Relationship between sampling duration and concentration of culturable airborne mould and bacteria on selected culture media


Diana Godish, Department of Physiology and Health Science, Ball State University, Muncie, IN 47306, USA. E-mail:


Aims:  This study was designed to evaluate potential effects of sampling duration on observed concentrations of airborne culturable mould and bacteria on selected media.

Methods and Results:  Airborne culturable mould and bacteria from lightly to moderately contaminated environments were collected on selected culture media using two co-located, concurrently operated, Andersen N-6 samplers for five sampling durations in the range of 1–10 min. Differences in mean concentrations, as well as linear relationships between sampling duration and both concentration and variability, were evaluated using nonparametric procedures. For the five sampling durations, there were no significant differences in mean concentrations of mould; for bacteria, there were significant differences, with a trend of decreasing concentrations as sampling duration increased. Data variability decreased with increasing sampling duration for both mould and bacteria.

Conclusions:  Airborne culturable mould concentrations were similar for sampling durations in the range of 1–10 min. Airborne bacteria concentrations tended to trend downwards with sampling durations exceeding 3 min.

Significance and Impact of the Study:  This study has shown that sampling durations of 1–10 min are appropriate for collection of airborne culturable mould on malt extract agar (MEA) and dichloran glycerol agar (DG-18); based on the apparent trend of decreasing bacterial sample concentrations associated with increasing sampling duration, sampling durations of ≤3 min may be more appropriate than 5 or 10 min for bacteria on trypticase soy agar (TSA).


Fungal spores and hyphal fragments as well as bacterial cells are commonly observed in aerosol samples collected in ambient (outdoor) and indoor environments (Burge 1995; Muilenberg 1995). Exposure to bioaerosol particles associated with common fungi (mould) and, to a lesser extent, bacteria in the indoor environments of residences, office buildings, and schools (Miller et al. 2000) has increasingly become a public health concern in North America. This has resulted from increased understanding of the potential relationship between respiratory symptoms (Brunkreef et al. 1989; Dales et al. 1991; Brunkreef 1992; Dales and Miller 1999), fungal sinusitis (Deshazo et al. 1997), and asthma (Strachan et al. 1990; Jaakkola et al. 1993,2002; Black et al. 2000; Douwes and Pearce 2003) and dampness factors such as water-damaged building materials, elevated relative humidity, and visible mould growth (Andersson et al. 1997; Institute of Medicine 2002). This public health concern has been heightened by the widely reported presence of toxigenic mould species such as Stachybotrys chartarum on water-damaged building materials (Etzel et al. 1998; Centers for Disease Control 2000).

Subsequently, a significant building owner/occupant demand has developed for bioaerosol (particularly mould) sampling and analytical services that are provided by private and public sector professionals. Culturable, spore trap, and quantitative polymerase chain reaction (QPCR) methods are commonly used for such sampling activities (Nevalainen et al. 1993; Dillon et al. 1996, 1999).

Culturable methods have the longest history of use. They are amenable to the collection of mould or bacteria on selected nutrient-containing media. Culturable mould and bacteria particles may be collected by impaction using multiple-hole sieve impactors [Anderson N-6, Anderson two- stage, and Surface Air System (SAS) samplers], centrifugal impactors (Reuter), or impingement in water with subsequent transfer to an agar medium (Cox and Wathes 1995; Dillon et al. 1996; Hung et al. 2005). On collection, media are incubated at room temperature (several days for bacteria and 7–10 days for mould). Colonies are counted and may be identified to the genus, and in the case of mould to the species, level.

Spore trap methods are used to sample airborne mould by collection on microscope slides or adhesive-covered coverslips with subsequent microscopic analysis (Aizenberg et al. 2000; Godish 2001; Hung et al. 2005). Both viable and nonviable mould species are enumerated. Although spores of some genera are easily identified, spores of Aspergillus and Penicillium cannot easily be identified to either genus or species. These methods are not suitable for the collection, enumeration, and identification of bacteria as spore trap samplers have a relatively large 50% cutoff diameter (c. 2·4 μm) (Aizenberg et al. 2000).

QPCR analyses of airborne mould are increasingly being used to identify and quantify mould species in airborne samples (Haugland et al. 2002; Hung et al. 2005). Although QPCR can be used to identify mould particles to the species level, its use is limited by analysis costs (several hundred dollars per sample) and target mould species selected.

Culturable bioaerosol sampling employing the Andersen N-6 sampler is often the method of choice for airborne mould and bacteria in indoor air quality investigations because of its long historical availability and inclusion in the NIOSH Manual of Analytical Methods (Cassinelli and O'Connor 1994). The NIOSH method recommends the use of 2% malt extract agar (MEA) and dichloran glycerol agar (DG-18) for airborne mould and trypticase soy agar (TSA) for airborne bacteria. A sampling duration of 10 min, with shorter durations for more heavily contaminated environments is recommended. Sampling durations using the Andersen N-6 sampler reported in the literature have varied from 1 to 20 min (Stancevich and Petersen 1990; Moschandreas and Storino 1996; Meklin et al. 1996; Neumeister-Kemp et al. 1999), with sampling durations of 2–5 min more commonly used (Dillon et al. 1996).

