69 (2014) 9–17
were also analyzed by Bloom et al. (2009). They reported that themost commonly found mold taxa were Aspergillus, Cladosporium andPenicillium.
Increased levels of airborne indoor microbes (bioaerosols) aftermajor flooding events have been reported in a number of studies(Adhikari et al., 2009; Chew et al., 2006; Fabian et al., 2005; Hoppeet al., 2012; Hsu et al., 2011; Khan and Wilson, 2003; Rao et al., 2007;Ross et al., 2000; Schwab et al., 2007; Solomon et al., 2006). The averageindoor and outdoor spore concentration levels varied significantly fromstudy to study and from region to region. In addition to different climat-ic regions, and sampling methodology, one explanation for this largevariation in concentration may also be the different response timesand methods used to clean-up after flood. However, there is very limit-ed information available on these parameters in previous studies. Chewet al. (2006) conducted a study to characterize airbornemold and endo-toxins throughout all phases (before, during and after) the cleanup pro-cess in three houses in New Orleans, which sustained between 0.3 and1.8 m of flood damage from Hurricanes Katrina and Rita. They reportedthat after the intervention, which included disposing of damaged fur-nishings and drywall, cleaning surfaces, drying the remaining structureand treatment with a biostatic agent, the levels of mold and endotoxinswere generally lower than pre-intervention levels. Recently, Hoppeet al. (2012) showed that proper post-flood remediation led to im-proved air quality and lower exposures among residents living inflooded homes.
In January 2011, about 22,000 Brisbane homes and 7600 businessesacross 94 suburbs experiencedmajor or partial inundation by floodwa-ters from the Brisbane River. After the flood waters had receded, thelocal authorities organized an immediate and extensive clean-up oper-ation to remove wet materials and dry out the building structures. Toprovide a better understanding of the impact of the flood and to testthe hypothesis that the cleaning prevented high post flood contamina-tions, the main objectives of this work were: 1) to assess the effect offlooding on indoor and outdoor PN and PM10, airborne culturablefungi and bacteria concentrations, as well as fungal flora; 2) to investi-gate the effect of flooding on the concentration of inorganic elementsin indoor dust; 3) to analyze the correlations between indoor and out-door concentrations of the pollutants, as well as indoor inorganic ele-ments in indoor dust; 4) to analyze the role of different cleaningapproaches on improving indoor air quality; and 5) to compare the re-sults with the limited data currently available in the literature.
2. Experimental methods
On 13 January 2011, flood waters in the Brisbane River peaked at4.46 m in Brisbane City and remained elevated until 14 January. Theheight of water in the flooded houses ranged from 5 to 270 cm. TheBrisbane City Council Local Disaster Management Group organized a“MudArmy” of volunteers to assistwith clean-up activities immediatelyafter the flood waters receded.
Approximately 23,000 volunteers registered for the first weekend(15–16 January) of the clean-up. They were allocated to sectors andthen transported to themby Council bus to assist residents and businessownerswith debris removal and other cleaning activities. On the secondweekend (22–23 January 2012), a large number of parks were cleaned.Council's call for assistancewas also answered bymany volunteers whodid not register. It is estimated that there were between 50,000 and60,000 volunteers who assisted over the second weekend clean-upactivities.
2.1. The sampling sites and houses
Nine residential suburbs of Brisbane located along the banks of theBrisbane River that were affected by the flood were chosen as the mea-surement sites. Median family income, as reported by the 2011 Census,ranged from AU$52,208 to AU$123,968 for the nine residential suburbs.
Wedelivered almost 600 invitation letters in this area and sent an emailinvitation via the QUT's media office (with a mailing list size of 2000 re-cipients). Following this, a total of 41 houses were enrolled, of which 24were flooded and 17 were not flooded. The latter were used as controls.The houses represented a variety of age, building material and designstyle. The general house characteristics are described in Table S1 inSupporting information.
2.2. Instrumentation and methodology
2.2.1. Airborne particulate matterIndoor and outdoor total supermicrometer PN concentrations (from
0.54 to 19.81 μm) were measured by a TSI Model 3312A UltravioletAerodynamic Particle Sizer (UVAPS) (TSI Incorporated, St. Paul, MN,USA), with the time resolution of 20 s.
Two TSI Model 8520 DustTrak aerosol monitors (TSI Incorporated,St. Paul, MN, USA) were used to simultaneously measure indoor andoutdoor PM10 concentrations. Since the instrument does not measureactual gravimetric values, and in order to obtain values closer to truePM10 concentrations, all of the DustTrak data were corrected based oncomparison of the DustTrak readings with readings from a TEOMmon-itor (50 °C R&P 1400a) at QUT.
2.2.2. BioaerosolsCulturable, viable fungi were collected using a Biotest RCS HIGH
FLOW (Biotest Hycon, Art. No. 940210, Ser. No. 30709) centrifugal im-pact air sampler, for 20 L, 50 L and 100 L air samples, at flow rate of100 Lmin−1. Since the concentration of culturable molds in the floodedhouses was not known, three sampling volumes were used for the firstfive houses, in order to ensure that the sufficient amount of material wascollected. Rose Bengal agar strips were used for collecting the samples,which were incubated at 28 °C for four days, prior to counting by directvisual inspection. Partial identification as Penicillium, Cladosporium andAspergillus to a genus level was conducted after seven days incubation.The chosen fungal genera were those frequently occurring indoors, inboth Australia and other places in the world (Hargreaves et al., 2003;Jo and Seo, 2005; Mandal and Brandl, 2011; Salonen et al., 2007; Wuet al., 2000). Aspergillus and Penicillium species were targeted, as theycan be toxic at elevated levels, due to their ability to producemycotoxins.Cladosporium has been known to cause several different types of infec-tions, including skin, eye, sinus, and brain infections. Cladosporium hasalso been associated with allergies and asthma (CEN, 2014).
Culturable bacteria were assessed using the same Biotest RCS HIGHFLOW centrifugal impact air samplers, for 100 L samples. Tryptic soyagar strips were used and incubated at 32 °C for three to four days topermit quantification. The results of culturable fungi and bacteria wereexpressed as colony-forming units per cubic meter of air (cfu m−3).
