PAPER www.rsc.org/jem | Journal of Environmental Monitoring

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online / Journal Homepage / Table of Contents for this issue

Background culturable bacteria aerosol in two large public buildings usingHVAC filters as long term, passive, high-volume air samplers

Nicholas J. Stanley,*a Thomas H. Kuehn,a Seung Won Kim,b Peter C. Raynor,b

Senthilvelan Anantharaman,c M. A. Ramakrishnanc and Sagar M. Goyalc

Received 14th December 2007, Accepted 19th February 2008

First published as an Advance Article on the web 7th March 2008

DOI: 10.1039/b719316e

Background culturable bacteria aerosols were collected and identified in two large public buildings

located in Minneapolis, Minnesota and Seattle, Washington over a period of 5 months and 3 months,

respectively. The installed particulate air filters in the ventilation systems were used as the aerosol

sampling devices at each location. Both pre and final filters were collected from four air handing units

at each site to determine the influence of location within the building, time of year, geographical

location and difference between indoor and outdoor air. Sections of each loaded filter were eluted with

10 ml of phosphate buffered saline (PBS). The resulting solutions were cultured on blood agar

plates and incubated for 24 h at 36 �C. Various types of growth media were then used for

subculturing, followed by categorization using a BioLog MicroStation (Biolog, Hayward, CA, USA)

and manual observation. Environmental parameters were gathered near each filter by the

embedded on-site environmental monitoring systems to determine the effect of temperature,

humidity and air flow. Thirty nine different species of bacteria were identified, 17 found only in

Minneapolis and 5 only in Seattle. The hardy spore-forming genus Bacillus was the most commonly

identified and showed the highest concentrations. A significant decrease in the number of species

and their concentration occurred in the Minneapolis air handling unit supplying 100% outdoor air

in winter, however no significant correlations between bacteria concentration and environmental

parameters were found.

Introduction

Increasing emphasis is being placed on the detection of bioaero-

sols in indoor environments.5,9,11,17,22,26,27,29,32,37,48,49 Bioaerosols

are omnipresent in the ambient environment as they are

produced and dispersed through various natural processes which

occur both indoors and outdoors. Bioaerosols can penetrate an

indoor environment from outdoor sources through purposeful

opening of doors and windows, cracks in the building envelope,

or the outdoor air intake of air handling units (AHU).3 Biocon-

tamination of indoor materials occurs as airborne microorga-

nisms are deposited on indoor surfaces. This allows building

occupants to be exposed through dermal contact, ingestion,

and inhalation, although adverse health affects are most

commonly the result of inhalation.5

The microorganisms that compose bioaerosols, such as

bacteria, fungi, and viruses, can cause adverse health affects as

well as contribute to poor indoor air quality. Wallemia sebi is

a fungus suspected to be a causative agent of farmer’s lung

aEnvironmental Division, Department of Mechanical Engineering, Instituteof Technology, University of Minnesota, 111 Church Street SE,Minneapolis, MN, 55455, USA. E-mail: stan0332@umn.edu; Fax:+1-612-625-6069; Tel: +1-612-625-1510bDivision of Environmental Health Sciences, School of Public Health,University of Minnesota, 420 Delaware Street SE, Minneapolis, MN,55455, USAcDepartment of Veterinary Population Medicine, College of VeterinaryMedicine, University of Minnesota, 1333 Gortner Avenue, St. Paul, MN,55108, USA

