Covariation between air chiller and chicken product microbiota in a poultry processing facility

Birgitte Moen; Annette Fagerlund; Sophie Marie Pursti; Merete Rusås Jensen; Solveig Langsrud

doi:10.1371/journal.pone.0344423

Abstract

Chicken products are prone to microbial spoilage, and early detection of deviations in contaminating microbiota can improve quality while reducing waste and customer complaints. In industrial poultry processing, the chiller is a key post-slaughter step but may serve as a source of microbial contamination. In this study the variation in microbial status of the air chiller surface was profiled together with freshly produced and end-of-shelf-life chicken legs and skinless breast fillets across 18 production batches over eleven weeks. Samples were taken early and late shifts on Mondays and Thursdays. The air chiller was cleaned weekly on Fridays. Products were stored under modified atmosphere (60% CO₂/40% N₂) for 21 days at 4˚C, and a subset underwent temperature abuse (8˚C) during the last four days of storage. Microbial community composition was determined using partial 16S rRNA amplicon sequencing. Higher bacterial loads were found on air chiller swabs and fresh products collected at late shift on Thursdays than Mondays (p_adj < 0.05). A similar trend was observed for products at the end of shelf-life. At the end of shelf-life, the microbiota in chicken legs was dominated by Carnobacterium, Yersiniaceae and Lactobacillales, whereas breast fillets were dominated by Lactobacillus, Lactobacillales and Carnobacterium. The air chiller microbiota was dominated by Acinetobacter, Psychrobacter and Pseudomonas. Across weeks and sampling times, microbial load and community composition covaried between the air chiller and products, indicating that temporal shifts in the air chiller environment influenced or reflected the product microbiota. These findings identify the air chiller as a critical control point for targeted monitoring and intervention to mitigate spoilage risk. Recommended measures include mid-week cleaning, condensation and aerosol management, and MAP-matched incubation of environmental swabs to reveal CO₂-tolerant spoilage bacteria.

Citation: Moen B, Fagerlund A, Pursti SM, Jensen MR, Langsrud S (2026) Covariation between air chiller and chicken product microbiota in a poultry processing facility. PLoS One 21(3): e0344423. https://doi.org/10.1371/journal.pone.0344423

Editor: Wadzani Palnam Dauda, Federal University Gashua, NIGERIA

Received: September 12, 2025; Accepted: February 19, 2026; Published: March 10, 2026

Copyright: © 2026 Moen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The sequencing data have been deposited in the National Center for Biotechnology Information (NCBI) archives under BioProject no. PRJNA1308980. All other relevant data are within the manuscript and its Supporting information files.

Funding: This work was supported by the Norwegian Fund for Research Fees for Agricultural Products (FFL) through the strategic programs FutureFoodControl and MicroPack (grant no. 314743 and 354136), and European Union’s Horizon Europe research and innovation programme under grant agreement N° 101136248 MICROORC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The consumption of poultry meat (chicken, turkey, and duck) is steadily increasing worldwide driven by affordability, favorable nutrition and low calorie density, and broad cultural acceptability. In the European Union, poultry production amounts to around 14 million tons per year and is projected to keep gaining market share relative to red meats [1–4]. Poultry products are, however, easily prone to microbial spoilage, resulting in consumer complaints related to off-odors, economic losses for producers and retailers, and increased environmental burdens due to food waste.

Food spoilage microorganisms commonly originate from raw materials or the processing environment, where surfaces of equipment and tools have a significant impact on the microbiota of the final food products [5–7]. Recent shotgun-metagenomics work found no evidence of farm-to-fork transmission of the carcass microbiome or resistome after chilling, instead suggesting the processing plant as main source of post-chill carcass microbiome [8]. Together with facility-wide environmental mapping that links equipment and surface microbiomes to product profiles [5], this points to the need for targeted within‑facility studies of potential critical control points that can shape product microbiota.

Although spoilage bacteria often represent only a small fraction of the initial microbiota, their selective growth during storage can lead to microbial levels that render food unsuitable for human consumption [9,10]. Multiple factors influence both the introduction and growth of spoilage organisms, including animal-associated microbiota, seasonality, the slaughter process, and storage conditions such as temperature and packaging atmosphere [3,11]. Depending on these factors, the spoilage microbiota of poultry products commonly includes Pseudomonas, Enterobacteriaceae, Brochothrix thermosphacta or lactic acid bacteria (LAB), mainly belonging to the genera Lactobacillus, Carnobacterium, Leuconostoc and Lactococcus [5,11–14]. In addition, Photobacterium has been isolated from spoiled modified atmosphere packed poultry [15].

The shelf-life of aerobically stored chicken is short (five to six days), and modified atmosphere packaging (MAP) is therefore commonly used to extend product durability. In Norway, most chicken products are packed under high-CO₂ atmospheres (50–70% CO₂), or vacuum, with shelf lives of 16–20 days depending on product type and producer. Under CO₂/N₂ atmospheres, aerobic bacteria are suppressed and high CO₂ levels inhibit the growth of key aerobic, psychrotrophic spoilage organisms, such as Pseudomonas spp., thereby prolonging the shelf-life [16–18]. Combined with refrigeration, high-CO₂ MAP selects for a diverse cold-, anaerobic- and CO₂-tolerant microbiota at the end of shelf-life, typically dominated by LAB and varying compositions of B. thermosphacta, Janthinobacterium spp., Serratia spp., Shewanella spp., Yersinia spp., Klebsiella spp., Acinetobacter spp. and Hafnia spp. [16]. Spoilage under MAP is not defined solely by bacterial numbers or which ones dominate, but by the capacity of CO₂‑tolerant taxa to generate spoilage‑associated metabolites, and that capacity varies with the specific MAP gas composition [19–21].

