Portal de Periódicos da CAPES

The microbial signature of aerosols produced during the thermophilic phase of composting

2009; Oxford University Press; Volume: 108; Issue: 1 Linguagem: Inglês

10.1111/j.1365-2672.2009.04427.x

1365-2672

Olivier Le Goff, Valérie Bru‐Adan, H. Bacheley, Jean‐Jacques Godon, Nathalie Wéry,

Odor and Emission Control Technologies

Journal of Applied MicrobiologyVolume 108, Issue 1 p. 325-340 Free Access The microbial signature of aerosols produced during the thermophilic phase of composting O. Le Goff, O. Le Goff INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorV. Bru-Adan, V. Bru-Adan INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorH. Bacheley, H. Bacheley Veolia Environnement Recherche et Développement, 291 avenue Dreyfous Ducas, Limay, FranceSearch for more papers by this authorJ.-J. Godon, J.-J. Godon INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorN. Wéry, N. Wéry INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this author O. Le Goff, O. Le Goff INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorV. Bru-Adan, V. Bru-Adan INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorH. Bacheley, H. Bacheley Veolia Environnement Recherche et Développement, 291 avenue Dreyfous Ducas, Limay, FranceSearch for more papers by this authorJ.-J. Godon, J.-J. Godon INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this authorN. Wéry, N. Wéry INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, FranceSearch for more papers by this author First published: 10 December 2009 https://doi.org/10.1111/j.1365-2672.2009.04427.xCitations: 54 Nathalie Wéry, INRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne F-11100, France. E-mail : weryn@supagro.inra.fr AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Aims: The microbial diversity of bioaerosols released during operational activities at composting plants is poorly understood. Identification of bacteria and fungi present in such aerosols is the prerequisite for the definition of microbial indicators that could be used in dispersal and exposure studies. Methods and Results: A culture-independent analysis of composting bioaerosols collected at five different industrial open sites during the turning of composting piles in fermentation was performed by building 16S rDNA and 18S rDNA libraries. More than 800 sequences were analysed. Although differences in the phylotypes distribution were observed from one composting site to another, similarities in the structure of microbial diversity were remarkable. The same phyla dominated in the five bioaerosols: Ascomycota among fungi, Firmicutes and Actinobacteria among bacteria. For each phylum, some dominant phylotypes were common to at least four bioaerosols. These common phylotypes belonged to Thermomyces, Aspergillus, Penicillium, Geobacillus, Planifilum, Thermoactinomyces, Saccharopolyspora, Thermobifida and Saccharomonospora. Conclusions: The microbial signature of aerosols produced during the thermophilic phase of composting was determined. The similarities observed may be explained by the selection of thermophilic and sporulating species. Significance and Impact of the Study: Several bacteria and fungi identified in this study may represent potential indicators of composting bioaerosols in air. Introduction Composting is a waste management method based on the biological degradation and stabilization of organic matter under aerobic conditions. It results in a sanitized and stabilized product rich in humic substances that can be beneficially applied to the land (Sykes et al. 2007). The degradation of organic matter is carried out by a complex and highly dynamic microflora containing Gram-positive and Gram-negative bacteria, as well as fungi (Ishii et al. 2000; Ryckeboer et al. 2003; Hansgate et al. 2005). Normal operations taking place at composting plants can be the source of nuisance or pollution involving odors, noise, dust, leachate and bioaerosols (Sanchez-Monedero et al. 2005). Bioaerosols are composed of inert particles and adsorbed live micro-organisms (algae, bacteria, fungi, yeasts) as well as microbial components such as mycotoxins, glucans and endotoxins (Fracchia et al. 2006; Peccia and Hernandez 2006). Bioaerosols generated at composting plants are released during processes that involve vigorous movement of material: fresh waste delivery, shredding, compost pile turning and compost screening. The release of bioaerosols is especially significant for composting plants operating in the open because their bioaerosols are released directly into the surrounding area without any pretreatment such as biofilters or bioscrubbers (Sanchez-Monedero et al. 2005). Open-air windrow systems still remain the predominant composting technology. The emission of bioaerosols from such