474 | J. Environ. Monit., 2008, 10, 474–481

disease.53 Mycobacterium tuberculosis is a huge public health

issue and can easily be transmitted from person to person as

an aerosol.7 If a microorganism is not pathogenic it still contri-

butes to poor indoor air quality and may be a contributing factor

for building-related illness (BRI). Certain allergies can also be

triggered by a variety of microorganisms.24

Various methods have been used to sample bioaerosols in

order to determine microbial viability. Portable bioaerosol

sampling devices were used and compared by Yoa and

Mainelis with varying results.52 Other instruments such as

impingers,27,29,33,40 impactors,8,9,14,17,37,41,48,49 and a combination

of instruments2,6,11,23,26,30 have been utilized for measuring cultu-

rable airborne microorganisms.7,33,34 A qPCR assay method,

along with a Nucleopore filter for aerosol sampling was used

by Chen et al. and proved to be accurate for real-time microor-

ganism quantification.7,47 Sampling bioaerosols can also be

accomplished using filtration media, including the existing venti-

lation filters within the AHU of a building.5,7,12,15,20,22,35,39,43,51,53

These filters are used to prevent the transmission of particles

through AHUs, but are not 100% efficient. Filters also do not

maintain the viability of microorganisms as well as other

sampling methods,6,30 and can serve as sources of indoor

airborne bacteria as organic/inorganic nutrients that may

support growth are deposited on the filter along with microor-

ganisms.20,25,35,43,51 Microbial growth can also occur when

particulate loaded filters are exposed to high humidity or

a high moisture situation (e.g. if standing water were in contact

with the filter).21 Kemp et al. found an equilibrium point where

This journal is ª The Royal Society of Chemistry 2008

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

the number of microorganisms dying on the filter was equal to

the number being captured plus the small number added by

reproduction.20

Other indoor sources are present as well; humans have been

identified as a primary contributor9,48 and local microbial growth

has been indicated as a possible source.37,49 Point sources are

difficult to identify, so most remain unknown and can be mobile,

such as a sick person. Even so, viable bacteria aerosol concentra-

tion has been shown to constitute less than 1% of total ambient

particle concentration.48 Measured indoor concentrations of

viable bacteria range from 101 to 8.8 � 106 CFU m�3 as reported

by various sources.26,33 Specific species of bacteria identified

through several methods have been published in the literature

and are shown in Table 1.

Environmental parameters play a role in microorganism

survival on filter media.36 In laboratory studies, relative humidity

has been shown to affect the survivability of various bacterial

aerosols;35,40 and temperature and relative humidity have been

suggested to affect growth rates of bacteria on loaded filter

media.25 However, field trial data between indoor viable bacteria

and temperature,9,14,37,48 humidity,9,14,37 and seasonal varia-

tions9,22,26,37,41,49 are inconclusive. Room temperature affects

the dispersal of bacterial aerosols48 but indoor survivability

is not affected by temperature or humidity.9,14,37 Seasonal

airborne bacteria differences have been identified in some

studies22,26,41 with concentrations in summer higher than in

winter. No overall seasonal correlations were observed in other

studies9,37,49 but seasonal variations may depend on the specific

residence.37 Storage conditions and filter life may also affect

Table 1 Indoor culturable bacteria aerosols identified in the literature

Ref Author(s) Environment Method

15 Gorny et al. Human dwellings Impactor

42 Pastuszka et al. Homes Impactor

2 Awad Flourmill Impinger andgravimetric samp

38 Moschandreas et al. Homes Impactor

51 Tsai and Macher Large officebuilding

Impactor

23 Kim et al. Schools Filters (small nu

35 Lin et al. Office building Impinger8 Chen and Li Hospital isolation

roomsSmall filter withanalysis

40 Nilsson et al. Homes Electrostatic dus

This journal is ª The Royal Society of Chemistry 2008

the viability of microorganisms collected on filter media.30,53 Li

et al. recovered hardy spores from filters more readily than

sensitive strains and found storage effects were insignificant for

hardy spores.30

Outdoor airborne bacterial concentrations have been found to

be different than those indoors in most cases,9,11,17,37,49 while

comparable in others.26,29 As with indoors, outdoor concentra-

tions show a seasonal dependency,22,23,26,41 with higher values

in summer than in winter. Kujundzic et al. found a 9-fold

decrease from summer to winter.26 There have been no specific

correlations between ambient temperature and ambient outdoor

bacterial concentrations; however more microbial activity has

been suggested to be the result of warmer conditions.22 Directly

comparing indoor and outdoor concentrations shows indoor

concentrations to be higher in some studies11,17,37 and outdoor

levels higher in others,11,22 but this difference does not seem

to depend on season or location. Outdoor viable fungal

concentrations show a correlation between season and relative

humidity.31 Overall, the correlations between ambient

environmental parameters and bioaerosol concentrations are

inconclusive.

Different geographical locations may be conducive to the pres-

ence of different strains of bioaerosols.18 The natural

surrounding areas (agricultural, urban, rural, etc.) impact the

background bioaerosol concentrations.4,10,13,16,32,42,50

The goal of this study was to quantify background culturable

bacteria aerosols in public buildings at two geographical

locations in the United States. The presence of potential biolog-

ical threat agents and their near neighbors was a high priority.

Bacteria Species Location

Species of: Micrococcus/Kocuria, Staphylococcus,Bacillus,Pseudomonadaceae,Aeromonas and Nocardia

Southern Poland

Species of Micrococcus,Staphylococcus, and Bacillus(and many other species)

Poland

lersBacillus, Micrococci,Staphylococci, Streptococci,Klebsiella, Pseudomonas,Corynebacteria, Diplococci,Sarcina, Tetrads,Acinetobacter, Alcaligenes,and Enterobacter

Giza, Egypt

Gram positive bacteria andStaphylococcus

Chicago, Illinois, USA

Gram positive rods (Bacillusspp. and Actinomycetes),Gram positive cocci, Gramnegative rods, and Grampositive cocci

100 locations in USA

cleopore) Species of Bacillus,Streptomyces, andPseudomonas

Sweden

Actinomycetes North Carolina, USAqPCR Mycobacterium tuberculosis Taipei, Taiwan

t sampler Species of Bacillus,Pseudomonas, andStreptomyces

Sweden

J. Environ. Monit., 2008, 10, 474–481 | 475

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

The approach was to use the existing filters within the building

AHUs, which required no additional samplers, and removed

the bacteria aerosols from large air volumes. This method

allowed the detection of very low concentration levels.12 Results

were compared with average environmental conditions experi-

enced by each filter during the service life. Other comparisons

of culturable bacteria aerosols included: outdoor vs. indoor,

recovery from prefilters vs. final filters, fresh vs. shipped filters,

geographical location and seasons of the year.