2.2.3. Dust elemental compositionDust samples were collected by passive sampling on a 1 m2 glass
panel, which was placed in the living room of each house for oneweek. The KimWipe tissues used for the dust collection were first pre-cleaned by sonication for 5 min in 3:1 volume to volume mixture ofacetone and Milli-Q water (18.2 MΩ cm). After drying the tissueswere placed in pre-cleaned sampling tube and weighed. The dust wassampled at each house using the tissue to wipe the dust from the glassand weighed to determine the mass of dust collected. The tissue withthe dust was microwave digested in 15 mL of concentrated HNO3 for15 min at 180 °C. The diluted digest was analyzed by inductivelycoupled plasma mass spectroscopy (ICP-MS) (Model Agilent 7500ce)according to the method described in Lim et al (2006). The concentra-tions of 24 elements, including: Li, Be, Na, Mg, Al, K, Ca, As, Sr, Ba andPb, together with the transition metals Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn,Mo, Cd, Ir and Hg were determined. Prior to each analysis the ICP-MSwas auto-tuned and calibrated using standards prepared from TraceCert(SigmaAldrich) standard solutions. All reported elemental concentrations
11C. He et al. / Environment International 69 (2014) 9–17
were field blank corrected, using blank solutions prepared by digestionof clean KimWipe tissues.
2.2.4. Other parametersA TSI Model 7545 Q-trak was used for the measurement of indoor
and outdoor CO, CO2, temperature and relative humidity. Weatherdata from the Bureau of Meteorology, and ambient air quality datafrom the Department of Science, Information Technology, Innovationand the Arts (DoSITIA) Queensland, were also obtained for the durationof the sampling periods.
2.2.5. Housing questionnaireA questionnaire was developed and used in this study to gather in-
formation in relation to the impact of different clean-upmethods onmi-crobiological and particle concentrations, as well as the relationshipbetween indoor air quality and house characteristics. The questionnaire(website: www.qut.edu.au/research/ilaqh/floodprojectquestionnaire/)was approved by the QUT Human Research Ethics Committee.
2.3. Study design
Since fungi and bacteria concentrations are known to be highly sea-sonal (e.g. Adhikari et al., 2009; Cho et al., 2008; Mentese et al., 2012),and given that Adhikari et al. (2009) observed decreases in fungi con-centration for up to two-years following the flooding event in NewOrleans, two sets of measurements were conducted, one from 21March–03 May, 2011, (two months after the flood) during theAustralian autumn, and the second from 18 July–12 August, 2011 (sixmonths after the flood), during thewinter. The second set was conduct-ed in 26 houses (15 owners/residents were not willing to continue theirparticipation in this study), of which 16 were affected by the flood.
Sampling at each house was conducted for approximately 3 h. Thetwo DustTrak measured simultaneously indoor and outdoor PM10 con-centrations respectively. The UVAPS measured outdoor air for 15 min,followed by 15 min of indoor measurements in the living room. Eachof these UVAPSmeasurementswas repeated twice (in total 3 times out-door and 3 times indoor). The Q-Trakmeasured indoor and outdoor CO,CO2, temperature and relative humidity with the UVAPS. For culturablefungi, eighteen indoor samples (nine from the living room and ninefrom the main bedroom) and nine outdoor samples (from the outdoorcontrol site) were collected for each of the first five houses. For the re-maining houses, six indoor samples (three from the living room andthree from the main bedroom) and three outdoor samples (from theoutdoor control site) were collected for each house. For culturable bac-teria, a total of nine samples (three outdoor, three in the living roomandthree in the main bedroom) were collected for each house. Additionalculturable fungi and bacteria measurements were conducted in the ga-rage of 15 houses during the first round measurements. After the mea-surements and sampling were completed in each house, a 1 m2 glasspanel was placed in the living room for one week, for the purpose ofcollecting dust samples. A schematic diagram of the instrumental setup for indoor and outdoor measurements and fungi and bacteria sam-pling is provided in Supporting information Fig. S1. During outdoormeasurement, instruments were placed 1 to 5 m away from thehouse. The sampling heights for both indoor and outdoor instrumentsware 1.4 to 1.7 m.
2.4. Data processing and analysis
Houses were classified as those affected (AFH) and not affected bythe flood (NFH). The flood affected houses were further classified ac-cording to the cleaning methods used and/or the progress of theclean-up, including time taken to clean up, the use of detergent andwhether remediation works had been completed or were still in prog-ress. Comparisons between these groups were made using the Mann–Whitney U test (nonparametric equivalent of Student's t-test) for
median values and two sample independent t-tests for mean values.Correlations were investigated using Pearson's correlation analysis.The p=0.05 level of significancewas used in this study, and all analysesand plots were performed using IBM SPSS statistical software (v19).
3. Results and discussion
The range of average meteorological parameters during the firstround of measurements was 2–19 km/h for wind speed, 14–23 °Cfor minimum temperature, 22–34 °C for maximum temperatureand 34–91% for relative humidity. During the second round, con-ducted during the winter, the ranges were: 2–20 km/h, 6–12 °C,18–24 °C and 25–80%, respectively. Overall, it can be seen that, due toBrisbane's subtropical climate, the main difference between the twomeasurement rounds was the temperature, with a relatively small dif-ference of 10 °C.
The indoor and outdoor PN, PM10, fungi and bacteria concentrationsfor AFH and NFH in the first and second rounds are summarized as boxplots in Fig. 1. A detailed description of the results is given in the follow-ing sections.
3.1. Particle number and mass concentrations
A summary of the indoor and outdoor average and median PN andPM10 concentrations, as well as indoor/outdoor (I/O) ratios and flood-affected/non-affected ratios is given in Table S2 in Supporting informa-tion for both rounds. It can be seen from Fig. 1 that, in general, for bothrounds, the total indoor and outdoor average andmedian PN concentra-tions in the AFH were comparable to those in the NFH, and the analysisshowed that there were no statistically significant differences betweenthem. In the first round, average indoor PN concentration for AFH wassignificantly higher than outdoor concentration, however the medianvalue was not. In the second round average indoor PN concentrationsfor both AFH and NFH were significantly higher than outdoor levels,which implies that there were indoor particle sources in the housestested.