In a study conducted by NIOSH investigators Stancevich and Petersen (1990), an apparent relationship between increasing sampling duration and decreasing sample colony counts was reported. Because of the long (10 min) sampling duration recommended in the NIOSH bioaerosol sampling method (Cassinelli and O'Connor 1994) and the reported decrease in sample concentration with increasing sampling time reported by Stancevich and Petersen (1990), a need to further evaluate the potential effects of sampling duration on both airborne mould and bacteria was evident. The primary objective of this study was to systematically evaluate potential relationships between sample concentration and sampling duration over the range of 1–10 min. The null hypothesis tested was that there would not be any significant effect of sampling duration on culturable mould or bacteria sample concentrations, means, or variability.

Materials and methods

Potential effects of sampling duration on concentrations of airborne culturable mould and bacteria were evaluated by collecting air samples on three nutrient agar media. These were: 2% MEA [12·75 g maltose, 2·75 g dextrin, 2·36 g glycerol, 0·78 g peptone, and 15 g agar in 1 l of H2O (Difco Laboratories, Detroit, Michigan, USA); DG-18 [220 g glycerol, 10 g glycerol glucose, 5 g peptone, 1 g KH2PO4, 0·5 g MgSO4.7H2O, 0·002 g 2,6-dichloro-4-nitroaniline, and 15 g agar in 1 l of deionized H2O (Pitts and Hocking 1980)]; and TSA [15 g Bacto-tryptone, 5 g Bacto-soytone, 5 g NaCl, and 15 g agar in 1 l of H2O (Difco Laboratories)]. MEA is recommended for a variety of hydrophilic and mesophilic fungal species, DG-18 for xerophilic fungal species, and TSA for mesophilic bacteria (Dillon et al. 1996).

All sampling was conducted in four living spaces (rooms) in two occupied residences known to have minor mould infestation. Windows were closed a minimum of 24 h prior to sampling. Airborne samples were collected using two co-located N-6 Andersen samplers with calibrated sampling rates of 28·3 and 26·5 l min−1. Samplers were located in a central area of each room.

Each set of samples was comprised of three subsets, i.e. five time periods for each of the three sampling media. Each subset was collected concurrently on a given medium within a 14-min time period. For each sampling subset, one sampler was used to collect the 10-min sample while the other was operated so that the 1-, 2-, 3-, and 5-min collection times overlapped the 10-min sample by 60–100%.

Samplers were sterilized with 70% ethanol and air dried before initial sampling. After a complete subset of samples was collected, samplers were rotated to reduce potential bias, sterilized again by wiping thoroughly with 70% ethanol, and air dried.

Collected samples were incubated at room temperature (c. 22 °C for 7 days and counted at 40× magnification. Concentrations were calculated as colony forming units per cubic metre (CFU m−3) after positive-hole correction for possible multiple mould spore/particle impactions at the same hole in the multiple-hole sampling orifice plate (Macher 1989).

Sampling was conducted in two series. The first involved collection of sample sets of culturable airborne mould and bacteria in each of the three rooms (kitchen, dining room, and bedroom) in one residence on five different days between January and April, 2002; n = 15 complete sets of samples for this series.

A second sampling series was conducted between December 2002 and February 2003 to increase sample size. Sampling locations were two rooms in the original residence (dining room and bedroom) and a single room (living room) in a second residence. Sample sets were collected twice in each of the two rooms of the first residence on three different sampling days and three times on two different sampling days in the second residence; n = 18 sets of samples.

Sampling data were evaluated for normality by calculating skewness values. All data were evaluated using nonparametric statistical procedures. Differences among means of airborne culturable mould and bacteria for each sampling duration on each of the three media were evaluated using Friedman's test. Potential linear relationships between increasing sampling duration and both sample concentrations and standard deviations (SD) were evaluated by Spearman's rank correlation test. A P value of <0·05 was accepted as significant.


Evaluation of data from two sampling series

Pooled data (n = 33 sets) from the two sampling series evaluated by descriptive statistical procedures are summarized in Table 1. Mean and median culturable moulds were relatively low. Bacteria concentrations were notably higher than mould concentrations, with a less pronounced difference between mean and median values. All minimum, maximum, and SD values indicated considerable data variability. Minimum culturable mould values were below the limit of detection (BLD) in 5% of the samples collected, with the BLD value as well as the number of BLD values observed decreasing with increasing sampling duration. Unlike mould, only one airborne bacterial sample was BDL.