Methods

The use of existing AHU filters as bioaerosol collection devices

was selected because of the low sampling cost, ease of implemen-

tation, and the ability to be applied to nearly any building. The

sampling process is unobtrusive and does not affect normal

day-to-day activities within a building. The main disadvantage

is the potential loss of culturability as the bacteria may reside

on the filter for a long period of time prior to elution and

culturing. For the current study, all the filters in each filter

bank containing 12–40 filters were changed on a schedule set

by each site. Although only one filter from each bank was used

for analysis, all the filters were changed when a sample filter

was removed to ensure uniform loading across all the filters in

the bank.

The project was conducted in 2 stages. During the preliminary

stage, methods for filter removal and elution were developed

using 16 filters from the field along with several filters challenged

in the laboratory. The methods were refined and implemented

during the final stage using 48 filters from the field. The locations

of the buildings used in the final stage were Minneapolis, MN

and Seattle, WA, USA, to examine a mid-continental and

maritime climate, respectively. Four AHUs were selected for

sampling and filter retrieval within each building. The filters in

the three mixed air AHUs were supplied (and filtered) with

a combination of return air from the building and outdoor air.

The fourth AHU supplied 100% outdoor air. The existing

building environmental monitoring systems were used to gather

data from each AHU to provide mixed air temperature,

humidity and air flow rate experienced by each filter bank.

Corresponding outdoor air data were obtained from the

National Climatic Data Center (NCDC) for each site.38 Each

AHU contained a pair of filter banks composed of pre and final

filters. The prefilter bank utilized pleated panel filters with an

ASHRAE 52.1-1992 dust-spot efficiency of�30%. The final filter

bank consisted of bag filters with a dust-spot efficiency of �95%.

These efficiencies were provided in the manufacturer’s literature

for each filter.

A variety of filters were used at each site and in each AHU.

Airguard Type DP 40 (Airguard, Louisville, KY, USA) and

Aerostar (Filtration Group, Joliet, IL, USA) pleated filters

were used as prefilters and Airguard Clean-Pak (Airguard,

Louisville, KY, USA) and Aerostar SoniQ (Filtration Group,

Joliet, IL, USA) bag filters were used as final filters in Seattle.

In Minneapolis, Glasfloss Z-line HV (Glasfloss, Dallas, TX,

USA) and Koch K-40 series (Koch Filter Co., Louisville, KY,

USA) pleated filters were used as prefilters and Glasfloss Excel

G-7 (Glasfloss, Dallas, TX, USA) and Koch Multi-Sak (Koch

Filter Co., Louisville, KY, USA) bag filters were used as final

476 | J. Environ. Monit., 2008, 10, 474–481

filters. Each filter was composed of a cotton/polyester blend,

synthetic fibers, or fiberglass media.

Elution

The method used to recover culturable bacteria was almost

identical to the method outlined by Farnsworth et al.12 Entire

filters were not eluted because analyzing such large volumes of

eluate would have been too difficult. The bioaerosol concentra-

tion in the air stream was assumed to be uniform; therefore

a representative sample of each filter was removed, eluted and

tested instead of using the entire filter. Several random 1 inch2

sections were cut from each bag filter (final filter) and prefilter

using sterile scissors (�0.1–0.6% of each filter was removed for

elution). These samples were then placed in a large plastic tube

and eluted with 10 ml of phosphate buffered saline (PBS)

containing 1.0 mg ml�1 concentration of Fungizone (amphoteri-

cin B and Deoxycholate) to suppress the growth of fungi. The

samples were then agitated by hand shaking for 30 s and vortex

mixing for 60 s. The eluate was then decanted and used for

bacterial culture.

Bacteria culturing

50 ml of each volume of eluate was inoculated on blood agar

plates for initial culturing followed by incubation for 24 h at

36 �C. The bacterial colonies were then counted and observed

for morphology. Subculturing was conducted on various types

of agar media for colonies representing different morphologies.

Following this subculture, each species was identified using

a BioLog MicroStation (Biolog, Hayward, CA, USA) as

described below. The number of colonies with the same

morphology were counted in order to estimate the concentration

of each species of bacteria identified in the air stream.

The Biolog MicroStation

After initial categorization using simple gram staining and

biochemical tests (catalase, oxidase, and TSI slant tests), the

subcultures were inoculated onto 96-well plates purchased

from Biolog containing different carbon sources in each well.