Although total indoor and outdoor average and median PM10 con-centration levels for AFH were somewhat higher than those for NFH,these differences were not statistically significant. Indoor medianPM10 concentration levels for AFH were the same as for NFH. For bothAFH and NFH, average and median indoor PM10 concentrations weresignificantly higher than outdoor concentrations.
A comparison of the average indoor and outdoor PN and PM10 con-centrations for the 26 houses which participated in both rounds is pre-sented in Fig. 2. A summary of the concentration ratios is given inTable S3 in Supporting Information. Both indoor and outdoor averagePN concentrations in the first round of measurements were higherthan those in the second round, for both groups of houses (seeTable S3 and Fig. 2). Except for indoor PN concentrations in NFH, thesedifferences were statistically significant. With the exception of outdoorPM10 concentrations for AFH, whichwere significantly lower in the firstround compared to the second round, there were no statistically signif-icant differences between PM10 concentrations for both rounds. Therewere also no statistically significant differences between the averageand median values for these 26 houses.
These results suggest that, in terms of PN and PM10 concentrations,there was no quantifiable effect of the flood on indoor air quality inAFH. However, the results from the first and second rounds of measure-ments showed that there were some indoor PM10 and PN sources inboth groups of houses, which makes differentiating their effects fromthose due to flood difficult.
There was no PM10 and only one PN concentration data available forAFH in literature. Chew et al. (2006) reported that the average total PNconcentration (size range from0.3 to 20 μm) for anAFHwas higher dur-ing the renovation (about 200 p cm−3) than before renovation (about70 p cm−3). Morawska et al. (2001) reported that the average indoor
Fig. 1. Indoor and outdoor PN, PM10, fungi and bacteria concentrations for affected houses (AFH) and non-affected houses (NFH) in the first and second rounds. Note: 1IN, first round in-door; 1OutA, first round outdoor; 2IN, second round indoor; 2Out, second round outdoor. The lower and upper boundaries of the box specify the 25th and 75th percentiles, respectively.The line within the box indicates the median and the whiskers above and below the box indicate the maximum and minimum, respectively.
12 C. He et al. / Environment International 69 (2014) 9–17
supermicrometer PN concentration in 16 residential houses inBrisbane was 2.5 ± 1.76 p cm−3. The PN levels observed in thisstudy (from 1.81 ± 0.99 to 3.22 ± 1.52 p cm−3) are similar to those
0
1
2
3
4
5
6
Indoor Outdoor
PN
(p
cm
-3)
First round AFH Second Round AFH
First round NFH Second Round NFH
0.E+00
5.E+02
1.E+03
2.E+03
2.E+03
3.E+03
3.E+03
4.E+03
4.E+03
5.E+03
5.E+03
Indoor Outdoor
Fu
ng
i (cf
u m
-3)
First round AFH Second Round AFH
First round NFH Second Round NFH
Fig. 2. Comparison of the average indoor and outdoor PN, PM10, fungi and bacteria concentratideviation.
previously reported byMorawska et al. (2001), but lower than those re-ported by Chew et al. (2006). These results imply that the flood did notaffect indoor concentration levels in relation to PN concentration.
0
5
10
15
20
25
30
35
40
Indoor Outdoor
PM
10 (µ
g m
-3)
First round AFH Second Round AFH
First round NFH Second Round NFH
0
50
100
150
200
250
300
350
400
450
Indoor Outdoor
Bac
teri
a (c
fu m
-3)
First round AFH Second Round AFH
First round NFH Second Round NFH
ons for the 26 houses that participated in both rounds. The vertical lines present standard
13C. He et al. / Environment International 69 (2014) 9–17
3.2. Fungi and bacteria concentrations
A summary of the average andmedian culturable fungi and bacteriaconcentrations, as well as I/O ratios, for both rounds is presented inTable S4 in Supporting information. Since both the indoor and outdoorbacterial sampling strips for House 37 were overgrown with spreadingcolonies, there is no bacterial data available for this house. Althoughboth indoor and outdoor average fungi concentrations in AFH are higherthan those in NFH, the differences were not statistically significant. Forboth groups, average indoor fungi concentrations were lower thanthose outdoors. However, the difference was only statistically signifi-cant for NFH.
The most frequently isolated fungi genus from the indoor air in bothAFH and NFH was Penicillium, followed by Cladosporium. In outdoor air,the prevalent fungal genus for AFHwas Cladosporium, while for NFH theoccurrence of Cladosporium and Penicilliumwas similar. The occurrenceof Aspergillus was much lower for both types of houses, in both indoorand outdoor air. The occurrence of other fungal genera in indoor andoutdoor was higher in NFH than in AFH.
There were no statistically significant differences between AFH andNFH for both indoor and outdoor average bacterial concentrations;however the outdoor average bacterial concentrations in AFH werehigher than those in NFH, while their median values were the same. Itcan also be seen from Table S4 in Supporting information that the aver-age indoor bacterial concentrations were significantly higher (76% and117%, p b 0.05) than average outdoor bacterial concentrations.
A comparison of living room and bedroom fungi and bacteria con-centrations showed that, with the exception of fungi concentrationsfor NFH, there were no significant differences between fungi and bacte-ria concentrations in these two rooms. There were significant correla-tions between the living room and bedroom fungi and bacteriaconcentrations for both AFH and NFH (data not shown).
An additional 15 culturable fungi and 4 bacteriameasurements werejust conducted in the garages of 15 houses in thefirst round. The averagefungi concentrations in indoor and garage air were 928 ± 421 cfu m−3
and 1579±1419 cfum−3 in 13 AFH, and 651±12 cfum−3 and 1427±32 cfum−3 in 2NFH, respectively. The average bacteria concentrations inthe indoor and garage air were 153 ± 41 cfu m−3 and 73 ± 48 cfu m−3
in 4 AFH, respectively. Although the median garage fungi concentrationswere higher than those in the indoor air, these differences were notstatistically significant. The differences in bacteria concentration werealso not statistically significant.