composting plants considerably increases the concentration of micro-organisms in the air. Shredding of fresh green waste, pile turning and screening of mature compost can lead to an increase on site by two logarithmic units (Sanchez-Monedero et al. 2005). Bioaerosols released from composting plants are a cause of concern because of their potential impact on the health of workers and the public living in close proximity to such facilities. The biological hazards potentially associated with bioaerosol emission from composting activities include fungi, bacteria, actinomycetes, endotoxin and 1-3 β-glucans (Douwes et al. 2003). Effects of exposure to bioaerosols on respiratory health can include aspergillosis in immuno-compromised individuals, organic dust toxic syndrome, extrinsic allergic alveolitis, allergic rhinitis, asthma, upper airway irritation and mucous membrane irritation (Sykes et al. 2007). If the association between composting bioaerosol exposure and adverse health effects has been demonstrated for compost workers in a small number of cases (Herr et al. 2003; Bünger et al. 2007), the situation is not so clear concerning nearby residents. Some studies have found no relationship between respiratory symptoms and place of residence (Cobb et al. 1995) or with the concentration of Aspergillus fumigatus (Browne et al. 2001), while others have shown that residents located within 150-200 m of a composting plant were affected, suffering from irritative respiratory complaints similar to mucous membrane irritation (Herr et al. 2003). Therefore, even if it is clear that exposure to some airborne micro-organisms present in composting bioaerosols can lead to human health impacts, the interpretation of air sampling results remains difficult (Marchand et al. 1995). The risk assessments undertaken to date have relied on air dispersion modelling to estimate downwind concentrations of bioaerosols at a distance and to permit comparisons with measured data upwind or at background locations (Taha et al. 2006). Bioaerosol concentrations decrease rapidly with distance from their source and it becomes difficult to verify that measurements at a distance related to a specific activity, as opposed to other noncompost sources (Taha et al. 2005). Monitoring microbial indicators in the air of a specifically compost origin would help in improving bioaerosols’ dispersion analysis and, therefore, risk assessment. Adequate understanding of the microbial diversity of composting bioaerosols is a prerequisite to the definition of microbial indicators useful in monitoring composting bioaerosols. However, the microbial diversity of composting bioaerosols remains poorly understood despite their potentially associated sanitary risks. Most studies have been carried out using culture methods, while only a small fraction of total microbial diversity (less than 1%) can be cultivated in bioaerosols (Peccia and Hernandez 2006). Furthermore, previous studies have quantified global microbial groups (total bacteria, total fungi) (Heida et al. 1995; Van Tongeren et al. 1997) or a few specific groups (actinomycetes, Aspergillus fumigatus) (Millner et al. 1980; Lacey 1991; Gumonski et al. 1992; Darragh et al. 1997; Fischer et al. 1998, 1999, 2000), but the full range of microbial diversity in composting bioaerosols remains poorly described. The aim of this study was to characterize the microbial diversity of bioaerosols, collected at five open industrial composting plants, treating different types of waste (green waste, municipal solid waste, sludge from wastewater treatment plants, etc.) by a culture-independent approach. A common microbial signature was then looked for, and potential indicators of compost aerosols emitted into the air during pile turning were defined. Materials and methods Description of composting sites Five French composting plants were chosen based on their different operational characteristics (type of waste treated, type of aeration, turning technology, geographical location). Five bioaerosol samples were collected, one at each of the five composting plants. All composting processes were operated in open windrow. Bioaerosols were collected during the turning of compost piles in the thermophilic phase. The turning was performed either by a wheel loader or a windrow turning machine. Table 1 lists the characteristics of the compost pile that was manipulated (waste, age, temperature), the technology used for the turning and the type of aeration. The compost piles were in trapezoidal cross-sectional windrows with an average height of 3 m. The size of compost piles at site B, treating solid municipal waste, was smaller (2 m on average). Therefore, several composting piles (five) were turned during the sampling to get enough material for analysis. For this site, the