The chicken slaughter process involves multiple steps (reception and stunning, slaughter, scalding and plucking, evisceration, chilling, cutting and packaging), each of which can reshape carcass microbiota. Among these, air chilling has been identified as a hotspot for contamination [6,22]. Condensation on walls and dripping on floors supports growth of psychrophilic bacteria such as LAB and Pseudomonas, which can aerosolize during washing or be redistributed by air movement in the air chiller [6,16,23,24]. Consistent with this, Vihavainen et al., [24] detected airborne LAB in the air-chilling area and identified similar LAB groups in late-shelf-life retail broiler products. Notably, the products were retail samples from three different manufacturers, while the environmental air sampling was conducted in a single processing facility. Despite numerous chicken shelf-life investigations, few have paired environmental and product sampling within the same facility and time window. A recent comparison of environmental microbiota and stored products also drew materials from different factories and even different countries, limiting facility level conclusions [5].

The microbial composition of processed raw chicken products, both immediately after processing and at the point of consumption, can show considerable variation despite rigorous hygiene management along the production chain. To assess processing‑related factors that contribute to this variability, we sampled 18 production batches of legs and skinless breast fillets over five weeks, collecting products during early and late shifts on Mondays and Thursdays. In parallel, air chiller surfaces (steel plates below the carcasses) within the same facility were swabbed. Both fresh and end‑of‑shelf‑life chicken products were analyzed, and agar plates were incubated under the same CO₂/N₂ MAP gas used for consumer packaging, to emphasize recovery of CO₂‑tolerant spoilers. Because the air chiller is cleaned weekly on Fridays and left to dry over the weekend, we reasoned that microbial loads could accumulate and be redistributed via aerosols and condensate during the production week, potentially shaping downstream product microbiota. We therefore hypothesized that air chiller bacterial loads would increase during the production week and that air chiller microbial load and community composition would covary with those of fresh and end‑of‑shelf‑life products across weeks and shifts within the same facility.

Materials and methods

Chicken products (legs and skinless breast fillets) and cloth swabs from the air chilling tunnel (air chiller) were obtained from a single Norwegian broiler processing plant. The chicken products were packed in modified atmosphere (60% CO₂/40%N₂) at the processing plant and delivered to the laboratory at Nofima late afternoon the same day. The swabs from the air chiller were collected from steel plates below the carcasses, not product contact. The air chiller was cleaned once a week on Fridays and left to dry during the weekend. This observational study was conducted exclusively post‑slaughter within the processing plant (air‑chilling area) and on finished products; no live birds were handled, and farm‑level husbandry metadata were not collected.

Experimental design

Swab samples from the air chiller and chicken products (legs and skinless breast fillets) were collected at 18 different sample times (batches) over a period of five non‑consecutive weeks from November 2021 to January 2022. Two parallel swabs from the air chiller and two parallel legs or breast fillets were collected from the processing line at each sample time (Table 1). The legs and breast fillets were packed individually in modified-atmosphere packaging trays, similar to the consumer packaging gas used in the processing plant. To maximize the potential variation of the process, samples were collected twice a week (Mondays and Thursday) to account for day-to-day process variability, and at the start of production (early) and mid-afternoon (late) to account for batch variation (Mondays (early = M1 and late = M2) and Thursdays (early = T1 and late = T2), see Table 1. Mondays and Thursdays were chosen to represent products processed immediately after cleaning the air chiller and after several days without cleaning (the production plant cleaned their air chiller once a week, at the end of production on Fridays and allowed the tunnel to air dry over the weekend). Swabs from the air chiller and MAP chicken products (60% CO₂/40%N₂) were transported in cooler units with ice packs from the processing plant to the lab at Nofima the same afternoon. The swabs from the air chiller were analyzed the same afternoon, while the chicken products were analyzed the following morning (day 1) or after storage at 4˚C for 21 days. The stated end-of-shelf-life (as determined by the company) was 19 days, but 21 days were chosen to investigate the potential of a longer shelf-life. Day 21 is therefore referred to as the end of shelf-life in this paper. Some products were also moved from 4˚C to 8˚C after 17 days for further storage to simulate a higher temperature (temperature abuse) in a consumer fridge.

Download:

Table 1. The total number of samples collected and analyzed. Two parallel samples were collected from each sample type (air chiller (cloth swabs), breast fillet or leg (two packages)) from different weeks, and sample times.

https://doi.org/10.1371/journal.pone.0344423.t001

The processing plant received birds from many different producers, and two broiler breeds were received during this study: Ross 308 and Liveche. Ross 308 is a standard, fast growing breed and was in this study slaughtered at the age of 34 days. The Liveche broiler is a slower growing hybrid broiler and was slaughtered at the age of 52 days. The first batch on Thursdays (T1) were Liveche and the others Ross308, which was the standard production breed at the facility. Due to different circumstances that were out of our control, some samples are missing from the dataset: Thursday of Week 3 (both T1 and T2) and early Monday (M1) fillet samples of Week 5. In addition, all samples collected on Mondays of Week 2 were moved to 8°C after 17 days by a mistake. A detailed overview is presented in Table 1.