For the second round, there was no statistical difference in fungi andbacteria concentrations between the two group houses, for both indoorand outdoor air (see Table S4). However, average and median indoorfungi concentrations in AFH were lower (p b 0.05) than those in out-doors, while average and median indoor bacteria concentrations inAFH were higher (p b 0.05) than those in outdoors.
A comparison of the average indoor and outdoor fungi concentra-tions, and indoor and outdoor bacteria concentrations for the 26 housesis also presented in Fig. 2. A summary of the concentration ratios is alsogiven in Table S3 in Supporting information. It can be seen from Fig. 2and Table S3 that both indoor and outdoor fungi and bacteria in thefirst round exhibited lower concentrations than those in the secondround for both groups of houses. These differenceswere only statistical-ly significant for indoor and outdoor bacteria, aswell as outdoor fungi inNFH. A comparison of the percentage of the isolated fungi genera in thefirst and second rounds is given in Supporting information Fig. S2. It isapparent that the percentage of Penicillium increased in the secondround (from 22–36% to 40–63%) for both indoor and outdoor air, andfor both groups of houses, while, in contrast, the indoor percentage ofCladosporium decreased. Since fungi concentrations are known to behighly seasonal (Bartlett et al., 2004; Frankel et al., 2012), seasonalitymay be one possible explanation for these changes.
These results further imply that there was no measureable effect ofthe flood in relation to either fungi or bacteria concentrations. However,
it should be noted that in subtropical areas outdoor sources of fungiwere the main contributors to indoor fungi concentration levels in allseasons, and the “normal” background fungal concentrations frequentlyexceed 1000 cfu m−3 (Hargreaves et al., 2003). Thus, the difference be-tween outdoor and indoor concentrations or association between in-door fungal concentrations and moisture damages were not alwaysdetected, although there may be mold/moisture damages in buildingstructures. For culturable bacteria, the main contributors were indoorsources.
At present, culturable fungi concentrations in AFH have been report-ed in 8 studies, under different sampling conditions (including numberof houses tested, period of sampling after flooding, stage of the remedi-ation process, etc) and a summary of these studies is presented inTable 1 and compared with the results of the present study.
The earliest work to assess the effects of flooding was that of Pearceet al. (1995), who conducted long-term monitoring of indoor and out-door culturable fungi concentrations in 8 AFH. Their results suggestedthat simply cleaning, disinfecting and drying flooded surfaces wouldnot preventmold contamination frompersisting long after thefloodwa-ters had subsided, unless some strategy for decontaminating hiddensurfaces could be devised. Curtis et al. (2000) conducted long-term con-tinuousmeasurements (monthly) in a total of 45 houses and found thatfungi concentrations were not significantly greater in the AFH versusNFH. Ross et al. (2000) measured indoor levels of mold spores over a7 month period in 44 asthmatics' homes (17 of which were AFH),starting 1 year after the flood. Indoor and outdoor culturable fungiconcentrations in 8 cleaned and reoccupied houses (plus 1 NFH) weretested several months after a major flood by Fabian et al. (2005), whofound significantly higher airborne microorganism levels in the AFHversus the NFH, and in many cases the difference was between 2 and3 orders of magnitude. The authors suggested that the flooded buildingmaterials were sustaining high aerosol bioburdens and contributing topoor indoor air quality more than 3 months after the structures hadbeen remediated following flood damage. Chew et al. (2006) measuredindoor and outdoor culturable fungi concentrations before, during andafter remediation in 3 AFH, 4–6 months after the flooding and reportedthat indoor culturable fungi concentrations decreased after remediationhad been completed. Rao et al. (2007)measured airbornemold concen-trations in 20 houses (15 moderately/heavily water-damaged housesand 5 mildly water-damaged houses) and found that culturable fungiconcentrations were significantly higher in the moderately/heavilywater-damaged houses than in the mildly water-damaged houses.Rabito et al. (2008) measured indoor and outdoor mold levels in 54houses after remediation twice (6–7 months and 8–9 months after theflooding) and found that both indoor and outdoor culturable fungiconcentrations decreased. Recently, Hsu et al. (2011) compared thedifference between fungal concentrations before and after a flood in14AFH, and found that the average total culturable fungal concentrationdecreased after the flood event, although the I/O ratio showed a visibleincrease.
Several conclusions can be derived from inspection of the data pre-sented in Table 1. Firstly, indoor andoutdoor average levels of culturablefungi concentrations vary widely, within one to three orders of magni-tude, in all of these studies, as well as between the studies. Secondly,flooding can increase indoor and outdoor culturable fungi concentrationlevels in moderately/heavily water-damaged houses and surroundingareas (Chew et al., 2006; Fabian et al., 2005; Pearce et al., 1995; Raoet al., 2007). However, after remediation, indoor culturable fungi con-centration levels were likely to decrease (Chew et al., 2006; Rabitoet al., 2008 and this study) and about one year after the flood, the effectis likely to disappear all together (Curtis et al., 2000). Thirdly, in general,I/O ratios of culturable fungi concentrations in AFHwere N1 immediate-ly after flooding, however following remediation, these ratios were b1in both AFH and NFH. Finally, the most frequently isolated fungi genusfrom indoor and outdoor air differed between these studies, especiallyin Fabian et al. (2005), where Trichodermawas the predominant indoor
Table 1A summary of the indoor and outdoor average culturable fungi concentrations (×103 cfu m−3) in flood affected house studies.
