Wildland fire as an atmospheric source of viable microbial aerosols and biological ice nucleating particles

Abstract

The environmental sources of microbial aerosols and processes by which they are emitted into the atmosphere are not well characterized. In this study we analyzed microbial cells and biological ice nucleating particles (INPs) in smoke emitted from eight prescribed wildland fires in North Florida. When compared to air sampled prior to ignition, samples of the air-smoke mixtures contained fivefold higher concentrations of microbial cells (6.7 ± 1.3 × 10⁴ cells m^-3) and biological INPs (2.4 ± 0.91 × 10³ INPs m^-3 active at temperatures ≥ -15 °C), and these data significantly positively correlated with PM₁₀. Various bacteria could be cultured from the smoke samples, and the nearest neighbors of many of the isolates are plant epi- and endophytes, suggesting vegetation was a source. Controlled laboratory combustion experiments indicated that smoke emitted from dead vegetation contained significantly higher numbers of cells, INPs, and culturable bacteria relative to the green shrubs tested. Microbial viability of smoke aerosols based on formazan production and epifluorescent microscopy revealed no significant difference in the viable fraction (~80%) when compared to samples of ambient air. From these data, we estimate each fire aerosolized an average of 7 ± 4 × 10⁹ cells and 2 ± 1 × 10⁸ biological INPs per m² burned and conclude that emissions from wildland fire are sources of viable microbial aerosols to the atmosphere.

Fig. 1. Cell and CFU concentrations in the sampled smoke–air mixtures.

Total DNA-containing cells (a) and cell concentration per mass of PM₁₀ (b) are shown for samples collected prior to ignition (i.e., ambient) and during the prescribed burns (n = 8). The bottom panels contain the concentrations of cells (c) and CFUs (d) per mass of PM₁₀ in smoke from the laboratory burns of litter and duff (Grass/Soil O horizon), runner oak, or saw palmetto. Boxes represent the interquartile range, the middle horizontal line is the median, and the whiskers represent minimum and maximum values. Outliers are 1.5 interquartile ranges below the first quartile and above the third quartile. Asterisks indicate level of significance as follows: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; not significant (ns) p > 0.05.

Fig. 2. Concentration of INP aerosols measured in smoke–air mixtures.

a Cumulative concentrations of total and biological INPs sampled in air prior to and during each prescribed burn (n = 6). b Cumulative concentrations of total and biological INPs per mass of PM₁₀ sampled in smoke from laboratory experiments that burned senesced grass and soil O horizon (GS), runner oak (RO), and saw palmetto (SP); n = 3.

Fig. 3. Viability of the microbial aerosols sampled in smoke and ambient air from units H-8a/b (n = 3).

The CTC data represent inferred initial concentrations of viable cells based on the number of formazan-accumulating cells after for 24 h and the assumption of exponential growth, as described in the text. DAPI (n = 3, burn units H-8a/b) and SYBR Gold (n = 3, burn units H-8a/b) data represent total cell concentrations. Stars depict significance level between corresponding measurements.

Fig. 4. Phylogenetic analysis of 16S rRNA genes sequences (positions 40–1340, Escherichia coli numbering) amplified from pure cultures of bacteria isolated from smoke and ambient aerosol samples.

Triangle and square symbols indicate type strains that were originally isolated from phyllosphere, plant tissue, or root nodule sources, Bootstrap values are shown as a percentage of 1000 replicates. The phylogenetic tree is based on maximum likelihood and the scale bar represents 0.03 substitutions per site. GenBank accession numbers are shown in parentheses.