The outdoor environment near the building was highly vegetated and consisted of a tree-lined street with lawn and flower gardens. There were no green plants in the room, nor were they common throughout the building. During the occupied and vacant experiments, all windows and doors were closed and conditioned air was delivered by the building HVAC system to the room through a 0.9 · 0.05 m air register situated above the door. Ventilation exhaust ports were located along the floor and near the wall opposite to the ventilation air inlet. Based on six deliberate CO2 releases and tracer gas studies performed prior to air sampling and under vacant, well-mixed conditions, the average air-exchange rate ± s.d. was 5.5 ± 1.3/h. AER was also calculated immediately after occupancy based on the decay of exhaled CO2 from occupants. This AER averaged 6.2 ± 0.9/h, which is not statistically different than the AER derived from tracer gas studies. During the four occupied and four vacant experimental days, the temperature was 23.5 ± 1.1 C indoors and 13.4 ± 2.4C outdoors, and the average relative humidity was 28 ± 7%indoors and 45 ± 16% outdoors. The reported air exchange rate is in the upper range for ventilated commercial buildings . The average occupant density of 4.7 persons in 30 m2 is within ranges common for many indoor spaces with moderately dense occupancy .Figure 1 shows the size distributions of indoor and outdoor particle mass and the bacterial and fungal genome copies obtained from the impactor measurements under occupied and vacant conditions. These data reveal important insights regarding concentrations and size distributions that result from room occupancy. First,plant grow racks human occupancy was associated with substantially increased airborne concentrations of total particles, bacterial genomes, and equivalent fungal genomes.
For each measured size range, the indoor concentrations are higher when the room is occupied compared with being vacant . Averaged across all size ranges, the indoor air occupied-to-vacant ratios of particulate matter mass concentrations, bacterial genome concentrations, and equivalent fungal genome concentrations were 15, 170, and 2.3, respectively. Only minor differences were observed between the outdoor levels when comparing occupied and vacant sampling days, which indicates that the increases in the indoor concentrations during occupancy were not caused by concomitant changes in outdoor levels. In addition, during vacant conditions, indoor size-distributed concentrations of particulate matter, bacterial genome, and equivalent fungal genome concentrations were qualitatively similar to those measured outdoors . In support of these observations from filter-based sampling, the real-time optical particle counter data reinforce the finding that occupancy is a major contributor to super micron particle concentrations. Comparison of Figures S1–S4 to Figures S5–S8 in the online supporting information demonstrates the substantial increases in super micron particle concentrations when the room is occupied. The estimated mass-based particle size distributions obtained from analysis of the OPC data reveal a profile that is similar to the filter-based measurements with respect to total concentration and the mass distribution being weighed toward the largest particles. Examining the particle size distributions, Figure 1 shows that the increase in total particle mass during occupancy is dominated by the largest measured super micron particles, that is those greater than 9 lm in aerodynamic diameter. The distribution of fungal genome copies was somewhat different than that of total particulate matter, with fungal peaks during occupancy near the typically cited aerodynamic diameters of unicellular and multicellular fungal spores . The largest multiplicative increase in airborne equivalent fungal genome concentrations associated with occupancy was observed in the largest stage. In contrast, for bacterial genomes, a strong peak during occupancy was observed on the stage corresponding to the 3–5 lm aerodynamic diameter range. Figure 2 plots the ratio of estimated bacterial mass to particulate matter mass for all the measured size ranges.
The results demonstrate that indoor airborne particulate matter during occupancy was enriched in bacteria, especially in the 3–5 lm range, and that the contribution of bacteria to particulate matter mass was about 0.5% in this size range. Overall, bacteria contributed approximately 0.1% to indoor airborne particle mass during occupancy. In these calculations, bacterial mass was estimated assuming the weight of a bacterium to be 655 femtograms . An analogous mass estimate was not conducted for fungi owing to the known variability of 18S rDNA gene copy numbers per genome among different fungal species and the large variability in the size of unicellular and multicellular spores.The material balance model applied to determine emission rates assumes that within a specific size range, the concentration of a species in indoor air can be estimated as a fraction of the level in outdoor air, plus the contribution from indoor emissions. The contribution from the indoor source is estimated from a balance between the emission rate and the rates of removal . To determine the size-specific infiltration factor, f, we analyzed the indoor–outdoor ratio for each size range during vacant conditions and assumed that the same ratios apply for each respective size in the occupied case. Also, during spans of time on the occupied monitoring days when the classroom was vacant , relatively consistent relationships were observed between the indoor and outdoor particle number concentrations for finer particles. During these same unoccupied intervals, the OPC data also demonstrate that indoor concentrations of super micron particles are small . The estimated, per person, size-resolved emission rates of particulate matter mass, bacterial genomes, and equivalent fungal genomes attributable to human occupancy are presented in Figure 3. The size-distributed emission rates are similar in shape to the time averaged concentrations during occupied periods. The majority of emissions for total mass are associated with particles larger than 9 lm, whereas bacterial genome emissions are predominantly in the 3–5 lm aerodynamic diameter range. It is important to acknowledge that these rates represent aggregate emissions associated with human occupancy, potentially including contributions from resuspension from the carpeted floor and from other surfaces plus direct shedding of microorganisms and particles from humans.