If the inoculated bacteria metabolized the carbon source in

a particular well, a dye appeared in the well due to the respiration

of the microorganism. The MicroStation reader detected these

color changes and compared the color change pattern to a data-

base assembled for a broad range of bacterial species, including

pathogens. The best match to the database was considered the

most likely candidate for species identification. In some cases,

different strains of the same species could be identified with the

BioLog database. These were grouped together in their common

species for all subsequent results and discussion.

Estimation of average airborne concentration

The time-averaged concentration of culturable bacteria in the air

stream passing through a given filter bank was estimated by first

determining the number of culturable bacteria on the filter media

samples (raw bacteria counts) associated with each species. The

filter life, average air flow rate through the filter, filter fraction

This journal is ª The Royal Society of Chemistry 2008

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

sampled (FF), fraction of eluate used for culturing (%e), and

filter efficiency (h) were all considered in the following equations.

Filter Fraction ðFFÞ ¼ Surface area of pieces eluted

Total media surface area of filter

(1)

Actual Number ¼ Raw Counts

FF�%e(2)

Actual Number ¼ actual number of bacteria colony forming

units (CFU) on the filter

Raw Counts ¼ raw counts of bacteria colonies from the initial

culturing

%e ¼ Percent of eluate used for culturing, 0.005 in all cases

(dimensionless)

Several assumptions were made in these equations. The

recovery rate (RR) of bacteria from each filter section was

assumed to be 100%. The bacteria on the filter were assumed

to experience no loss of culturability or proliferation during

the life of the filter. The rated manufacturers’ dust spot efficien-

cies were used although filter efficiency varies with particle size

and filters become more efficient with dust cake build-up during

their life. There is currently no information available to adjust

for these assumptions; however they allow for a fair and consis-

tent comparison between filter sets and between bacteria species.

Once the actual number of colonies present on each filter is

obtained from eqn (2), the airborne concentration can be

estimated using the following expression.

Concentration ¼ ActualNumber � 35:32

RR � Q � Dt � hf

(3)

Concentration ¼ time averaged concentration of culturable

bacteria strain (CFU m�3)

Q ¼ average standard volumetric flow rate through the filter

(SCFM)

Dt ¼ life of the filter in the AHU (minutes)

hf ¼ filter efficiency (30% and 95% efficiency used for pre and

final filters, respectively)

RR ¼ recovery rate (dimensionless) ¼ 100%

35.32 is the conversion factor between m3 and ft3 (1 m3 ¼ 35.32

ft3)

The concentration of each species of bacteria identified was

determined for each set of filters. In Minneapolis, 3 sets of

prefilters were sampled in summer (08/01/05–9/14/05), fall

(9/14/05–11/01/05), and winter (11/01/05–12/15/05). One set of

final filters from Minneapolis (08/01/05–11/01/05) and one set

of pre and final filters from Seattle (9/14/05–12/14/05) were

sampled as well. The Minneapolis and Seattle change-out sched-

ules were staggered to provide a more uniform work load for

laboratory personnel. To begin the current study, fresh and ship-

ped filters were examined and compared from the Minneapolis

location. Two preliminary sets of filters were removed from the

Minneapolis site on 8/01/05 to determine the possible effects

shipping time has on bacteria culturability. One set was packed

and shipped to our laboratory by the standard procedure and

the other set of ‘‘fresh filters’’ was hand delivered, eluted and

cultured the same day. The lifetime of these initial filters is

This journal is ª The Royal Society of Chemistry 2008

unknown, as well as the airflow rate through each filter, so the

bacteria concentrations could not be determined.

Results and discussion

The shipped filters were not eluted until three days later, which

provided additional time for the microorganism concentrations

to change. The number of culturable colonies was expected to

be larger on the fresh filters than the shipped filters caused by

loss of culturability with time. However, the concentration of

most species was slightly higher on the shipped filters than the

fresh filters. During transit, the bacteria may have proliferated

instead of dying off or perhaps they had time to recover from

the air velocity shear and other adverse conditions within the

AHU. This is contrary to laboratory experiments conducted

during the preliminary stage of this study, where the loss of

culturability of Bacillus subtilis over time was documented using

clean filters.

During the fresh/shipped comparison B. subtilis was found on

only the fresh filters, indicating a loss of culturability over time

did occur with this species. There were four other species of

bacteria that appeared only on fresh filters; however all but

one (Staphylococcus hominis) of these were recovered from

shipped filters later in the study, including B. subtilis. Therefore

shipping time is not the only reason these species were initially

not recovered and survivability seems to be species dependant.

