Indoor airborne bacterial communities are influenced by ventilation, occupancy, and outdoor air source

Abstract

Architects and engineers are beginning to consider a new dimension of indoor air: the structure and composition of airborne microbial communities. A first step in this emerging field is to understand the forces that shape the diversity of bioaerosols across space and time within the built environment. In an effort to elucidate the relative influences of three likely drivers of indoor bioaerosol diversity - variation in outdoor bioaerosols, ventilation strategy, and occupancy load - we conducted an intensive temporal study of indoor airborne bacterial communities in a high-traffic university building with a hybrid HVAC (mechanically and naturally ventilated) system. Indoor air communities closely tracked outdoor air communities, but human-associated bacterial genera were more than twice as abundant in indoor air compared with outdoor air. Ventilation had a demonstrated effect on indoor airborne bacterial community composition; changes in outdoor air communities were detected inside following a time lag associated with differing ventilation strategies relevant to modern building design. Our results indicate that both occupancy patterns and ventilation strategies are important for understanding airborne microbial community dynamics in the built environment.

Keywords: Airborne bacterial community; Bioaerosol; Built environment; Indoor microbial ecology; Natural ventilation.

Fig 1

Indoor air phylogenetic diversity closely resembles outdoor air. Faith's phylogenetic diversity (PD) values for all samples (900 sequences per sample) are overlaid throughout the week. Outdoor air experienced temporal community changes throughout the sampling period, and this outdoor influence is experienced inside of rooms regardless of ventilation strategy

Fig 2

Bacterial communities in unoccupied rooms were more similar to concurrent outdoor air than occupied rooms. When all indoor samples are compared with outdoor air samples taken during the same time period, occupied rooms were more dissimilar (P = 0.018). Marginal tick marks show the distribution of dissimilarity values in either group. Box plots delineate (from bottom) minimum value, Q1, median (Q2), Q3, and maximum value; notches approximate 95% confidence around median values

Fig 3

Unventilated rooms retain legacy of outdoor air communities. Gray points on the bottom panel show the average outdoor community dissimilarity (±1 s.d.) from each initial outdoor sample at the beginning of the sampling period shown (Thursday 12 a.m.–8 a.m.); blue and red points represent fully ventilated (Night-flush) and conventionally ventilated (No Night-flush) classrooms, respectively. Blue and red bars above the points show ventilation regime, with night-flush (blue) vents open throughout the sampling period, and non-night-flushed (red) vents only open during occupied hours. Arrows show hourly wind direction and speed (length) during the sampling period

Fig 4

Twenty-five most abundant OTUs exhibit the temporal community changes detected in Figure 3. Relative abundances of the 25 most common bacterial taxa (mean = 42.6% of sequences ± 8.5 s.d.) shift at different rates based on ventilation regime. The height of each taxon band is equivalent to the relative abundances of each taxon, and they are colored by whether they were more abundant during the beginning of the sampling period or at the end. OTU names are from BLAST identification to the genus level, and all identifications shown were ≥99% similar to reference NCBI sequences. Red bars indicate the status of ventilation dampers in non-night-flushed rooms; dampers in night-flushed rooms were open throughout the sampling period shown