The indoor mycobiome of daycare centers is affected by occupancy and climate

Many children spend considerable time in daycare centers and may here be influenced by indoor microorganisms, including fungi. In this study, we investigate the indoor mycobiome of 125 daycare centers distributed along strong environmental gradients throughout Norway. Dust samples were collected from doorframes outside and inside buildings using a citizen science sampling approach. Fungal communities in the dust samples were analyzed using DNA metabarcoding of the ITS2 region. We observed a marked difference between the outdoor and indoor mycobiomes. The indoor mycobiome included considerably more yeasts and molds compared to the outdoor samples, with Saccharomyces, Mucor, Malassezia and Penicillium among the most dominant fungal genera. Changes in the indoor fungal richness and composition correlated to numerous variables related to both outdoor and indoor conditions; there was a clear geographic structure in the indoor mycobiome composition that mirrored the outdoor climate, ranging from humid areas in western Norway to drier and colder areas in eastern Norway. Moreover, the number of children in the daycare centers, as well as various building features, influenced the indoor mycobiome composition. We conclude that the indoor mycobiome in Norwegian daycare centers is structured by multiple factors and is dominated by yeasts and molds. This study exemplifies how citizen science sampling enables DNA-based analyses of a high number of samples covering wide geographic areas. Importance With an alarming increase in chronic diseases like childhood asthma and allergies, there is an increased focus on the exposure of young children to indoor biological and chemical air pollutants. Our study of 125 daycares throughout Norway demonstrates that the indoor mycobiome not only reflects co-occurring outdoor fungi but includes a high abundance of yeast and mold fungi with an affinity for indoor environments. A multitude of factors influence the indoor mycobiome in daycares, including building type, inhabitants, as well as the outdoor environment. Many of the detected yeasts and molds are likely associated with the human body, where some have been coupled to allergies and respiratory problems. Our results call for further studies investigating the potential impact of the identified daycare-associated mycobiomes on children health.

In many countries, children spend considerable time in daycare centers, where they are exposed to 58 indoors microorganisms, including fungi. Since young children often vector organic material such 59 as soil and litter from nature, daycare centers may accumulate extra organic substrates promoting 60 fungal growth, as compared to other indoor environments. In line with this, it has previously been 61 shown that the concentration of fungi in daycare centers is higher compared to homes (9). In several studies, the outdoor environment has been reported as the main source of indoor fungi (10-63 13) due to the influx of spores through windows, entrances and the ventilation system. Hence, the 64 vegetation and climate that structure the outdoor fungi will indirectly also structure the indoor 65 mycobiome (11). In correspondence with this, in a recent DNA metabarcoding study performed in 66 271 private homes across Norway, we showed that outdoor climate was one of the main drivers of 67 the indoor dust mycobiome (13). A similar observation was done by Barberán   body and may therefore be prevalent indoors (16)(17)(18)(19). Which fungi that are associated with the 74 human body may, to some extent, be age-dependent. For instance, the basidiomycete yeast 75 Malassezia seems particularly prevalent on adults (20), while children tend to have a more diverse 76 skin-associated mycobiome, including genera like Aspergillus, Epicoccum, Cladosporium, 77 Cryptococcus and Phoma, in addition to Malassezia (18).

92
The impact long-term exposure to indoor fungi can have on human health highlights the need to 94 better characterize the indoor mycobiome from an early age. In this study, we aim to analyze the 95 indoor mycobiome associated with daycare centers. We ask (1) which outdoor and indoor factors 96 drive the daycare mycobiome, and (2) which fungal groups dominate in the daycare centers, as 97 compared to outdoor samples. To address these research questions, we chose a citizen science 98 approach, where daycare personnel collected dust samples according to our instructions. We   (Table 1). In addition, information about the local climate and vegetation were extracted based on 116 the geographic coordinates of the daycare centers (29). Considered individually, numerous of these 117 variables correlated significantly with the compositional variation in the indoor mycobiome ( Fig.   118 2c), including variables related to the daycare centers such as daycare type, construction year, 119 number of departments, pests and building type. Climatic variables such as temperature and total 120 insolation were also significantly correlated to the indoor mycobiome composition, as well as 121 spatial variables that likely mirror additional regional environmental variability (Fig. 2c, d). Many 122 of the inferred variables were associated with the major climate gradient stretching from humid, 123 oceanic areas in western Norway, to inland, continental areas in eastern Norway (Fig. 2c, d).
that longitude (mirroring the regional climate gradient), presence of pest/rodents, construction year 126 of the daycare center and number of children were the main drivers of the fungal community 127 composition, with very low interaction effects (<0.01%). However, these factors accounted 128 altogether for only 7% of the variation in mycobiome composition ( Table 2). The indoor fungal 129 richness, calculated on a sample-basis, was significantly higher in the bathroom compared to the 130 main room, and there was a significant positive correlation between indoor fungal richness and the 131 maximum outside temperature during May at the sampling location, as well as the proximity to 132 coast (see the Mixed Effect Model presented in Table 3).  Interestingly, in our recent study on seasonality of the indoor mycobiome, the indoor environment composition. In addition to building variables, regional climate related factors such as maximum 183 temperature in June, mean temperature of the coldest quarter and total insolation, also correlated 184 significantly with the indoor mycobiome composition. Longitude, an approximation for regional climate variability, also had explanatory power. Throughout most of Norway, longitude mirrors a climate gradient from oceanic and humid areas in the west, to areas with dryer, colder and high 187 temperature seasonality conditions in the east. Our findings mirror the observations by Barberán

248
In the current study, we carried out a citizen science sampling approach for obtaining the material 249 from daycare centers. In line with our previous study (13), only a few outlier samples occurred, 250 and the indoor and outdoor dust samples were largely separated in species composition, indicating 251 a low influence of sampling bias. Moreover, very few samples were discarded due to low DNA 252 yields. Our results supported that citizen science sampling is a powerful approach to obtain 253 samples from a widespread geographic area during a short time span. We advocate for further 254 citizen science studies for evaluating biological and chemical air pollutants, which will also help 255 to raise public awareness on air quality problems in buildings.     (Table 1).

333
Annotation of fungal (OTUs) growth characteristics 334 We annotated the 1 593 most abundant OTUs, defined as those with >20 sequences and taxonomic   To investigate OTU-richness trends, a linear mixed effect model was applied using the NLME 376 package (52), including daycare ID as a random contribution. Collinear variables were excluded 377 as described above (|r| > 0.6), however, to further avoid multicollinearity in the mixed effect model     Goodness-of-fit statistics (r2) for variables that significantly (p<0.05) account for variation in the 576 composition of the indoor mycobiome. Variables related to regional climate are listed in the upper part of 577 the table, while variables related to the specific daycares are listed in the lower part. The category saprotrophs represents litter and wood decay fungi. 584 The black lines indicate standard error. 591