waste was first passed for 2–3 days in a rotating drum (DANO® drum tube), screened and was then formed into piles. Table 1. Characteristics of compost windrows and composting process Type of waste Compost age (weeks) Compost temperature (°C) Technology used for windrow turning Forced aeration Site A Green waste 8 50 Wheel loader No Site B Solid municipal waste 3–4* 59–67 Windrow turning machine No Site C Green waste and sludge 3 60 Wheel loader Yes Site D Green waste, industrial biowaste, manure, mycelium, fats, waste from cosmetic industries 6 55 Wheel loader Yes Site E Green waste and sludge 6 60 Wheel loader Yes *Several compost piles were turned Air sample collection Air samples were collected through the air biocollector Coriolis®μ (Bertin Technologies, Montigny-le-Bretonneux, France). The air sampler was located a few metres away from the turning machine, directly in the dusty cloud, at a height of 1·2 m above the ground. The sampling was based on cyclonic forces which enabled the biological particles to be collected from the air into a sterile solution of Tween 20 (Sigma, St Louis, MO, USA) 0·005% in 1 l milliQ water. The air samples were gathered over 30 min at a flow rate of 300 l min−1. The solution was then filtered under sterile conditions on a 47-mm, 0·2-μm polyethersulfone filter (Supor 200; Pall Corporation, Ann Arbor, MI). Sterile collecting liquid was used to wash the cone surface and recover remaining biological particles, and then filtered on the same filter. Each filter was kept frozen at −80°C in a 2 ml sterile tube until DNA extraction. Extraction of total genomic DNA DNA extraction was performed directly on the filter using the protocol described by Godon et al. (1997), with slight modifications. The filter was ground to powder using a sterile plastic stick, while maintained frozen. Volumes of 400 μl of 4 mol l−1 guanidine thiocyanate in 0·1 mol l−1 Tris–HCl (pH 7·5) and 120 μl of 10% N-lauroylsarcosine were then added. The filter was frozen at −20°C. A volume of 500 μl of 5% N-lauroylsarcosine in 0·1 mol l−1 phosphate buffer (pH 8) was added to the frozen filter, and the tube was immediately held at 95°C for 10 min. After the thermic treatment, samples were put in a vibroshaker along with zirconia beads of different sizes (1–0·06; 0·1; 0·5 mm in respective proportions of 1/2/1 in mass). Polyvinylpolypyrrolidone was then incorporated to remove inhibiting substances. DNA was precipitated using isopropanol, and RNA was removed by RNAse treatment. Finally, the DNA was purified using the QIAamp DNA Kit (Qiagen, Hilden, Germany) and collected in 30 μl of sterile water. PCR Amplification and cloning of 16S rDNA and 18S rDNA fragments SSU ribosomal DNA fragments with an average size of 500 bp were amplified to build five bacteria and five Eukaryota libraries (Godon et al. 1997). The bacterial forward primer W18 (5′-GAGTTTGATCMTGGCTCAG-3′, positions F9-26) and the universal reverse primer W31 (5′-TTACCGCGGCTGCTGGCAC-3′, positions R500-517) were used to amplify the 16S rDNA. Eukaryotic forward primer W16 (5′-CTTAATTTGACTCAACACGG-3′, positions F960–979) and universal reverse primer W02 (5′-GNTACCTTGTTACGACTT-3′, positions R1492–1509) permitted the amplification of the 18S rDNA fragment. The reaction mixture was made up of 35 μl of water, 5 μl of Red Taq buffer (10X), 4 μl of dNTPs solution (2·5 mmol l−1), 2 μl of each primer solution (100 ng μl−1), 1 μl of Red Taq (Sigma), (2·5 U μl−1) and 1 μl of DNA. The PCR conditions were as follows: an initial denaturation step of 2 min at 94°C; 25 cycles (bacteria) or 30 cycles (Eukaryota) for 1 min at 94°C, 30 s at 50°C for the bacteria or 1 min at 47°C for Eukaryota and 1 min at 72°C; and a final elongation for 10 min at 72°C. The size of the amplified products was checked visually in a 0·7% agarose gel with 1% of ethidium bromide. The amplified products were then purified by the QIAquick PCR purification kit (Qiagen). The PCR products purified were cloned into E. coli using the TOPO TA cloning kit for sequencing (Invitrogen, Carlsbad, NM, USA), according to the manufacturer’s instructions. Each E. coli clone was purified on LB agar plates supplemented with kanamycin (100 μg ml−1) and then sent for sequencing after overnight incubation at 37°C (MilleGen, Toulouse, France). On average, 135 bacterial sequences and 35 eukaryotic sequences were obtained for each composting site. Sequence analyses Each sequence was cleaned from primers and plasmid sequences using BioEdit ver. 7.0.9.0 (Hall 1999). Sequences were checked for chimeras using the chimera detection program from Ribosomal Database Project (RDP) II (Cole et al. 2003) and Bellerophon (Huber et al. 2004). Sequences were aligned by Clustal W (Tompson et al. 1994), and a distance matrix was built with DNAdist (Felsenstein 1989) under