The average outdoor air temperature in Minneapolis changed

from 21.7 �C in the summer sample to �0.5 �C in the winter

sample but the average daily relative humidity remained nearly

constant changing from 65.9% to 69.9%. The average daily

temperature and relative humidity of the mixed air filters

changed from 22.0 �C and 55.6% to 17.0 �C and 19.3% between

summer and winter (averaged across the 3 mixed AHUs). In

Seattle, considerable variation in air flow rate through the

HCT and HMT AHUs was observed, whereas the other two

were more stable. The Minneapolis AHU airflow rates were

fairly consistent during the current study. Although a significant

change in environmental parameters was observed for the 3

seasons in Minneapolis, no statistically significant correlations

were found between culturable bacteria concentration and

environmental parameters (temperature, relative humidity,

humidity ratio, or percent outdoor air). In the current study,

these parameters were time averaged over 6–7 weeks; however,

environmental parameters changed significantly throughout the

day.

Several authors have reported variations in concentrations

that show a local area source can be responsible for a bioaero-

sol.10,13,16,18,32,42,50 This concentration variation is commonly

similar for pollen, inertly released fungal spores, and bacteria.18

Site surveys for relative locations of outdoor sources such as

agricultural practices,16,32 sewage treatment works,42 and

composting centers13 were not performed in this study. Sites

such as these have been found to contribute to outdoor ambient

bioaerosol concentrations, although agricultural practices only

significantly contribute during harvest and planting times.50

Pillai et al. concluded that there was little risk for airborne

transmission of bacterial bioaerosols 6 km away from sewage

treatment plants.42 Dowd et al. drew this line at 10 km downwind

from a sewage treatment plant.10 Folmsbee et al. found a point

J. Environ. Monit., 2008, 10, 474–481 | 477

Fig. 1 Concentration of bacteria fromMinneapolis prefilters, 8/01/05–9/

14/05 from different air handling units.

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

downwind from a compost center to have up to a 10-fold

increase in bioaerosols over locations unaffected by the compost

center.13 Factors such as filter shipping duration or site specific

environmental and ecological conditions may contribute to the

difference between the two locations more than their proximity

to specific outdoor sources.

The results from most field trials agree well with our data;

showing a seasonal difference but no correlation between bacte-

rial aerosols and environmental parameters.2,14,22,23,26,30,35,40,41,49

The species identified in the current study (Table 2) agree well

with bacteria identified in the ambient air from previous studies

(Table 1). Estimated airborne concentrations are on the low end

of the range determined in previous studies; however the filter

durations were very long compared to the sampling duration

used previously. Each filter sampled a large volume of air and

it is likely that most captured bacteria did not survive the entire

filter duration. Different species of bacteria exhibit unique

behavior under various environmental conditions. Some may

survive while others die off. Therefore it is not too surprising

to see no correlation between mean environmental conditions

and bioaerosol concentration when sampling a variety of

bacteria species. Comparing daily culturable bacteria concentra-

tions with daily environmental parameters in a real-time

measurement setup may reveal some beneficial correlations, see

Fig. 1 and 2.

Microorganism source depletion and die-off most likely

occurred during the winter months in Minneapolis. There were

only two bacterial colonies detected in the 100% outdoor AHU

in Minneapolis during this time (see Fig. 3), whereas in previous

filter samples the 100% outdoor AHU (S23) provided the most

species detected and the highest concentrations for most of the

Table 2 Culturable bacteria identified in Minneapolis and Seattle

Minneapolis & Seattle Minneapolis only Seattle only

Bacillusamyloliquefaciens

Bacillus badius Dermacoccusnishinomiyaensis

Bacillus cereus/thuringiensis

Bacillus fastidiosus Macrococcuscaseolyticus

Bacillus circulans Bacillus halodurans Paenibacillus popilliaeBacillus laevolacticus Bacillus racemilacticus Paenibacillus validusBacillus licheniformis Brevibacterium otitidis Tsukamurella

inchonensisBacillus maroccanus Curtobacterium

flaccumfaciensBacillus megaterium Deinococcus

radiopugnansBacillus mycoides Kurthia gibsoniiBacillus pumilus Microbacterium

laevaniformansBacillus sphaericus Paenibacillus

azotofixansBacillus subtilis Paenibacillus maceransBrevibacillus brevis Paenibacillus pabuliGeobacillusstearothermophilus

Staphylococcusarlettae

Geobacillusthermoglucosidasius

Staphylococcusepidermidis

Micrococcus luteus Staphylococcushominis

Paenibacillus larvae sslarvae

Staphylococcus warneri

Paenibacilluspolymyxa

Virgilbacilluspantothenticus

Fig. 2 Concentration of bacteria from Minneapolis prefilters, 9/14/05–

11/01/05 from different air handing units.

Fig. 3 Concentration of bacteria from Minneapolis prefilters, 11/01/05–

12/15/05 from different air handling units.

478 | J. Environ. Monit., 2008, 10, 474–481 This journal is ª The Royal Society of Chemistry 2008

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

culturable species. This could be due to lack of ideal survival

conditions for the bacterial colonies. Not only did the tempera-

ture and relative humidity change significantly from summer to

winter, but the amount of organic matter in the air and captured

on the filter could have also played a role. No correlations

between environmental parameters and culturable bacteria

concentrations were apparent. Source depletion and microor-

ganism die-off are most likely to have caused the change.