Study Number of house(flooded/total)a
Sampling taking afterflooding (month)
Indoor Outdoor Indoor dominatedspecies b
Outdoor dominatedspecies
Remediationprocessc
RFF
Chew et al. (2006) 3/3 4–6 353 (22–515) 8.00 Pe, As, Pa Pe, As, Pa Before clean Clear3/3 4–6 3 ~ 20 10.0 After
Curtis et al. (2000) 18/45 12–19 2.24 (10–48.45) 4.22 (0.050–41) As, Pe, Cl As, Pe, Cl Both Clearf
Fabian et al. (2005) 8/8 Several 2.20 (±2.24) 0.95 (±0.88) Tr Pe After Clear0/1 0.035 0.071
Hsu et al. (2011) 14/14 0.5–1 13.4 (±11.0) 6.92 (±7.85) As, Pa, Cl, Pe As, Pa, Fu, Cl Both NoPearce et al. (1995) 2/2 5–6 9.17 (±7.65) 0.10 (±0.02) Pe, As Cl, Pe Both Clear
2/2 11–15 1.28 (±1.53) 2.67 (±2.08) Cl, Pe, As Cl BothPearce et al. (1995) 6/6 8–9 0.91 (±1.15) 0.32 (±0.22) Pe, As, Cl Cl, As, Pe Both
6/6 16–17 0.45 (±0.32) 2.31 (±1.45) Pe, As, Cl Cl, As, Pe BothRabito et al. (2008) 20/54 6–7 0.08d 0.78d Cl, Pe, Tri, Pa Cl, Pe, Tri, Pa Both No
20/54 8–9 0.04d 0.15d NoRao et al. (2007) 11/11 1 6.95 (±0.015)e 0.98 (±0.004e) Pe, As, Cl, Ba As, Pe, Tr, Pa N/A ClearRao et al. (2007) 20/20 1 32.7 (±0.013)e N/ARoss et al. (2000) 17/44 12–19 2.19 (±3.34) As, Pe, Cl Cl, Pe Both N/AThis study 24/24 2–3 1.04 (±0.93) 1.34 (±1.54) Pe, Cl Cl, Pe Both no
N/A: not available; RFF: relationship between flooding and indoor fungi concentrations.a 0 is for non-affected house.b As: Aspergillus; Ba: Basidiopores; Pa: Paecilomyces; Pe: Penicillium; Cl: Cladosporium; Fu: Fusarium; Tr: Trichoderma.c In general, cleaning is following by remediation process, both: including both non-finished and finished.d Estimated data based on figure.e Geometric mean (geometric standard deviation).f But not clear 1 year later.
14 C. He et al. / Environment International 69 (2014) 9–17
molds that was found. However, the local context in which floodingoccurred could be expected to influence these results.
Only 3 studies have reported indoor andoutdoor bacteria concentra-tions in AFH. Ross et al. (2000) reported indoor bacteria concentrationlevels (1258 ± 786 cfu m−3) in 44 asthmatics' homes (17 of whichwere AFH), however no outdoor data was reported in their study. InCurtis et al. (2000), the mean airborne bacterial concentrations of 45homes (18 of which were AFH) were 1310 cfu m−3 indoors, with arange of 0–11,600 cfu m−3, and 1110 cfu m−3 outdoors, with a rangeof 0–9100 cfu m−3. Fabian et al. (2005) found that bacterial coloniescultured on impactor-mounted TSA plates ranged between 130 and1100 cfu m−3 indoors, and between 35 and 2700 cfu m−3 outdoors for8 AFH, with average indoor and outdoor values of 574 ± 382 cfu m−3
and 197 ± 135 cfu m−3, respectively. For 1 NFH, Fabian et al. (2005) re-ported that both indoor and outdoor bacteria concentrations were35 cfu m−3. The bacteria concentrations of the present study are lowerthan in the above studies, except for Fabian's NFH. While only threestudies have reported bacteria CFU, there were others that measured en-dotoxin, a proxy for gramnegative bacteria exposure. For example inNewOrleans homes after Hurricane Katrina (Chew et al., 2006; Riggs et al.,2008) the concentration of endotoxin in indoor air was much higher inAFH than in NFA.
In Brisbane, Hargreaves et al. (2003) found that the average indoorand outdoor fungi concentrations were 810 ± 389 cfu m−3 and1133 ± 759 cfu m−3, respectively, in 14 residential suburban housesunaffected by flooding. They also reported that themost frequently iso-lated fungi genera were Cladosporium, Curvularia, Alternaria, Fusariumand Penicillium. Based on a comparison of thesefindingswith the resultsof this study, it can be seen that both the outdoor and indoor averagemold concentrations reported by Hargreaves et al. (2003) were withinthe similar range, and in general, were lower than the concentrationsin AFH and higher than the concentrations in NFH during the firstround of this study. However, the most isolated fungal genus in indoorair was different between the two studies.
The indoor air mycoflora largely reflected the fungal flora presentin outdoor air. This study agrees with previous studies, in thatCladosporium was the dominant genus in outdoor air — a result thathas been found around the world during all seasons (WHO, 2009).
This study also supports earlier findings that Penicillium, a common fun-gal genus in indoor air, can easily grow on wet material, and thus, it isthe most common fungal genus detected in moisture damaged areas(Hyvärinen et al., 2002).
Generally, there are no uniformly accepted, or validated, quantita-tive environmental sampling methods with which to assess exposureto mold and other agents associated with damp indoor environments(ACGIH, 2009; Frankel et al., 2012; Mazur and Kim, 2006). The RCSHigh Flow instrument has a particle diameter cut off size (d50) of 2–5 μm (Millipore, 2003), which meets the cut off size requirements formost of the fungal spores in indoor environments (2 to 4 μm in aerody-namic diameter) (Reponen et al., 1994, 2001), and thus, the RCS HighFlow instrument is suitable for collecting indoor fungal spores.
Sampling duration is an important consideration in the determina-tion of collection efficiency. Short sampling times are associatedwith in-creased variability among samples and with results that in general arenot representative of the prevailing air microbiota. Longer samplingtimes are associated with lower limits of detection, and therefore aredesirable. Sampling techniques that support long-term samples havebeen emphasized (Flannigan et al., 2001). However, increased samplingduration was associated with decreased recovery of most generacollected. According to Saldanha et al. (2008) in situations where lowairborne spore loads are expected, sampling times of up to 6 min maybe reasonable for the RCS as well as for Andersen N6, but longersampling times may yield distorted results.