Table 1.   Summary statistics for culturable mould and bacteria on selected sampling media for five sampling durations (n = 33 sampling sets)
Sampling mediumSampling duration (min)Mean (CFU m−3)Median (CFU m−3)Minimum (CFU m−3)Maximum (CFU m−3)SD (CFU m−3)Number of samples BDL
  1. *Below limit of detection (BDL) = 35–38.

  2. †BDL = 18–19.

  3. ‡BDL = 12–12·6

  4. §BDL = 7·0–7·5.

  5. **BDL = 3·5–3·8.

  6. Abbreviations: MEA, malt extract agar; DG-18, dichloran glycerol agar; TSA, trypticase soy agar.


Data for the two sampling series were also evaluated using descriptive statistical procedures to determine whether they were from similar sampling populations. These analyses indicated that median culturable mould values in the first sampling series for MEA and DG-18 (128 and 151 CFU m−3, respectively) were higher than in the second sampling series (38 and 69 CFU m−3, respectively), suggesting they may have been from different airborne mould concentration populations. For bacteria, however, median values were relatively similar (318 vs 297 CFU m−3), indicating no apparent differences in concentration values from the two sampling series.

Comparisons of bioaerosol concentrations in samples collected for different sampling durations


Summary statistics for all sampling data are presented in Table 1. These data were not normally distributed, as indicated by the relatively large differences between mean and median concentrations and confirmed by skewness values (c. 3·7–4·7 for MEA and 3·9–4·4 for DG-18). On application of Friedman's nonparametric test, no significant differences were observed among means of airborne culturable for each of the five sampling durations mould (MEA, P = 0·94 ; DG-18, P = 0·98).

No linear relationship was observed between sample concentrations and sampling duration when these data were evaluated using Spearman's rank correlation test. However, significant negative trends were observed for data variability (as indicated by SD) (Table 2).

Table 2.   Spearman's rank correlation analyses of relationships between sampling duration and concentrations and standard deviations of airborne culturable mould and bacteria on selected sampling media
Sampling mediumSample concentration (n = 165) SD (n = 5)
  1. *Significant at P < 0·05.

  2. †Significant at P < 0·01.

  3. Abbreviations: MEA, malt extract agar; DG-18, dichloran glycerol agar; TSA, trypticase soy agar.



Summary statistics for airborne culturable bacteria on TSA are also presented in Table 1. Skewness values, which varied from 1·1 to 1·9, indicated that these data were not normally distributed. Application of Friedman's test indicated that the mean concentration of the 5-min samples was significantly different from those for sampling durations of 1, 2, and 3 min (P = 0·002). Although trending in a decreasing direction, the mean concentration for the 10-min sampling duration was not significantly different (P = 0·08) from means for the first three sampling durations. Data in Table 2 indicate a weak, nearly significant, negative relationship between sampling duration and sample concentration and a strong negative relationship between sampling duration and data variability.


Stancevich and Petersen (1990), in their study of potential effects of sampling duration on culturable mould concentrations on MEA, reported that mean concentrations of airborne mould decreased significantly with increased sampling duration; 1-min samples had nearly twice the mean concentration of 3- and 5-min samples. In contrast, no significant differences in mean airborne mould concentrations on MEA or DG-18 were observed for sampling durations in the range of 1–10 min used in this study. These results are consistent with sampling durations (≤10 min) recommended in the NIOSH bioaerosol method (Cassinelli and O'Conner 1994). Results of this study for culturable airborne bacteria indicate that increasing sampling duration beyond 3 min has the potential to significantly decrease count concentrations, and thus accuracy. These findings suggest a need for further investigation and review of NIOSH guidelines for sampling culturable airborne bacteria.

Decreases in data variability associated with increasing sampling duration (and thus volume of air collected) observed in this study and reported by Stancevich and Petersen (1990) are consistent with increases in averaging time and decreases in the number of samples that were BDL. In the latter case, this was primarily because of the decrease in the limit of detection from 35 CFU m−3 for a 1-min sampling duration to 3·5 CFU m−3 for 10 min.

Although the current study found no decrease in culturable airborne mould concentrations associated with increased sampling duration, this was not the case with airborne culturable bacteria. Reasons for this difference are not known. However, Dillon et al. (1996) have suggested that the significantly lower culturable mould concentrations reported by Verhoeff et al. (1990) for air samples collected using SAS samplers as compared with those collected using Andersen N-6 samplers may have been because of reduced bioaerosol particle viability caused by desiccation. SAS samplers use very high sample volumes (450–900 l) as compared with the much lower sample volumes (142 l) for the Andersen N-6. If reduced bioaerosol viability is associated with high sample volume collections, then results of our studies appear to suggest that viable airborne bacteria may be more sensitive to such effects than viable mould spores/particles.