Sampling and microbial analysis

The chicken meat was aseptically sampled by excising a 9 cm² × 1 cm (approximately 10 grams) section with a sterile scalpel, diluted by approximately 90 ml peptone water until 1/10 dilution was attained before homogenized in a stomacher for 60 s, and 10-fold dilutions were made and spread on Plate count agar (PCA) plates (Merck). For the chilling tunnel, 600 cm² of the horizontal steel surfaces exposed to drippings and air were sampled using sterile Sodibox swab cloths (Sodibox, France). Pre-moistened (with neutralizing broth) cloths were used to sample at the start of production and dry cloths were used in the afternoon. The dry cloths were 20 g lighter than the pre-moistened cloths and added 75 ml peptone water before homogenization while 55 ml were added to the pre-moistened swabs. The cloths were homogenized in a stomacher instrument for 60 s. Plating was performed either by an automated plater (Whitley Automated Spiral Plater, Don Whitley Scientific limited, Bingley, United Kingdom) or manually with single-use L-shaped spreaders. The detection limits were 1.25 Colony Forming Units (CFU)/cm² (0.1 log CFU/cm²) and 50 CFU/g (1.7 log CFU/g) for cloths and chicken meat products, respectively. When comparing the bacterial load from the air chiller and product samples, it was assumed that one cm² approximately equals one gram. The bacterial loads are referred to Total Viable Counts (TVC) in the text.

The agar plates were incubated at 15˚C either aerobically for 7 days or in 60% CO₂/40%N₂ (corresponding to the MAP used for the packed products) for 10 days. Since storing products in MAP already selects CO₂-tolerant bacteria, end-of-shelf-life samples were not incubated in high CO₂. The aerobically incubated and 60% CO₂/40%N₂ packed agar plates are further referred to as aerobic and MAP, respectively in tables and figures.

Identification of colonies

Approximately ten randomly picked colonies (that is all colonies within a randomly selected sector on the agar plate) from each sample were selected for direct analysis by MALDI-TOF MS. When plates had less than 10 colonies, all colonies were picked. From the air chiller samples, only colonies from samples collected early on Mondays (M1) and Thursdays (T1) were identified. In total, 1914 colonies were picked (1035 colonies from fresh products, 726 from the air chiller, and 153 from end-of-shelf-life products (only from week 4, M1 and M2)). In total 1563 were successfully identified by MALDI TOF MS (82%). The unidentified colonies were labelled NoID.

The direct transfer protocol was followed to obtain mass spectra. Briefly, ∼0.1 mg of cell material was directly transferred from a bacterial colony to a target plate and overlaid with 1 μL of matrix solution (10 mg/mL a-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 2.5% trifluoroacetic acid). MS analysis was performed on an Biotyper MALDI-TOF mass spectrometer (Bruker Daltonics, Germany) using MBT Compass 4.1 and FlexControl 3.4 software (Bruker Daltonics). Calibration was carried out with the Bacterial Test Standard (Bruker Daltonics, Germany). All MS spectra were measured automatically using Flex Control software according to the standard measurement method for microbial identification, MBT-autoX.axe autoExecute. The BioTyper 3.1 software (Bruker Daltonics) equipped with MBT 9607MSP Library (released November 2020) together with our in-house library were used for classification. According to the manufacturer’s guidelines, identification scores of ≥2.0 were considered reliable for species-level identification, while scores between 1.7 and 2.0 were accepted for genus-level identification. Scores below 1.7 were deemed unreliable and registered as “NoID”.

DNA extraction and Illumina partial 16S rRNA amplicon sequencing

Samples from the air chiller and stored products were subjected to Illumina partial 16S rRNA amplicon sequencing. Two aliquots of 2 mL homogenate per sample were collected directly after homogenization and pelleted by centrifugation at 13000 × g for 5 min. The pellet was frozen at −20˚C until the time of analysis. DNA extraction from pellets was done with the DNeasy PowerLyser PowerSoil kit (Qiagen) following the kit protocol, the pellets were homogenised by tissue homogeniser (Precellys Evolution 24 Tissue homogeniser) for 3 × 40 seconds at 7400 rpm with 10 seconds between intervals.

16S rRNA gene PCR (V4 region) and paired end sequencing (2x150 bp) using the MiSeq Reagent Kit v3 on a MiSeq instrument (Illumina) was performed using the protocol presented by [25] as previously described [26].

Sequence analysis

The sequences were processed in QIIME2 (qiime2–2020.11) [27]. Briefly, the data processing was: demultiplexed using demux, paired ends were joined using vsearch, quality filtered based on a q-score above 30, denoised using deblur 16S, and taxonomy was achieved using classify-sklearn with the SILVA database (Silva 138 99% OTUs from the 515F/806R region) [27–33]. Singletons were removed and one outlier (bloated package) was removed, in addition to mitochondrial- and chloroplast sequences. The dataset was also split based on sample type and product type. The genus level taxonomy table was exported to text files and further processed in Excel. The genus table was converted to relative values and taxa below an average across all samples of 0.1% was represented as “Other”. A selection of representative sequences was submitted to BLAST nucleotide search (https://blast.ncbi.nlm.nih.gov/) (assessed 05.07.2024) to obtain additional information about possible species. The standard nucleotide collection database and the highly similar sequences option were used.

Gas analysis

Prior to opening, the CO₂ and O₂ in headspace of the packages were analysed at each sampling time by a CheckMate 9900 O₂/CO₂ analyser (PBI Dansensor, Ringsted, Denmark). The accuracy of the instrument is 0.01% O₂/CO₂. The mean gas composition of product packages was 64% CO₂/36% N₂ (CO₂/N₂ ratio min. = 55/45; max. = 69/39) and 58% CO₂/42% N₂ (CO₂/N₂ ratio min = 47/53; max = 65/35) after respectively one and 21 days of storage. One outlier had a gas composition of 4% CO₂, 79% N₂ and 17% O₂, which is near atmospheric levels of oxygen (21%). This sample (breast fillet stored 21 days at 4°C collected mid-afternoon on Monday (M2), week 2) was removed from further statistical analyses. The mean gas composition after 10 days incubation of the CO₂ agar plates was 38% CO₂–61% N₂ with 1% O₂ (CO₂/ N₂ ratio min. = 37/60, max = 38/61.

