Season, Vegetation Proximity and Building Age Shape the Indoor Fungal Communities' Composition at City-Scale

Abstract

Exposure to particular microbiome compositions in the built environment can affect human health and well-being. Identifying the drivers of these indoor microbial assemblages is key to controlling the microbiota of the built environment. In the present study, we used culture and metabarcoding of the fungal Internal Transcribed Spacer ribosomal RNA region to assess whether small-scale variation in the built environment influences the diversity, composition and structure of indoor air fungal communities between a heating and an unheated season. Passive dust collectors were used to collect airborne fungi from 259 dwellings representative of three major building periods and five building environments in one city-Lausanne (Vaud, Switzerland)-over a heating and an unheated period. A homogenous population (one or two people with an average age of 75 years) inhabited the households. Geographic information systems were used to assess detailed site characteristics (altitude, proximity to forest, fields and parks, proximity to the lake, and density of buildings and roads) for each building. Our analysis indicated that season was the factor that explained most of the variation in colonies forming unit (CFU) concentration and indoor mycobiome composition, followed by the period of building construction. Fungal assemblages were more diverse during the heating season than during the unheated season. Buildings with effective insulation had distinct mycobiome compositions from those built before 1975 - regardless of whether they were constructed with pre-1945 technology and materials or 1945 - 1974 ones. The urban landscape-as a whole-was a significant predictor of cultivable Penicillium load-the closer the building was to the lake, the higher the Penicillium load-but not of fungal community composition. Nevertheless, the relative abundance of eleven fungal taxa detected by metabarcoding decreased significantly with the urbanization gradient. When urban landscape descriptors were analyzed separately, the explanatory power of proximity to vegetation in shaping fungal assemblages become significant, indicating that land cover type had an influence on fungal community structure that was obscured by the effects of building age and sampling season. In conclusion, indoor mycobiomes are strongly modulated by season, and their assemblages are shaped by the effectiveness of building insulation, but are weakly influenced by the urban landscape.

Keywords: built environment; culture; indoor air quality; metabarcoding; urban mycobiome.

Figure 1

Geographical location of the buildings for which the mycobiome was characterized, in black for those for which molecular and cultivable data were obtained, and in red for those for which only the culture provided results. Their belonging to an environmental typology is indicated by a triangle for peri-urban cluster, by a circle for the suburban cluster, by a plus for the historic downtown cluster, by a diamond for the lakefront downtown, and by a square for the lakefront. Elevation lines every 100 m and 20 m are indicated with thick and thin brown lines, respectively.

Figure 2

Visualization of the similarity in genus composition of fungal communities across the studied buildings. Each point represents the fungal community in a building, such that those that are closer together share more identified genera in common than those farther apart. Fungal community composition and relative abundance of genera tend to cluster according to season (a), period of building construction (b), or proximity to green space in the surroundings (c).

Figure 3

Boxplots showing distribution in relative abundance of the nine most abundant and frequent genera across samples. The values observed during the heating period are shown in red, and those during the unheated period are shown in blue. The whiskers in a box-and-whisker plot are the adjacent values that correspond to the highest value not greater than p75 + 3/2 IQR and the lowest value not less than p25 − 3/2 IQR, where IQR is the inter-quartile range, the box covers the values between the first and third quartile, and the line in the box marks the median value.

Figure 4

Visualization of the proportion (a) and log transformed mean relative abundance (b) of indoor fungal genera that differed significantly between the unheated and heating periods (p < 0.05). Values observed during the heating period are shown in red, and those observed during the unheated period are in blue.

Figure 4

Visualization of the proportion (a) and log transformed mean relative abundance (b) of indoor fungal genera that differed significantly between the unheated and heating periods (p < 0.05). Values observed during the heating period are shown in red, and those observed during the unheated period are in blue.

Figure 5

Visualization of the proportion (a) and log transformed mean relative abundance (b) of indoor fungal genera that differed significantly between buildings constructed at different periods (p < 0.05). Shown in grey are taxa observed in buildings constructed before 1944, in red are those observed in buildings constructed between 1945 and 1974, and in yellow are taxa observed in buildings constructed between 1975 and 2014.

Figure 6

Visualization of the proportion (a) and mean log transformed relative abundance (c,d) of indoor fungal genera that differed significantly between buildings with and without green space in the surroundings (p < 0.05)—(c) during the unheated period (d) during the heating period. The distribution of relative abundance among samples is illustrated for few taxa in (b). In green, values for buildings with green space in the surroundings, and in black, values for buildings without green space in the surroundings.

Figure 6

Visualization of the proportion (a) and mean log transformed relative abundance (c,d) of indoor fungal genera that differed significantly between buildings with and without green space in the surroundings (p < 0.05)—(c) during the unheated period (d) during the heating period. The distribution of relative abundance among samples is illustrated for few taxa in (b). In green, values for buildings with green space in the surroundings, and in black, values for buildings without green space in the surroundings.