The species recovered from the Seattle pre and final filters are

shown in Fig. 4 and 5, respectively. The results from the Seattle

pre and final filters can be directly compared as they were in

service simultaneously. This trend was observed on the Minne-

apolis filters as well, although three species recovered from the

final filters were not found on the prefilters. This was not the

case in Seattle, where all of the species identified on the final

filters were also found on the prefilters. In Minneapolis, two

prefilters were used during the life of each final filter so the results

are not directly comparable.

Fig. 4 Concentration of bacteria from Seattle prefilters, 9/14/05–12/14/

05 from different air handling units.

Fig. 5 Concentration of bacteria from Seattle final filters, 9/14/05–12/

14/05 from different air handing units.

This journal is ª The Royal Society of Chemistry 2008

Results from the preliminary stage of this project also showed

a higher concentration of culturable bacteria found on the prefil-

ters than the final filters. However the prefilters are listed at 30%

efficient, the final filters are listed at 95% efficient, and these are

taken into account within our concentration calculations.

Removing these values from the calculation would bring the

prefilter and final filter concentration values closer, however

including these values is the only way to mathematically differen-

tiate between the two filter types. The prefilters remove much of

the airborne bacteria leaving little to be collected by the final

filters. The lowest airborne culturable bacteria concentration

determined in this study was 0.025 CFU m�3. Concentrations

below this value are too small to be detected.

The difference between our culturable bacteria results and the

listed filter efficiencies may be due to the dust that accumulated

on the prefilter forming a ‘‘dust cake.’’ These dust cakes were

observed on prefilters during the elution process, but not on

the final filters. A dust cake could provide adequate nutrients

for bacterial survival,1,43,44,45,51 and may shield the bacteria

from desiccation and other loss mechanisms. Dust cakes also

increase the filter capture efficiency. Airborne bacteria spores

could also be attached to larger particles or clumped together.46

If culturable bacteria travel as particles larger than single spores,

they should be captured by prefilters at a higher efficiency than

the rated filter dust-spot efficiency even when the filters are

new. Bacteria capture efficiency and survival within the filter

dust cake should be investigated.

Table 2 shows the bacteria identified from each location.

There were 39 different species of bacteria identified from the

filter media. 17 of the species appeared on filters from both

locations, 17 were found only on the Minneapolis filter media

and 5 were found only on the Seattle filter media. The most

common bacteria recovered were spore-forming species of the

genus Bacillus. Other identified species of interest include some

non-spore-forming bacteria such as: Brevibacterium otitidis

(Minneapolis), which is a cause of ear infections; Curtobacterium

flaccumfaciens (Minneapolis), which is associated with soybeans;

Deinococcus radiopugnans (Minneapolis), which is highly resis-

tant to radiation and desiccation; and Tsukamurella inchonensis

(Seattle), which is commonly associated with marine biofilms.

The previous studies outlined in Table 1 identify Bacillus,

Micrococcus, and Staphylococcus species as the most commonly

detected indoor ambient bioaerosols. Species of Staphylococcus

were only detected in Minneapolis, however Micrococcus luteus

was detected in both locations. Bacillus species were the most

frequently detected, and must have the ability to survive under

various indoor environmental conditions (as well as Staphylo-

coccus and Micrococcus species to a lesser extent). Due to their

prevalence in the literature and in the current study, indoor

sources of these species may also exist.

The bacteria species that appeared only on the mixed AHU

filters and not on the 100% outdoor AHU filters have sources

that can be found indoors. These species of bacteria include:

Bacillus cereus/thuringiensis, Bacillus mycoides, Bacillus sphaeri-

cus, Bacillus pumilus, Geobacillus thermoglucosidasius, Kurthia

gibsonii, Micrococcus luteus, Paenibacillus macerans, and

Staphylococcus arlettae. Humans, potted plants (including

biological pesticide and potting soil), and food that has been

cooked, spoiled, or improperly refrigerated are potential sources

J. Environ. Monit., 2008, 10, 474–481 | 479

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

of these bacteria. As each of these sources can be found indoors,

each of the bacteria species listed above can be produced

indoors.

Data regarding the survivability of each species of bacteria

present in the ambient environment is currently unavailable.

Most species recovered were hardy spore-forming bacteria;

however more fragile species such as vegetative bacteria may

have been present in the air and not survive to be identified

(such as Staphylococcus hominis). Only one filter in each AHU

filter bank, consisting of 12–40 filters, was removed for testing,

a very small sample of each filter material was removed for

elution, and only 0.5% of the elution was cultured. However,

after all of this dilution there were still 39 different species

detected on 48 different filters, many of which appeared on

several samples from several filters at each location.