3.3. Dust elemental composition
In total, 41 indoor dust samples were collected, however 1 samplewas contaminated and re-collecting sample in this house was notachievable, leaving data for 40 samples, of which 23 and 17 were col-lected from AFH and NFH, respectively. The inorganic elemental con-centrations in the indoor dust from each house are shown in Fig. 3. Ofthe 24 elements assessed, 8 (Li, Be, V, Cr, Co, Ni, As and Ir) were belowthe detection limit in all samples. It can be seen from Fig. 3 that the con-centration of elements in the dust samples varied considerably for bothAFH and NFH. House 27 showed the highest mass concentration levelsfor all of the elements, which is likely to be the result of renovation
Fig. 3. Indoor dust inorganic elemental concentrations for each house.
15C. He et al. / Environment International 69 (2014) 9–17
works thatwere being conducted in the house during the sample collec-tion period, and therefore it was excluded from further statistical anal-yses. The variation of total mass concentration for all elements amongthe remaining houses ranged from 741 to 5840 μgm−2, with an averageof 2296±1328 μgm−2, for AFH and from 657 to 3086 μgm−2, with theaverage of 1454± 678 μgm−2, for NFH. Overall, the total mass concen-tration for AFH showed a wider variation than that for NFH. Statisticalresults indicate that the total mass, Ca and Sr concentrations for AFHwere significantly higher than those for NFH. Although Na, Mg, Al, Cu,Zn, Pb, as well as the total concentration of heavy metals (Mn, Fe, Cu,Zn, Cd, Pb) also showed higher concentrations in AFH than in NFH(see Fig. S3 in Supporting information), these differences were not sta-tistically significant.
The percentage contribution of each element to the total elementalconcentration for the two groups of houses is presented in Supportinginformation Fig. S4. The first 5 elements (Na, Ca, Al, Fe and K) constitut-ed a dominant fraction of the total dust (93.4%) in AFH (see Fig. S4). ForNFH, the 5 dominant elements were Na, Ca, K, Fe and Al (95.5%). There-fore, the elemental composition of dust from the two groups of houseswas similar.
Again, therewas nodust elemental composition data available in theliterature for AFH in the area of this study. In Brisbane, Robertson et al.(2005) collected and analyzed house dust from 12 residential houses(NFH). A list of themost abundant elements (%) in this and Robertson'sstudy is given in Supporting information Table S9, the most abundantelement, sodium (Na), was not measured and lead (Pb) was not detect-ed. Therefore, while direct comparison is not possible, it can be seen thatthe percentages of Fe, Pb and Zn in this study, in both AFH and NFH,were higher than those measured in Robertson's study.
Based on these results (some significant differences, some similari-ties, lacking literature data) and considering indoor renovation activitiescould have contributed to the observed differences between the twogroups of houses, no conclusion can be drawn from the effect of theflood by the elemental composition or concentration of indoor dustdate in AFH. While the results do not point to a marked effect of theflood, a more focused study would have been required to distinguishthe effects, if there were any.
3.4. Impact of remediation process on indoor air
Based on the information gathered from the questionnaire, all of theAFH were cleaned within one week after the flood water had receded.
The cleaning methods used in these houses included just water, waterplus detergent, water plus bleach, water plus disinfectant, water plusdetergent and bleach, water plus disinfectant, detergent and bleach,and water plus insecticide. In order to determine the impact of cleaningmethods on indoor air quality, all of the AFH were classified into twogroups: ‘water only’ and ‘all other cleaning methods’ (see Table S5).The results showed that indoor concentrations of PM10, PN and fungiin the water only houses were lower than those in houses that usedall other cleaningmethods, although these differences were not statisti-cally significant. Further analysis revealed that all six of the water onlyhouses were cleaned completely and immediately (one or two days)after the flood water receded. For indoor dust chemical concentrations,nearly all of tested elements were lower in the water only, than thosehouses that used all other cleaning methods. The ratios of elementalconcentrations in the houses that used water only versus all othercleaning methods were all less than one. Statistical results indicatedthat Na (0.64), Al (0.41), Ca (0.41), Fe (0.47), Zn (0.19), Sr (0.25) andtotal (0.51) elemental concentrations were significantly different be-tween the two groups of houses.
In order to compare the effect of remediation (or renovation) on airquality, all of the AFH were further classified into two groups, based onwhether the remediation had been completed or was still in progress.There were 8 ‘in-progress’ houses and 16 ‘completed’ houses in thefirst round of measurements. A comparison of the concentration resultsfor the two group houses is given in Table S6 in Supporting informationand shows that while there were some differences between the twogroups of houses, none of them were statistically significant, except in-door fungi concentration. Similarly, there were no significant differ-ences between the two groups of houses in terms of indoor dustchemical concentrations, although the total chemical elemental concen-tration was slightly lower in the houses with remediation still in prog-ress compared to those in which remediation was completed.
Further analysis showed that both detergent and bleach were usedin the cases for which the I/O bacteria ratio was lower than 0.80, whilehouses with an I/O bacteria ratio higher than 3.00 were cleaned onlywith water, only detergent or only bleach. Statistical analyses showedthat the average indoor bacterial concentration in the first group ofhouses was significantly lower than in the second, which implied thatusing both detergent and bleach to clean the houses was a betterway to reduce indoor bacteria concentration levels. In contrast, theresults gave no clear indication of the most effective cleaningmethods to reduce indoor fungi concentrations, since many factors
16 C. He et al. / Environment International 69 (2014) 9–17
(especially indoor sources and sampling location) can affect I/O fungiratios.
3.5. Correlations
The correlations between indoor and outdoor concentrations ofPM10, PN, fungi and bacteria concentrations were investigated for bothgroups of houses. Summaries of these results are given in Table S7 andTable S8 in Supporting information, respectively. For AFH, therewere statistically significant positive correlations between indoor andoutdoor PN concentrations, indoor PN and indoor and outdoor PM10
concentrations, indoor and outdoor fungi concentrations, and also be-tween outdoor fungi and outdoor bacteria concentrations.