Statistical analysis

Bacterial loads. Analyses were run separately by subset. For the air‑chiller swabs and fresh products, we fitted a General Linear Model (GLM; ANOVA framework) with fixed effects for Week (levels 1–5), Sample time (M1, M2, T1, T2), and Incubation atmosphere (aerobic vs MAP; 60% CO₂/40% N₂). For end‑of‑shelf‑life products, we fitted a GLM with fixed effects for Week (1–5), Sample time (M1, M2, T1, T2), and storage temperature (4 °C vs temperature‑abuse). TVC was log₁₀‑transformed prior to analysis.

For the effects of week and sample time, pairwise comparisons were performed on model‑based estimated marginal means (least‑squares means, LS‑means) using Tukey’s method to control the family‑wise error rate at α = 0.05. Throughout, “p” denotes GLM p‑values for main effects, while Tukey‑adjusted p‑values for pairwise contrasts are reported as p_adj* (padj). Model assumptions were assessed using residual diagnostics (residuals vs fitted values and normal Q–Q plots) and the Anderson–Darling normality test of residuals in Minitab. If a subset showed material departure from normality (e.g., Anderson–Darling p < 0.05), we applied a pre‑specified sensitivity analysis in which the GLM was re‑fitted after excluding the affected observations, and Tukey comparisons were repeated. Conclusions were drawn using two‑sided α = 0.05. All analyses were performed in Minitab (Minitab Statistical Software, Version 21.2/22.4). The TVC was illustrated using box plots (Minitab). The line inside the box represents the median, which is the middle value of your data set. The box itself spans from the first quartile (Q1) to the third quartile (Q3), representing the middle 50% of the data. The lines extending from the box (whiskers) show the range of the data, excluding outliers. The whiskers typically extend to the smallest and largest values within 1.5 times the interquartile range from the quartiles. Data points that fall outside the whiskers are considered outliers and are often marked with asterisks (*) or dots. Individual standard deviations were used to calculate the intervals.

Correlation analysis was performed to calculate the correlation between the TVC for different samples. The correlation coefficient (R value ranges from −1–1) is a measure of the strength and direction of the linear relationship between two variables in regression analysis.

Since Thursday samples were missing from week 3, calculations were performed both with and without week 3 samples. The tendency was that the variation between the sample categories was slightly higher (higher F-value) when excluding week 3 from the calculations, but with the same significant variables. The results presented therefore include week 3 samples. There are two samples that were excluded from the analysis; one due to bloated package and one sample that was missing plate counts.

Alpha- and beta diversity. All sub-operational taxonomic units (sOTUs) were aligned with mafft [34] (via q2‐alignment) and used to construct a phylogeny with fasttree2 (Price et al., 2010) (via q2‐phylogeny). The alpha- and beta diversity Principle Coordinate Analysis (PCoA) was estimated using q2‐diversity (core-metrics-phylogenetic) [27] with a sampling depth of 25,000 (or 50,000 for the air chiller). The results for the different metrics (weighted and unweighted UniFrac, Bray-Curtis, and Jaccard) were visually inspected using principal-coordinate analysis (PCoA) plots. The result from the Bray-Curtis metric was used for visualization because it captures abundance‑weighted compositional differences (emphasizing shifts in dominant taxa) while ignoring joint absences and without imposing a phylogenetic model; patterns were qualitatively consistent across metrics. Alpha-diversity differences among groups were evaluated in QIIME 2 (qiime2–2023.5) using the `alpha-group-significance` workflow (Kruskal–Wallis test). We report the Kruskal–Wallis H statistic (H) and the Benjamini–Hochberg false-discovery-rate–adjusted p-value (q). H quantifies the magnitude of rank differences among groups; p reflects the probability of observing those differences under the null hypothesis; and q controls for multiple testing across comparisons. Unless otherwise stated, two-sided tests were used and statistical significance was defined as q < 0.05.

Difference in bacterial community diversity across different processing and storage parameters were statistically evaluated using pairwise permutational multivariate analysis of variance (PERMANOVA) tests in QIIME2 (plugin: diversity beta-group-significance) [35,36]. The test is based on the prior calculation of the distance between any two objects in the experiment. We evaluated three distance metrics for microbial community comparison: Bray-Curtis, weighted UniFrac, and unweighted UniFrac. The test statistic is a pseudo F-ratio, similar to the F-value in ANOVA. It compares the total sum of squared dissimilarities (or ranked dissimilarities) among objects belonging to different groups to that of objects belonging to the same group. Larger F-ratios indicate more pronounced group separation, however, the significance of this ratio is usually of more interest than its magnitude. For each sample category (e.g., product, air chiller, end‑of‑shelf‑life breast fillets and legs), we first tested overall group differences (“group significance”) among sample part, week, sample time, and temperature with PERMANOVA (999 permutations). To investigate the difference between sampling times, the Bray-Curtis metric was used in pairwise PERMANOVA within the same category. We report pseudo‑F and permutation p‑values for each test. For pairwise contrasts, we control the family‑wise false discovery rate using Benjamini–Hochberg; adjusted values are reported as q‑values.

Results

The bacterial load of the air chiller and fresh breast fillets and legs was estimated by incubating agar plates both aerobically and under modified atmosphere (high CO₂ like the product MAP) to select for bacteria able to proliferate during MAP storage of the final products. The gas composition was measured before opening the product packages, as well as the packages containing the MAP agar plates. The bacterial composition was determined by identification of isolates by MALDI-TOF MS (air chiller, fresh samples, and a selection of end-of-shelf-life products) and Illumina 16S rRNA amplicon sequencing (air chiller and end-of-shelf-life products).