The approach of using ventilation filters as bioaerosol collec-

tors should also work for fungi identification. Fungal growth

can occur with nutrients deposited on filter media,43,51 and

a similar technique has been implemented to measure bacterial

and fungal survival and growth on filter media.20 Several

different fungal species were identified from studies involving

airborne microorganism capture (with an impinger, impactor,

or filter) and culture methods.2,9,11,14,17,20,22,23,28,29,33,35,41,43,53

A year-long study involving more locations should provide

more data for the development of seasonal and geographic corre-

lations. Summer has been shown to have the highest bioaerosol

concentrations and the harvest season could possibly have an

impact, but we do not have data to demonstrate this trend,

shown in other studies.19,50 Real-time sampling could possibly

show how very specific climate changes affect specific species

of microorganisms, as the current work provides time-averaged

bacteria aerosol concentration data. This is not very useful

when trying to determine correlations for environmental param-

eters, which change significantly on a daily basis.

Conclusions

Sampling large volumes of air ensures that bacteria at very low

concentration levels have the possibility of being detected,

provided they remain culturable between the time of capture

and final identification. A total of 39 different species were

detected in the current study with the use of typical ventilation

filters as the bioaerosol sampling devices. The ventilation systems

required no alteration in order to implement this method. Other

bioaerosol sampling methods have been shown to be effective in

previous studies; however these methods typically involve addi-

tional air sampling instrumentation, low air sampling volumes,

and a small number of culturable bacteria species detected.

Acknowledgements

This study was supported by the Technical Support Working

Group (TSWG) with funding support from the US Department

of Homeland Security through contract W91CRB-04-C-0035.

References

1 D. G. Ahearn, S. A. Crow, R. B. Simmons, D. L. Price, S. K. Mishraand D. L. Pierson, Curr. Microbiol., 1997, 35, 305–308.

2 A. H. A. Awad, Aerobiologia, 2007, 23, 59–69.

480 | J. Environ. Monit., 2008, 10, 474–481

3 D. H. Bennett and P. Poutrakis, J. Aerosol Sci., 2006, 37, 766–785.4 A. Bovallius, B. Bucht, R. Roffey and P. Anas, Appl. Environ.Microbiol., 1978, 35(6), 1231–1232.

5 M. P. Buttner, P. Cruz, L. D. Stetzenbach, A. K. Klima-Comba,V. L. Stevens and T. D. Cronin, Appl. Environ. Microbiol., 2004,70(8), 4740–4747.

6 P. S. Chen and C. S. Li, Aerosol Sci. Technol., 2005, 39, 231–237.7 P. S. Chen and C. S. Li, Aerosol Sci. Technol., 2005, 39, 371–376.8 G. V. Crawford and P. V. Jones, Water Res., 1979, 13, 393–399.9 J. A. DeKoster and P. S. Thorne, Am. Ind. Hyg. Assoc. J., 1995, 56,573–580.

10 S. E. Dowd, C. P. Gerba, I. L. Pepper and S. D. Pillai, J. Environ.Qual., 2000, 29, 343–348.

11 M. P. Fabian, S. L. Miller, T. Reponen and M. T. Hernandez,J. Aerosol Sci., 2005, 36, 763–783.

12 J. E. Farnsworth, S. M. Goyal, S. W. Kim, T. H. Kuehn,P. C. Raynor, M. A. Ramakrishnan, S. Anantharaman andW. Tang, J. Environ. Monit., 2006, 8, 1006–1013.

13 M. Folmsbee and K. A. Strevett, J. Air Waste Manage. Assoc., 1999,49, 554–561.

14 R. L. Gorny, J. Dutkiewicz and E. Krysinska-Traczyk, Ann. Agric.Environ. Med., 1999, 6, 105–113.

15 S. A. Grinshpun, A. Adhikari, T. Honda, K. Y. Kim, M. Toivola,K. S. R. Rao and T. Reponen, Environ. Sci. Technol., 2007, 41,606–612.

16 B. A. Holmen, T. A. James, L. L. Ashbaugh and R. G. Flocchini,Atmos. Environ., 2001, 35, 3265–3277.

17 W. K. Jo and J. H. Kang, Chemosphere, 2006, 65, 1755–1761.18 A. M. Jones and R. M. Harrison, Sci. Total Environ., 2004, 326, 151–

180.19 B. L. Jones and J. T. Cookson, Appl. Environ. Microbiol., 1983, 45,

919–934.20 P. C. Kemp, H. G. Neumeister-Kemp, G. Lysek and F. Murray,

Atmos. Environ., 2001, 35, 4739–4749.21 S. J. Kemp, T. H. Kuehn, D. Y. H. Pui, D. Velsey and A. J. Streifel,

ASHRAE Trans., 1995, 101(1), 305–316.22 J. L. Kim, L. Elfman, Y. Mi, G. Wieslander, G. Smedje and