For NFH, in addition to the correlationswhichwere observed in AFH,therewere significant positive correlations between indoor and outdoorPM10 concentrations, outdoor PN and PM10 concentrations, and nega-tive correlations between indoor fungi and indoor and outdoor bacteriaconcentrations.
For both AFH andNFH, therewere no statistically significant correla-tions between dust inorganic element concentrations and particle(PM10 and PN) or bioaerosol (fungi and bacteria) concentrations.There was also no correlation between particle and bioaerosolconcentrations.
4. Conclusion
This is the first study to report on comprehensive investigations ofthe following set of parameters in residential houses after a majorflooding event: indoor and outdoor PN, PM10, fungi and bacteria con-centrations, as well as indoor dust elemental concentrations. In general,the study showed that, as expected, the average indoor and outdoorlevels for these parameters varied by up to one order of magnitudenot only for the houses which were flooded, but also for the houseswhich were not flooded. Median I/O ratios were higher than or closeto one for PN, PM10 and bacteria, but not for fungi. Although therewere some higher average or median concentration levels in AFH thanin NFH, these differences were not statistically significant, except forsome inorganic elements. For both AFH and NFH, there was no any sta-tistically significant correlation between dust inorganic elemental con-centrations and other particle (PM10 and PN) concentrations, as wellas bioaerosol (fungi and bacteria) concentrations. Therewas also no sta-tistically significant correlation between particle (PM10 and PN) concen-trations and bioaerosol (fungi and bacteria) concentrations. However,there were statistically significant correlations between indoor andoutdoor PN and fungi concentrations, respectively, for AFH, as well asbetween indoor and outdoor PM10, PN, and fungi concentrations, re-spectively, for NFH.
The results of this study suggest that there were no quantifiable ef-fects of the flood in Brisbane on indoor air quality, in terms of PM10,PN, fungi and bacteria concentrations. The main reasons for these re-sults are likely to be the immediate removal of flooded materials andthe extensive clean-up, both indoors and outdoors, immediately afterthe flood water had receded. These results also show that the BrisbaneCity Council Local Disaster Management Group and the “Mud Army”of volunteers made a very important contribution to the after floodmanagement and recovery process.
Acknowledgments
The study was partially funded by the Institute of Health and Bio-medical Innovation and the Science and Engineering Faculty, Queens-land University of Technology. We would like to thank to: members ofthe ILAQH, QUT, in particular, Dr. James Smith, Ms Rachael ApplebyandMr. SamClifford, for their assistance in numerousways, Dr. CarolineDuchaine from University Laval, Quebec, Canada and Dr. Tiina Reponenfrom University of Cincinnati, USA for their helpful discussions in
relation to this project; Mr Don Neale from the Queensland Departmentof Science, Information Technology, Innovation and the Arts (DoSITIA)for supplying the EPA data, and the owners and occupants of the housesfor their help and assistance with this project.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.envint.2014.04.001.
References
ACGIH. Guide to occupational exposure values. Cincinnati, Ohio: American Conference ofGovernmental Industrial Hygienists; 2009.
Adhikari A, Jung J, Reponen T, Lewis JS, DeGrasse EC, Grimsley LF, et al. Aerosolization offungi, (1 → 3)-beta-D glucan, and endotoxin from flood-affected materials collectedin New Orleans homes. Environ Res 2009;109:215–24.
Ahern M, Kovats R, Wilkinson P, Few R, Matthies F. Global health impacts of floods: epi-demiologic evidence. Epidemiol Rev 2005;27:36–46.
Alderman K, Turner LR, Tong S. Floods and human health: a systematic review. EnvironInt 2012;47:37–47.
Bartlett K, Kennedy SM, Brauer M, Van Netten C, Dill B. Evaluation and a predictive modelof airborne fungal concentrations in school classrooms. Ann Occup Hyg 2004;48:547–54.
Bloom E, Grimsley LF, Pehrson C, Lewis J, Larsson L. Molds and mycotoxins in dust fromwater-damaged homes in New Orleans after hurricane Katrina. Indoor Air 2009;19:153–8.
CEN. Fungal diseases. Cladosporium. Other pathogenic fungi. Centers for Disease Controland Prevention; 2014 [Available online at: http://www.cdc.gov/fungal/diseases/other/cladosporium.html].
Chew GL, Wilson J, Rabito FA, Grimsley F, Iqbal S, Reponen T, et al. Mold and endotoxinlevels in the aftermath of Hurricane Katrina: a pilot project of homes in New Orleansundergoing renovation. Environ Health Perspect 2006;114:1883–9.
Cho SJ, Ramachandran G, Grengs J, Ryan AD, Eberly LE, Adgate JL. Longitudinal evaluationof allergen and culturable fungal concentrations in inner-city households. J Occup En-viron Hyg 2008;5:107–18.
Curtis L, Ross M, Persky V, Scheff P, Wadden R, Ramakrisnan V, et al. Bioaerosol concen-trations in the quad cities 1 year after the 1993 Mississippi river floods. IndoorBuilt Environ 2000;9:35–43.
Du W, FitzGerald GJ, Clark M, Hou X. Health impacts of floods. Prehosp Disaster Med2010;25:265–72.
FabianMP, Miller SL, Reponen T, HernandezMT. Ambient bioaerosol indices for indoor airquality assessments of flood reclamation. J Aerosol Sci 2005;36:763–83.
Flannigan B, Samson RA, Miller JD, editors. Microorganisms in home and indoor work en-vironments: diversity, health impacts, investigation and control. New York: Taylor &Francis; 2001.
Frankel M, Bekö G, Timm M, Gustavsen S, Hansen EW, Madsen AM. Seasonal variation ofindoor microbial exposures and their relations to temperature, relative humidity andair exchange rates. Appl Environ Microbiol 2012;78:8289–97.
Hargreaves M, Parappukkaran S, Morawska L, Hitchins J, He C, Gilbert D. A pilot investiga-tion into associations between indoor airborne fungal and non-biological particleconcentrations in residential houses in Brisbane, Australia. Sci Total Environ 2003;312:89–101.