BioEdit. Phylotypes were defined at the level of 97% similarity by the dotur program (Schloss and Handelsman 2005). Each phylotype was compared to sequences in the GenBank and embl databases with the Blastn algorithm available through the National Center for Biotechnology Information (Altschul et al. 1990). Phylogenetic distribution of clones belonging to the Bacteria domain was performed according to Rappé and Giovannoni (2003). For fungi, affiliation was carried out in accordance with Hibbett et al. (2007). The rarefaction curves were performed by Dotur at 97% of similarity. The Simpson and Schao1 indices were calculated using the Dotur program. Simpson was calculated as: Sobs representing the number of species observed, Si the number of individuals for one species and N the total number of species and Schao1 was calculated as: , Sobs representing the number of species observed, n1 the number of phylotypes with only one sequence and n2 the number of phylotypes with only two sequences. Good’s coverage was calculated as a percentage, according to the following relation C = [1 − (n/N)] × 100, where n represented the number of phylotypes appearing only once in a library and N being the library size (Ravenschlag et al. 1999). For each phylotypes, one sequence was deposited in the GenBank database. Accession numbers of bacterial and fungal nucleotide sequences were respectively FJ754668–FJ754887 and FJ754888–FJ754908. The bacterial and eukaryotic sequences were coded respectively as NA and EO for bioaerosol A, NC and EQ for bioaerosol B, NB and EP for bioaerosol C, Ni and ET for bioaerosol D and NE and ES for bioaerosol E. Identification of common phylotypes between the different bioaerosols was performed using the dotur program. To analyse the distribution in phyla of bacteria previously identified in compost (through culture and culture-independent approaches), 16S rDNA sequences were retrieved from the RDP public database (http://rdp.cme.msu.edu/) using the ‘compost’ keyword. Results Bacterial and fungal diversities in composting bioaerosols Bioaerosols were collected during the turning of composting piles on five composting platforms. Bacteria and Eukaryota rDNA libraries were built for each of the five bioaerosols. From the five bacterial 16S rDNA libraries, 685 bacterial sequences were obtained. They were distributed in 220 phylotypes defined at 97% of similarity threshold. Concerning eukaryotic sequences, the 173 sequences obtained were distributed into 21 sequences of insects (dipterans, springtails), five sequences of nematodes, 16 sequences of protozoa, four sequences of plants (conifer, grass), five sequences of Stramenopiles (three oomycetes and two algae) and 122 sequences of fungi. Twenty-one fungal phylotypes were defined. The microbial diversity of the bioaerosols was characterized using various parameters: rarefaction curves (Fig. 1), diversity indices and coverage (Table 2). Rarefaction curves indicated that the molecular analysis performed permitted only the identification of the dominant species in the bioaerosol, both for bacteria and fungi. The Simpson indices showed that the bacterial diversity was high for the five sites, while the Schao1 indices showed different abundancies. The bacterial coverage was higher than 70% for the A, B, D and E sites. The site C library had a much higher predicted bacterial diversity as given by Schao1 index than the other sites, together with a poor coverage rate (59%). The lowest values of the two diversity indices were obtained for the site B library. The fungal diversity was lower than the bacterial diversity and, for the B and E sites, one or two fungal species dominated (Simpson indices below 0·5). For the D and E sites, Schao1 values were the same (six phylotypes), while Simpson index values were different. Indeed, site E was dominated by one species in contrast to site D, which had a more diverse fungal diversity. The level of diversity of bacteria and fungi were very different, but the same tendency between sites was observed. Indeed, diversity was the lowest for site B and the highest for site C, for both bacteria and fungi. Figure 1Open in figure viewerPowerPoint Rarefaction curves determined from 16S rDNA (Bacteria) and 18S rDNA (Eukaryota) libraries for composting bioaerosol A ( △ ), B ( □ ), C ( ○ ), D ( • ) and E ( ). Table 2. Analysis of bacterial and fungal diversity in composting bioaerosols Sites Sequences number Phylotypes number Coverage C (%) Diversity indices Schao1 Simpson* Bacteria/fungi Bacteria/fungi Bacteria/fungi Bacteria/fungi Bacteria/fungi Site A 111/20 45/7 72/80 161/10 0·935/0·768 Site B 127/30 43/4 80/97 89/4 0·923/0·402 Site C 158/13 83/6 59/62 307/16 0·946/0·641 Site D 143/23 51/6 77/91 156/6 0·946/0·715 Site E 146/36 60/5 77/94 107/6 0·970/0·495 *Simpson indice values are presented as 1-D for best readability, the diversity increasing from 0 (one species) to 1 (maximal diversity). Identification and proportion of the dominant micro-organisms Table 3 gives the bacterial distribution at the phyla level for each composting site. Bacterial diversity was represented by eleven different phyla but was largely dominated by Firmicutes and Actinobacteria. Indeed, of the 220 bacterial phylotypes obtained overall for the five sites, 49·1% was assigned to Firmicutes and 37·2% to Actinobacteria. The remaining 13·7% was distributed between Proteobacteria (Alpha, Beta and Gamma), Bacteroidetes, Chloroflexi, Gemmatimonadetes, Planctomycetes, Thermotogae and TM7. TM7, Planctomycetes and Thermotogae were found only in one library. Even if the same tendencies were observed for the five sites (dominance of Firmicutes and Actinobacteria, presence of Firmicutes, Actinobacteria and Bacteroidetes in all libraries), differences were observed from one site to another. The proportion of Actinobacteria was much higher for site B (78% of the sequences), the only site treating solid municipal waste. Phyla other than Firmicutes and Actinobacteria were well represented in sites C and E. In particular, the representativity of Gamma-proteobacteria was much higher in the site E bioaerosol (16%) than in the others (0–2·5%). Table 3. Phylogenetic positioning and abundancy of bacterial phylotypes for the five composting bioaerosols. Only phylotypes with abundancy above 1% in one bioaerosol or more are presented. Phylotypes common to three, four and five bioaerosols are presented respectively in light grey shading, dark grey shading and in white-on-black shade. The table was ranked by decreasing abundancies of phylotypes Bacteria in the bioaerosols were then identified by analysing similarities with 16S rDNA sequences present in public databases. The phylogenetic identification of each phylotype along with its abundancy in the different bioaerosols is given in Table 3. Only phylotypes with abundancy above 1% have been included in Table 3. A table giving the identification of all bacterial phylotypes is included as supplementary material. The dominant bacteria were represented by Bacillus sp., Thermoactinomyces, Geobacillus and Planifilum for Firmicutes and by Saccharomonospora, Thermobifida and Saccharopolyspora for Actinobacteria. Gamma-proteobacteria highly present in the site E bioaerosol mainly corresponded to Acinetobacter sp. Differences from one site to another were observed. The dominant Actinobacteria were Saccharopolyspora in the A bioaerosol, Saccharomonospora and Saccharopolyspora in the B, Thermobifida in the C and D and Thermobifida, Corynebacterium and Microbacterium in the E bioaerosol. Among the Firmicutes, dominant species also varied from one site to another: Thermoactinomyces dominated in A, Thermoactinomyces and Planifilum in B and C, Bacillus in D and Geobacillus and Planifilum in E. The fungal diversity in the bioaerosols was largely dominated by Ascomycota although Basidiomycota, Microsporidia, Mucoromycotina and Entomophthoromycotina were also present (Table 4). From the total number of phylotypes (21), 67% was assigned to Ascomycota, 19% to Mucoromycotina and the remaining phylotypes being distributed among the three other phyla. In the E bioaerosol, all fungi identified belonged to Ascomycota. Mucoromycotina, as well as Ascomycota, were present in the A, C and D bioaerosols, and Entomophthoromycotina in the B bioaerosol. The C bioaerosol presented the highest number of phyla with species belonging to Ascomycota, Mucoromycotina, Basidiomycota and Microsporidia. Table 4 shows the percentage of sequences of the different Eukaryota among the five bioaerosols and their phylogenetic positioning. The three dominant fungi were Aspergillus, Penicillium and Thermomyces. As for bacteria, major differences were observed from one site to another concerning the distribution of the different phylotypes. For the A bioaerosol, Aspergillus fumigatus (30%) and Penicillium (30%) dominated, whereas Thermomyces was highly dominant in the other four. Table 4. Phylogenetic positioning and abundancy of fungal phylotypes. Phylotypes common to four and five bioaerosols are presented respectively in grey shading and in white-on-black shade. The table was ranked by decreasing abundancies of phylotypes Origin of the closest relatives in public 16S rDNA databases In order to decide whether the sequences obtained were affiliated to bacterial phylotypes previously identified in compost, the origin of the closest 16S rDNA sequences present in public databases was analysed for each phylotype. Origins obtained for each bioaerosol are gathered in Fig. 2. From the different sources involved (soil, water, compost, air, aerobic and