Bacterial load

The bacterial load was significantly higher (p < 0.001) in leg samples (with skin) than breast fillet samples both for fresh samples and at the end of shelf-life (Table 2). As expected, incubating the agar plates under high CO₂ (MAP) resulted in significantly (p < 0.001) lower TVC (0.5–0.9 log lower), since CO₂ inhibits growth of several bacteria. There was a tendency of higher TVC of products after temperature abuse (8°C the last four days of storage) compared to 4°C, but this difference was only statistically significant for leg samples (p = 0.001).

Download:

Table 2. Bacterial loads (log₁₀ TVC) for the air chiller and product samples (day 1 and end-of-shelf-life). The TVC was measured at day 1 and at the end-of-shelf-life by enumeration on PCA incubated at 15˚C either aerobically or in MAP. Product samples were stored either at 4°C or at 8°C the last four days of storage (abuse).

https://doi.org/10.1371/journal.pone.0344423.t002

Accumulation of bacteria during the week in the air chiller.

Since the air chiller is cleaned once a week on Fridays and left to dry during the weekend, it was hypothesized that the bacterial load would accumulate during the production week. The bacterial load in the air chiller was significantly different between weeks (p < 0.001 and among sampling times (p = 0.010) for both aerobic and MAP counts. There was an increase in the mean TVC from the first to the last sampling week, and a significantly higher TVC in samples collected late on Thursdays (T2) than Mondays (M1 and M2) (p_adj = 0.015–0.033; Fig 1A-1B). Although the MAP incubated plates had a lower TVC compared to the aerobic plates, there was a positive correlation between aerobic and MAP counts for the air chiller, R = 0.9 (p < 0.001), indicating the same trend independent of incubation atmosphere.

Download:

Fig 1. Total viable counts (TVC; log CFU/cm²) in the air chiller.

Boxplot for weeks (A) and sampling time (B) shown for aerobic and MAP counts. ** above brackets represent significant difference (p_adj < 0.05) based on Tukey‑adjusted p-values from the GLM (LS‑means; family‑wise α = 0.05).

https://doi.org/10.1371/journal.pone.0344423.g001

Variation in the bacterial load of chicken products.

Early and late batches were collected on Mondays and Thursdays to investigate if there was a random or systematic variation in the bacterial load compared to the air chiller. The bacterial load of both fresh and end-of-shelf-life products varied between sampling times (M1, M1, T1 and T2) and showed similar tendency independent of agar incubation atmosphere or storage temperature (Fig 2A-2D).

Download:

Fig 2. Total viable counts (TVC; log CFU/g) in product samples.

Boxplot of aerobic and MAP counts stratified by sampling time for day 1 breast fillets (A) and day 1 legs (B), and end-of-shelf-life under 4°C storage or temperature abuse conditions for breast fillets (C) and legs (D). ** above brackets represent significant difference (p_adj < 0.05) based on Tukey‑adjusted p-values from the GLM (LS‑means; family‑wise α = 0.05).

https://doi.org/10.1371/journal.pone.0344423.g002

For fresh breast fillets (day 1) there was a significant difference in TVC between weeks for aerobic counts (p = 0.030) and between sampling times for both aerobic- and MAP counts (p = 0.023 and 0.014, respectively). Samples collected late on Thursdays (T2) had a significantly higher TVC than M2 (p_adj = 0.045) and T1 (p_adj = 0.026) samples for aerobic counts, and a significantly higher TVC than M1 (p_adj = 0.045) and M2 (p_adj = 0.015) for MAP counts. For fresh legs there was a significantly different TVC between sampling times for MAP counts where samples collected late on Thursdays (T2) had a significantly higher TVC than M1 (p_adj = 0.020) and M2 (p_adj = 0.020) samples (Fig 2A-2B).

At the end-of-shelf-life, TVC in legs differed significantly across sampling times (p = 0.006) and between 4°C and temperature abuse (p = 0.001). For legs stored at 4°C, there was a tendency of higher TVC in samples collected late on Thursdays (T2) compared to early Mondays (M1), similar as for fresh leg samples. TVC in breast fillets exhibited greater variability than in legs and did not differ significantly by week, sampling time, or storage temperature, although a Thursday > Monday trend was observed (Fig 2C-2D).

Microbial diversity in the air chiller and fresh products (culture dependent)

Culturable bacteria in samples from the air chiller and fresh products was identified by MALDI-TOF to complement the 16S data. Plates were incubated aerobically and under MAP‑conditions to recover viable, spoilage‑relevant taxa expected under processing/packaging. A total of 778 colonies from fresh products and 636 isolates from the air chiller were identified by MALDI-TOF MS. Fresh product samples exhibited the highest diversity, with 41 identified genera. In contrast, 18 genera were detected in the air chiller.

From the fresh products and the air chiller, 16 genera (see Fig 3) were detected more than 10 times across all samples. Pseudomonas (29%), Escherichia (10%), Arthrobacter (8%) and Carnobacterium (6%) accounted for more than 50% of the total colonies identified. All isolates identified as Escherichia were E. coli. There was a higher diversity of bacteria isolated from aerobically incubated plates than from those incubated in MAP, in addition to a higher number of unidentified colonies. Aerobically incubated plates also showed a greater difference in the bacterial composition between the air chiller and product samples (Fig 3).

Download:

Fig 3. Relative abundance (%) of colonies identified from the air chiller, fresh breast fillet- and leg samples.