The nature of these experiments does not allow us to apportion occupancy associated emissions into their separate components for total particulate matter. However, phylogenetic analysis does provide some clues as to the origin of airborne bacterial genomes, as described next.The results of this study highlight two important concepts for understanding the sources of microbial aerosols in occupied indoor environments. First, human occupancy results in significant emissions of airborne particulate matter mass, bacterial genomes, and fungal equivalent genomes. Second, during occupancy, the bacterial phylogenetic analysis demonstrates a distinct indoor air signature of bacteria with associations to human skin, hair, and nostrils.Emissions of airborne bacteria and fungi in indoor environments have not been well characterized. However, studies based on surrogate measures for microorganisms in multiple residences, as well as chamber and full-scale room resuspension rate estimates for allergens, all indicate that biological particles can be resuspended or directly shed during human occupancy . In the present study, the characterization of particle size distributions, the estimates of per person emissions rates, and the phylogenetic results revealed new insights into the major sources of fungal and bacterial indoor aerosols. The total mass emitted during occupancy was predominantly associated with large particles . This size-dependent increase is broadly consistent with previously observed increases in resuspension rate with increased particulate matter size . In the case of fungi,sliding grow racks the largest increase in concentration during occupancy was also observed on the largest impactor stage . Unicellular and multicellular fungal spores are commonly present in indoor environments and thus provide for a potentially broad size distribution of fungal genomes. Multicellular fungal spores such as Alternaria spp. and Epicoccum spp. provide a source of larger fungal particles that can be preferentially resuspended during human activity. Under unoccupied conditions, one expects that these larger spores would be removed effectively by the filters in the building ventilation system, would penetrate through air leakage paths inefficiently, and would settle rapidly onto indoor surfaces compared to the smaller 2– 5 lm unicellular fungi, such as Aspergillus spp. or Cladosporium spp., that are common indoors . Relative to bacteria, the more modest overall increase in equivalent fungal genomes attributable to occupancy, especially below 9 lm aerodynamic diameter, is consistent with a recent study that demonstrated the roughly equivalent importance of both outdoor air concentrations and resuspension on influencing indoor air concentrations of Aspergillus spores . Bacterial genome emissions, in contrast, did not follow size distributions that would be expected from purely physical resuspension processes. While human occupancy led to a large total increase in bacterial genomes, the size distribution for bacterial genomes was not skewed toward the largest sizes, and increases in the largest impactor stage were substantially less than the average bacterial genome increase . The dominant size ranges for bacterial genomes were on the stage corresponding to 3.3–4.7 lm aerodynamic diameter. This size is larger than pure culture, single bacteria that have been previously screened for their aerodynamic diameter: Serratia marcescens 2.8 lm ; Bacillus subtilus 2.7 lm ; Mycobacterium parafortuitum 2.8 lm ; Pseudomonas fluorescens 0.8 lm . Thus, we hypothesize that the organisms are commonly attached to each other or to other small biotic or abiotic particles. The wide dispersion of single taxa across multiple size ranges also suggests that bacteria were mostly not present in particles as single cells. That indoor airborne bacteria may be found in aggregates has been previously reported by Lidwell et al. . The size distribution observed in this study is consistent with recent outdoor measurements of bioaerosol particle concentration determined by an ultraviolet aerodynamic particle sizer that demonstrated a pronounced peak at 3.2 lm geometric mean diameter, which the authors attributed to unicellular fungal spores and agglomerated bacteria .
The size distributions of bacterial genomes and fungal genomes clearly demonstrate that particulate matter mass is not a good surrogate for biological particles. During occupancy, the majority of total particle mass is larger than 9 lm in aerodynamic diameter; however, the dominant bacterial genome concentrations were between 3 and 5 lm, and the fungal genomes were broadly distributed. Other researchers have shown a lack of association between fungal spores and particulate matter mass . Focusing on bacteria in the 3- to 5-lm size range, such particles deposit relatively slowly to indoor surfaces. They would be removed efficiently by air filters common in mechanical ventilation systems . However, indoor emissions could cause indoor exposures whenever occupants share indoor spaces. When inhaled, 3- to 5-lm particles have a high deposition efficiency in the human respiratory tract with substantial penetration to and deposition in the pulmonary region of the lung . The qPCR and phylogenetic-based analysis of particle size distributions for airborne fungi and bacteria is novel. A major advantage of these DNA-based methods is the complete responsiveness to bacterial and fungal organisms, regardless of the viability or culturability of the cells. Limitations in detecting environmental bacteria and fungi through culture based methods are well documented , as is the inactivation of bacteria during impaction or desiccation in impactors that are commonly used to provide sizeresolved information . Another important advantage of using DNA-based methods is the fact that agglomerated organisms are accurately counted, rather than determined as one colony-forming unit. However, when using universal primers for fungi, gene copies cannot be reliably converted to whole fungal spore counts owing to the variable number of 18S rDNA gene copies for different fungal species and the lack of consensus on the average operon gene copy number among all fungi. Moreover,the PCTE filters used were not coated to ensure successful cell extraction and downstream enzymatic based PCR analysis.However, the qualitatively good agreement between the OPC results and the impactor-based particle mass measurements suggests that any such bias, if present, was not strong.A major contribution of the phylogenetic work reported here is to provide quantitative emission rates of bacteria that originate from human skin, hair, nostrils, and the oral cavity . Such analysis demonstrates that the bacteria shed from humans can comprise a significant portion of the airborne bacteria to which people are exposed in common occupied spaces. Public health stresses the importance of human hygiene and specifically hand washing to prevent transmission of disease.