Hoppe KA, Metwali N, Perry SS, Hart T, Kostle PA, Thorne PS. Assessment of airborne ex-posures and health in flooded homes undergoing renovation. Indoor Air 2012;22:446–56.
Hsu NY, Chen PY, Chang HW, Su HJ. Changes in profiles of airborne fungi in floodedhomes in southern Taiwan after Typhoon Morakot. Sci Total Environ 2011;409:1677–82.
Hyvärinen A, Meklin T, Vepsäläinen A, Nevalainen A. Fungal and actinobacteria inmoisture-damaged building materials— concentrations and diversity. Int BiodeteriorBiodegrad 2002;49:27–37.
Jo WK, Seo YJ. Indoor and outdoor bioaerosol levels at recreation facilities, elementaryschools, and homes. Chemosphere 2005;61:1570–9.
Khan NN, Wilson BL. An environmental assessment of mold concentrations and potentialmycotoxin exposures in the greater Southeast Texas area. J Environ Sci Health A ToxHazard Subst Environ Eng 2003;38:2759–72.
Lim MCH, Ayoko GA, Morawska L, Ristovski ZD, Jayaratne ER, Kokot S. A comparativestudy of the elemental composition of the exhaust emissions of cars powered by liq-uefied petroleum gas and unleaded petrol. Atmos Environ 2006;40:3111–22.
Mandal J, Brandl H. Bioaerosols in indoor environment — a review with special referenceto residential and occupational locations. Open Environ Biol Monit J 2011;4:83–96.
Mazur LJ, Kim J. Spectrum of noninfectious health effects from molds. Pediatrics 2006;118:E1909–26.
Mentese S, Rad AY, ArisoyM, Gullu G, ArisoyM, Gullu G. Seasonal and spatial variations ofbioaerosols in indoor urban environments, Ankara, Turkey. Indoor Built Environ2012;21:797–810.
Metts T. Addressing environmental health implications of mold exposure after majorflooding. AAOHN J 2008;56:115–20.
Millipore. Monitoring airborne microorganisms during food and beverage processing.Comparison of four microbial air samplers. Billerica, MA: Millipore; 2003. p. 1–4.[Technical brief].
17C. He et al. / Environment International 69 (2014) 9–17
Morawska L, He C, Hitchins J, Gilbert D, Parappukkaran S. The relationship between in-door and outdoor airborne particles in the residential environment. Atmos Environ2001;35:3463–73.
Pearce M, Huelman PH, Janni KA, OlsenW, Seavey RT, Velsey D. Long-term monitoring ofmold contamination in flooded homes. J Environ Health 1995;58:6–11.
Rabito F, Iqbal S, Kiernan M, Holt E, Chew G. Children's respiratory health andmold levelsin New Orleans after Katrina: a preliminary look. J Allergy Clin Immunol 2008;121:622–5.
Rao CY, Riggs MA, Chew GL, Muilenberg ML, Thorne PS, Van Sickle D, et al. Characteriza-tion of airborne molds, endotoxins, and glucans in homes in New Orleans afterHurricanes Katrina and Rita. Appl Environ Microbiol 2007;73:1630–4.
Reponen T, Hyvarinen A, Ruuskanen J, Raunemaa T, Nevalainen A. Comparison of concen-trations and size distributions of fungal spores in buildings with and without mouldproblems. J Aerosol Sci 1994;25:1595–603.
Reponen T, Willeke K, Grinshpun S, Nevalainen A. Biological particle sampling. In: BaronP, Willeke K, editors. Aerosol measurement: principles, techniques, and applications.New York: John Wiley and sons Inc.; 2001.
Riggs MA, Rao CY, Brown CM, Van Sickle D, Cummings KJ, Dunn KH, et al. Residentcleanup activities, characteristics of flood-damaged homes and airborne microbi-al concentrations in New Orleans, Louisiana, October 2005. Environ Res 2008;106:401–9.
Robertson S, Ayoko G, Duigu J. Elemental & PAH compositions of house dust in Brisbane,Australia. Proceedings: 10th International Conference on Indoor Air Quality andClimate, 1-5. 2005. p. 1536–40.
Ross MA, Curtis L, Scheff PA, Hryhorczuk DO, Ramakrishnan V, Wadden RA, et al. Associ-ation of asthma symptoms and severity with indoor bioaerosols. Allergy 2000;55:705–11.
Saldanha R, Saldanha R, Manno M, Saleh M, Ewaze JO, Scott JA. The influence of samplingduration on recovery of culturable fungi using the Andersen N6 and RCS bioaerosolsamplers. Indoor Air 2008;18:464–72.
Salonen H, Lappalainen S, Lindroos O, Harju R, Reijula K. Fungi and bacteria in mould-damaged and non-damaged office environments in a subarctic climate. Atmos Envi-ron 2007;41:6797–807.
Schwab KJ, Gibson KE, Williams DL, Kulbicki KM, Lo CP, Mihalic JN, et al. Microbial andchemical assessment of regions within New Orleans, LA impacted by HurricaneKatrina. Environ Sci Technol 2007;41:2401–6.
Solomon GM, Hjelmroos-Koski M, Rotkin-Ellman M, Hammond SK. Airborne mold andendotoxin concentrations in New Orleans, Louisiana, after flooding, October throughNovember 2005. Environ Health Perspect 2006;114:1381–6.
Taylor J, Biddulph P,DaviesM, Lai KM. Predicting themicrobial exposure risks in urbanfloodsusing GIS, building simulation, and microbial models. Environ Int 2013;51:182–95.
Taylor J, Lai KM, Davies M, Clifton D, Ridley I, Biddulph P. Flood management: predictionof microbial contamination in large-scale floods in urban environments. Environ Int2011;37:1019–29.
WHO. WHO guidelines for indoor air quality: dampness and mould. Copenhagen: WorldHealth Organization; 2009 [248 pages].
Wu PC, Su HJ, Lin CY. Characteristics of indoor and outdoor airborne fungi at suburbanand urban homes in two seasons. Sci Total Environ 2000;253:111–8.