The columns represent an average of several samples and n number of identified isolates from either aerobically- or MAP incubated agar plates. The figure also includes the relative abundance of colonies where no identification was obtained (noID). Genera identified less than 10 times are combined and represented as “Other”.

https://doi.org/10.1371/journal.pone.0344423.g003

The bacterial load in the air chiller was lower when the agar plates were incubated in MAP compared to aerobically, 4.3 log CFU/cm² and 5.2 CFU/cm² respectively (Table 2). Fig 3 shows that most isolates identified from the air chiller belonged to the genus Pseudomonas and that MAP suppressed growth of many bacterial genera. Growth of Pseudomonas on MAP plates, indicated that there might have been residual oxygen during incubation (the agar plates were not flushed with CO₂). Incubation in MAP also selected for growth of Carnobacterium, which was the dominating identified genus in end-of-shelf-life products (see Colony based bacterial composition of end-of-shelf-life products). From the air chiller, less than 1% of the isolates from aerobic plates were identified as Carnobacterium compared to 22% of the isolates from MAP plates. All Carnobacterium isolates identified from the air chiller from MAP plates belonged to Carnobacterium maltaromaticum. For aerobically stored plates two isolates belonged to C. maltaromaticum and one could not be assigned to strain level.

For fresh breast fillets the microbial diversity was much lower on MAP plates than aerobic plates, with a dominance of Pseudomonas, Escherichia and Carnobacterium on MAP plates. The percentage of Carnobacterium was 4% and 9% for aerobic and MAP plates, respectively, where approximately 90% belonged to Carnobacterium divergens and 10% C. maltaromaticum. A similar lower diversity was observed for MAP plates from fresh leg samples, with dominance of Escherichia and Pseudomonas. The percentage of Carnobacterium was below 1% for both incubation atmospheres, where all isolates from MAP plates were identified as C. maltaromaticum.

In contrast to the bacterial load, no clear trends were found in the cultured microbiota between weeks and sampling times.

Microbial diversity in the air chiller and end-of-shelf-life products (culture independent)

To compare the microbiota between the air chiller and end-of-shelf-life products, 16S rRNA amplicon sequencing was performed.

In total, 133 samples were analyzed and, altogether, 932 sOTUs were detected from a total of 10,244,028 sequences after filtering (see Materials and methods). The mean number of sequences obtained per sample was 77,022 (max 139,935 and min 25,375). The mean number of sequences per sOTU was 10,991 (max 163,8163 and min 9). The sequencing data have been deposited in the National Center for Biotechnology Information (NCBI) archives under BioProject no. PRJNA1308980.

The dominating genera in the air chiller was Acinetobacter (27%), followed by Psychrobacter and Pseudomonas (both 17%). In leg samples the most prevalent bacterial taxa were Carnobacterium (29%), followed by Yersiniaceae (16%) and Lactobacillales (12%), while Lactobacillus (52%), followed by Lactobacillales (19%) and Carnobacterium (12%) were most prevalent in breast fillet samples. Fig 4 shows the dominating microbiota in the air chiller and end-of-shelf-life products across sample types, weeks and sample times.

Download:

Fig 4. Microbiota in air chiller and end-of-shelf-life products.

Barplot of the relative abundance of the most abundant bacterial taxa in the air chiller and end-of-shelf-life products across sample types, weeks and sample time. The results are represented as an average of the two parallels for the different sample times. All taxa with average relative abundance > 0.1% across all samples are shown, the rest is represented as Other. The taxa are mostly at genus level but may also be at family (e.g., Enterobacteriaceae) or order (e.g., Lactobacillales) when a lower taxonomic identification was not possible.

https://doi.org/10.1371/journal.pone.0344423.g004

In the air chiller there was a tendency for a higher relative abundance of Acinetobacter in samples collected in weeks 4 and 5, especially prevalent for T1 and T2 (Fig 4). These sample points also had the highest bacterial loads (Fig 1). Fig 4 also shows a tendency for a higher relative abundance of Carnobacterium in leg samples collected in weeks 4 and 5, and a higher abundance of Yersiniaceae in samples collected in week 2. For the breast fillets there was a tendency for a higher relative abundance of Lactobacillus in samples collected in weeks 4 and 5 and Lactobacillales (possibly C. divergens) in weeks 1, 2 and 3.

Fourteen sOTUs were assigned to Acinetobacter and although the analysis did not result in species identification, the most abundant Acinetobacter species in the air chiller determined by colony identification was A. harbinensis. No Acinetobacter colonies were identified from end-of-shelf-life products, but the dominating Acinetobacter species on MAP plates from fresh breast fillets and legs were also A. harbinensis. Four different sOTUs were assigned to Carnobacterium, however only one had a high abundance across samples. This sOTU was indicated to be C. maltaromaticum according to a BLAST search against the NCBI nt database. Two sOTUs were assigned to Lactobacillales, also here only one sOTU was highly abundant across samples. According to a search in BLAST this sOTU matched Carnobacterium sp., C. divergens, Enterococcus and Vagococcus. With reference to the MALDI-TOF MS results, this sequence most likely represented C. divergens. For Pseudomonas 15 sOTUs were identified with some variation between the sample types.

Colony based bacterial composition of end-of-shelf-life products.

Colony based bacterial identification was performed to capture viable taxa that grow under storage, and to obtain species‑level IDs via MALDI‑TOF to complement 16S data. Due to having 16S amplicon results for end-of-shelf-life samples, colony based identification was only performed on a selection of samples (only week 4, M1 and M2). A total of 149 isolates were identified and the results showed a much less diverse microbiota than for fresh products, with only 7 bacterial taxa: Carnobacterium (66%), Lactobacillus (15%), Lactococcus (11%), Morganella (2%), Paucilactobacillus (2%), Stenotrophomonas (<1%) and Ewingella (<1%). The most abundant species identified (MALDI score ≥2.0) was C. maltaromaticum in both breast fillets (54%) and leg samples (64%), with the highest abundance in products stored at 4°C, followed by C. divergens (9% and 12% in breast fillets and legs, respectively). Other identified bacteria were Lactococcus piscium (11%), Lactobacillus curvatus (8%), Lactobacillus sakei (7%), Morganella psychrotolerans (2%), Ewingella americana (>1%), Lactococcus raffinolactis (>1%) and Stenotrophomonas rhizophila (>1%).