anaerobic treatment processes, etc.), the two origins mainly observed were soil and compost. The percentage of sequences affiliated to compost varied from 28% (bioaerosol B) to 70% (bioaerosol D). A significant percentage of sequences linked to several environments was obtained for bioaerosols A (17%) and E (20%). The fact that the closest relatives in public databases originated mainly from compost samples confirmed that the dominant bacteria identified in this study came from the compost and had been released during the turning of the composting pile. This compost origin was also confirmed by the dominance of thermophilic species, linked to the fact that the composts were in their thermophilic phase at the time of sampling. Indeed, the bioaerosols harboured a high diversity of thermophilic bacteria belonging to Geobacillus, Planifilum, Bacillus, Clostridium and thermophilic actinomycetes (Thermoactinomyces, Saccharomonospora, Saccharopolyspora, Thermobifida, Streptomyces, Thermomonospora, Nocardiopsis, Thermocrispum, etc.). Among the less dominant phylotypes, some sequences previously found in several environments corresponded to ubiquitous and/or psychrophilic bacteria (Planococcus, Micrococcus, Acinetobacter, Pedobacter, Chryseobacterium, etc.). Figure 2Open in figure viewerPowerPoint Environmental origin of the closest bacterial relatives in public databases according to 16S rDNA similarities (BLAST). The closest bacterial relatives recovered from several environments are shown under the ‘several environments’ denomination. The ‘not determined’ denomination corresponds to sequences for which the environment of the closest relatives is not provided in the database, as well as to sequences having less than 95% similarity with their closest relatives. The term ‘Bioreactor’ was used for aerobic and anaerobic treatment processes. Identification of core species common to several bioaerosols Beyond the description of microbial diversity in composting bioaerosols, the objective of this research was to look for phylotypes common to several bioaerosols that would represent a microbial signature of compost aerosols. A significant number of bacterial phylotypes were indeed common to several bioaerosols; they are presented in grey shadings in Table 3. Thirty-two phylotypes (gathering 150 sequences) were common to two sites, five phylotypes (76 sequences) to three sites and six phylotypes (176 sequences) to four sites. Only one phylotype was present in the five bioaerosols; it represented 6·9% of total sequences and had 97% similarity with Saccharopolyspora rectivirgula. The six phylotypes found in four libraries were close to Planifilum yunnanense, Geobacillus thermodenitrificans, Thermactinomyces intermedius, a Thermoactinomycetaceae sp., Thermobifida fusca and Saccharomonospora glauca. Fungi common to different bioaerosols could also been identified: three phylotypes were common to two sites (gathering nine sequences), two phylotypes to four sites (gathering 25 sequences) and only one phylotype was present in the five libraries. It was affiliated to Thermomyces lanuginosus, which gathered 49% of fungal sequences, and was differently represented in the five bioaerosols: 5% in A, 77% in B, 54% in C, 39% in D and 56% in E. The two phylotypes common to four libraries were close to a Penicillium sp. (10·6% of fungal sequences) and to Aspergillus fumigatus (10·6% of fungal sequences). Aspergillus fumigatus was found in all composting bioaerosols apart from C. Discussion Bioaerosols produced during composting have never been described in their overall microbial diversity. Using culture-independent techniques, this study reports the description of the diversity of bacteria and Eukaryota in bioaerosols released at five different composting sites during the turning of composting piles. The objective was to compare five bioaerosols originating from different industrial sites to see whether a common microbial signature could be defined, even when originating from composts made up of different wastes and managed by different processes. Origin of micro-organisms recovered from the bioaerosols During the turning of composting piles, bioaerosols are released into a nonsterile environment, air (Stetzenbach 2002). Thus, micro-organisms gathered by sampling may originate either from the compost or from the surrounding air. Because the aim of the study was to identify microbial indicators of a compost origin present in the air, the question of the origin of the micro-organisms identified was critical. Ascomycota and the dominant bacterial genera found in the bioaerosols (Bacillus, Geobacillus, Planifilum, Thermoactinomyces, Saccharomonospora, Saccharopolyspora, Thermobifida, Penicillium and Aspergillus) had been previously described in the literature as part of th