D. Norback, Indoor Air, 2007, 17, 153–163.23 K. Kruczalak, K. Olanczuk-Neyman and R. Marks, Pol. J. Environ.

Stud., 2002, 11(5), 531–536.24 T. H. Kuehn, D. Y. H. Pui, C. D. Berg, D. Vesley and M. Peloquin,

ASHRAE Trans., 1991, 97(2), 164–169.25 T. H. Kuehn, J. Sol. Energy Eng., 2003, 125, 366–371.26 E. Kujundzic, M. Hernandez and S. L. Miller, Indoor Air, 2006, 16,

216–226.27 E. Kujundzic, D. A. Zander, M. Hernandez, L. T. Angenent,

D. E. Henderson and S. L. Miller, J. Air Waste Manage. Assoc.,2005, 55, 210–218.

28 C. R. Kuske, Curr. Opin. Biotechnol., 2006, 17, 291–296.29 C. S. Li and T. Y. Huang, Aerosol Sci. Technol., 2006, 40, 237–241.30 C. S. Li and Y. C. Lin, Sci. Total Environ., 2001, 278, 231–237.31 C. M. Liao, W. C. Luo, S. C. Chen, J. W. Chen and H. M. Liang,

Atmos. Environ., 2004, 38, 4415–4419.32 B. Lighthart, Appl. Environ. Microbiol., 1984, 47, 430–432.33 X. Lin, T. A. Reponen, K. Willeke, S. A. Grinshpun, K. K. Foarde

and D. S. Ensor, Atmos. Environ., 1999, 33, 4291–4298.34 X. Lin, T. Reponen, K. Willeke, Z. Wang, S. A. Grinshpun and

M. Trunov, Aerosol Sci. Technol., 2000, 32, 184–196.35 R. Maus, A. Goppelsroder and H. Umhauer, Atmos. Environ., 2001,

35, 105–113.36 M. Moritz, H. Peters, B. Nipko and H. Ruden, Int. J. Hyg. Environ.

Health, 2001, 203, 401–409.37 D. J. Moschandreas, K. R. Pagilla and L. V. Storino, Aerosol Sci.

Technol., 2003, 37, 899–906.38 National Climatic Data Center (NCDC) web address (NCDC

website - http://www.ncdc.noaa.gov/oa/ncdc.html).39 A. Nilsson, E. Kihlstrom, V. Lagesson, B. Wessen, B. Szponar,

L. Larsson and C. Tagesson, Indoor Air, 2004, 14, 74–82.40 T. Paez-Rubio and J. Peccia, J. Environ. Eng., 2005, 113(4), 512–517.41 J. S. Pastuzka, U. Kyaw Tha Paw, D. O. Lis, A. Wlazlo and K. Ulfig,

Atmos. Environ., 2000, 34, 3833–3842.42 S. D. Pillai, K. W. Widmer, S. E. Dowd and S. C. Ricke, Appl.

Environ. Microbiol., 1996, 62, 296–299.43 D. L. Price, R. B. Simmons, S. A. Crow and D. G. Ahearn, J. Ind.

Microbiol. Biotechnol., 2005, 32, 319–321.

This journal is ª The Royal Society of Chemistry 2008

Publ

ishe

d on

07

Mar

ch 2

008.

Dow

nloa

ded

by T

empl

e U

nive

rsity

on

28/1

0/20

14 0

5:28

:50.

View Article Online

44 R. B. Simmons and S. A. Crow, J. Ind. Microbiol., 1995, 14, 41–45.45 R. B. Simmons, D. L. Price, J. A. Noble, S. A. Crow and

D. G. Ahearn, Am. Ind. Hyg. Assoc. J., 1997, 58, 900–904.46 N. J. Stanley, Master’s Thesis, University of Minnesota, June 2007.47 L. D. Stetzenbach, M. P. Buttner and P. Cruz, Curr. Opin.

Biotechnol., 2004, 15, 170–174.48 K. W. Tham and M. S. Zuraimi, Indoor Air, 2005, 15(Suppl 9),

48–57.

This journal is ª The Royal Society of Chemistry 2008

49 F. C. Tsia and J. M. Macher, Indoor Air, 2005, 15, 71–81.50 Y. Tong and B. Lighthart, Aerosol Sci. Technol., 2000, 32, 393–403.51 M. C. Verdenelli, C. Cecchini, C. Orpianesi, G. M. Dadea and

A. Cresci, J. Appl. Microbiol., 2003, 94, 9–15.52 M. Yoa and G. Mainelis, J. Exposure Sci. Environ. Epidemiol., 2007,

17, 31–38.53 Q. Y. Zeng, S. O. Westermark, A. Rasmuson-Lestander and

X. R. Wang, Appl. Environ. Microbiol., 2004, 70(12), 7295–7302.

J. Environ. Monit., 2008, 10, 474–481 | 481