Differences in bacterial community.

Alpha- and beta diversity. There seemed to be a higher bacterial diversity in the air chiller compared to the end-of-shelf-life product, as well as a different microbiota (Fig 4). This was confirmed by measuring the diversity both within a sample group (alpha diversity) and between sample groups (beta diversity). The alpha diversity (Shannon entropy) was significantly higher in the air chiller compared to product samples (H = 49.9, q < 0.001) (S1 Fig in S1 File). As expected, leg samples (with skin) had a significantly higher diversity compared to fillet samples (H = 43.7, q < 0.001) (S1 Fig in S1 File). There was a covariance in the alpha diversity between weeks and sample times for the air chiller and end-of-shelf-life breast fillets and legs, with a higher bacterial diversity in the first weeks of sampling and higher diversity on Mondays compared to Thursdays (S1 Fig in S1 File). Alpha diversity was inversely related to TVC where samples with lower TVC tended to show higher within‑sample diversity.

The bacterial diversity between samples (beta diversity) was illustrated using PCoA analysis (Bray-Curtis metrics) (Fig 5).

Download:

Fig 5. Principal coordinate analysis (PCoA) biplot.

PCoA biplot based on Bray Curtis metrics depicting taxonomic differences between sample types and weeks. The samples are colored according to sample type: orange = air chiller; red = breast fillets; blue = legs, while different symbols indicate different weeks: week 1; week 2; week 3; week 4 and week 5. The five taxa with the highest impact on the distribution of the samples are shown as arrows: red = Lactobacillales; blue = Carnobacterium; orange = Lactobacillus; green = Yersiniaceae; and purple = Acinetobacter.

https://doi.org/10.1371/journal.pone.0344423.g005

As expected from the overall microbial composition, the PCoA analysis showed that the air chiller and product samples clustered separately defined by the bacterial composition (Fig 5). The product samples were also mostly separated based on sample parts (leg- or breast fillet samples). The variation between samples was highest for breast fillet samples, with a tendency of clustering based on weeks (weeks 1 and 2, and weeks 4 and 5). The five most important sOTUs to explain the variation in the PCoA plot were associated with Acinetobacter, Lactobacillales (BLAST associated with C. divergens), Lactobacillus, Carnobacterium (BLAST associated with C. maltaromaticum) and Yersiniaceae, where Acinetobacter was more abundant in the air chiller, and the others in end-of-shelf-life products (Fig 5). In general, the breast fillets had a higher abundance of Lactobacillales and Lactobacillus, while the leg samples had a higher abundance of Carnobacterium and Yersiniaceae.

PERMANOVA. To further investigate the effect of the sample categories (sample type (air chiller or product); product part (breast fillet or leg), week, sampling time and temperature) PERMANOVA analyses were performed. The results varied depending on the metrics used (see S1 Table in S1 File for details and p-values). As expected, the greatest variation in microbial composition (as indicated by the highest test statistics) was observed between sample types (air chiller and product samples), followed by sample type (breast fillet or leg). For sample type and product type, weighted UniFrac revealed the most pronounced differences, indicating substantial variation in microbial communities in terms of both phylogenetic relationships and relative abundance. Consistent with the patterns observed in alpha diversity, the microbiota in the air chiller varied significantly across weeks and sampling times. For end-of-shelf-life breast fillets there was a significant difference in the bacterial composition between weeks (Bray Curtis and unweighted Unifrac, p = 0.001) and sample time (p = 0.009), but not for storage temperature. For leg samples there was a significant difference in the bacterial composition between weeks (p = 0.001–0.009), sample time (Bray Curtis and unweighted Unifrac, p = 0.001) and storage temperature (Bray Curtis and weighted Unifrac, p = 0.011–0.013). To investigate the difference between sampling times, the Bray-Curtis metric was used in pairwise PERMANOVA (S2 Table in S1 File). For the air chiller there was a trend that samples collected on Mondays were different from samples collected on Thursdays. For the air chiller, breast fillets and legs, samples collected early Mondays (M1) were significantly different from samples collected early Thursdays (T1) (q < 0.05). For breast fillet and leg samples, there was also a significant difference between samples collected late Mondays (M2) and early Thursdays (T1) (q < 0.05). These consistent patterns across TVC, alpha diversity, and beta diversity suggest a strong covariance in microbial dynamics between the air chiller and chicken products, influenced by sampling time and processing conditions.

Fecal indicators and potential pathogens

Other than E. coli, few bacteria related to fecal indicators or pathogens were detected from agar plates, e.g., no Campylobacter, Listeria or Salmonella was detected among the isolates identified. Although the sequencing technique used in this study lacks phylogenetic resolution to detect pathogens, and relatedness based on partial 16S rRNA amplicon sequencing is only a weak indicator for the pathogenic potential of the identified bacteria, it is of interest to look at the families and genera that could include potential pathogens or indicators of fecal contamination. Sequences assigned to Escherichia-Shigella were identified in 19 air chiller samples (53%) and nine product samples (9%) with a relative abundance below 0.9% and 0.3%, respectively. No further sequence match was performed, but all Escherichia isolates identified from the air chiller were identified as E. coli. Sequences assigned to Campylobacter were identified in three air chiller samples, but no product samples. Salmonella and Listeria were not identified. Enterobacteriaceae was identified in 81% of leg samples and 26% of breast fillet samples, but only in five air chiller samples (14%).

Discussion

Understanding the variation in microbial status in poultry processing facilities and products is essential for improving quality and securing shelf-life. In this study, air chiller and product samples covaried in both microbial load and community composition across weeks and sampling times, indicating that environmental shifts in the air chiller influenced the downstream product microbiota. Across 18 production batches we observed: (i) higher late‑week bacterial loads in the air chiller and in fresh products (late Thursday > Monday), with a similar tendency at end-of-shelf-life; (ii) covariance in both bacterial load and community composition between the air chiller and products across weeks and sampling times; (iii) end‑of‑shelf‑life communities under high‑CO₂ MAP were dominated by CO₂‑tolerant taxa (e.g., Carnobacterium, Lactobacillales); and (iv) air‑chiller microbiota dominated by aerobic psychrotrophic genera (Acinetobacter, Psychrobacter, Pseudomonas).

No significant odor was noticed when opening the packages at the end of the shelf-life. While total viable counts (TVC) give a practical benchmark, commonly 7 log CFU/g for acceptability [37,38] and noticeable odor changes above 8 log CFU/g [39–41], spoilage is ultimately driven by which bacteria establish under a given temperature and packaging atmosphere [16]. Here, average bacterial loads at the end-of-shelf-life were 7.5 log CFU/g in leg samples and 6.3 log CFU/g in breast fillets. The relatively low TVC combined with the dominance of lactic acid bacteria could indicate that the products were not yet spoiled. Sensory evaluation by a professional panel would be required to conclude about the degree of spoilage.

The dominating microbiota of end-of-shelf-life leg products included Carnobacterium, Yersiniaceae and Lactobacillales, while breast fillet samples were dominated by Lactobacillus, Lactobacillales and Carnobacterium. This pattern aligns with CO₂‑tolerant LAB and Brochothrix/Carnobacterium often displacing strict aerobes under MAP, a shift repeatedly observed across poultry systems [11,16,38].

Culture-dependent results from the air chiller and from MAP-incubated plates of fresh product showed that the most abundant bacteria was Pseudomonas spp. Aerobic psychrotrophs such as Pseudomonas can be introduced from water, air, and equipment surfaces, and can proliferate in wall condensations and drippings during air chilling or air movement, providing plausible transfer routs to product [16,23]. Consequently, fresh chicken frequently carries a high relative abundance of post‑processing contaminants, and air‑stored poultry typically exhibits Pseudomonas dominance at spoilage [6,13,16,42].

By contrast, 16S amplicon sequencing showed that Acinetobacter was the predominant genus detected in the air chiller, followed by Psychrobacter and Pseudomonas. All three genera include aerobic, psychrotrophic, moisture‑tolerant taxa frequently detected on cold, humid industrial surfaces and in processing‑area air, where they commonly predominate during processing [7,39,43]. Acinetobacter spp. are commonly found in food and drinking water [44] and in food processing environments [45], particularly in humid fish processing environments [46]. Kunert-Filho et al., [47] reported Acinetobacter dominance on final carcasses in slaughterhouses in Brazil. The discrepancy between the abundance of Acinetobacter identified from isolates and by culture-independent sequencing can arise from growth‑rate biases on PCA and competitive overgrowth by Pseudomonas, leading to under‑representation of certain taxa in culture relative to sequencing. Similar divergences between culture‑dependent and ‑independent profiles have been reported by others [43].

The microbiota of the air chiller differed from the end-of-shelf-life products, and taxa characteristic for end-of-shelf-life products bacteria occurred at low abundance in the air chiller. However, incubating air chiller environmental swabs under product-like MAP enriched CO₂‑tolerant bacteria and indicators relevant to late shelf-life. Spoilage and indicator bacteria such as Carnobacterium and Escherichia were more abundant on MAP plates, suggesting that conventional culture conditions may fail to recover CO₂‑tolerant taxa unless incubation atmosphere mimic package gases. Carnobacterium species, especially C. maltaromaticum and C. divergens, are major spoilage bacteria in high CO₂ MAP [48], with contamination routes suggested to be from the abattoir environment [24]. By contrast, Lauritsen et al. [11], using amplicon sequencing, did not readily detect Carnobacterium sequences on carcasses or equipment despite its abundance in spoiled meat. In this study, C. maltaromaticum was cultivated from air chiller and fresh leg samples, and C. divergens from fresh breast fillets, plausibly reflecting skin‑associated transfer into the chiller by dripping carcasses. These observations are consistent with proposed contamination routes for psychrotrophic LAB in chilling areas. The end-of-shelf-life leg microbiota also showed a higher relative abundance of Carnobacterium than breast fillets, although species were not resolved.

It is well known that both pre- and post-slaughter factors influence the poultry microbiome [6,16] where the processing plant has been identified as the main source of post-chill carcass microbiome [8]. We have shown higher late‑week bacterial loads in the air chiller and in fresh products and the same tendency for end-of-shelf-life products, but minor within-day differences. Mechanistically, this is consistent with bacterial accumulation in the air‑chiller environment followed by cross contamination to carcasses. Vargas et al. [49] observed higher bacterial counts during the second shift, conversely, studies standardizing flock and packaging conditions found minor differences at retail [11]. No direct association with shelf-life was observed, but repeated covariation between air chiller and product communities indicates that environmental changes in the chiller can influence downstream product microbiota. Together with facility‑wide mapping that links processing‑area surfaces and equipment to product microbiota [23,24], our findings highlight the air chiller as a practical focus for monitoring, without implying causation.

Environmental microbiome monitoring as a strategy to improve food quality and safety is not a new concept [50] but there is still little experimental evidence. Our study adds new insight into the microbial variation in chicken products, showing that microbial patterns in the air chiller can influence both fresh and end-of-shelf-life products, even if not directly predictive of shelf-life. Despite the absence of a direct correlation with shelf-life, the air chiller exhibited notable covariance with product microbial status, suggesting its potential as an indicator of overall microbial quality in the processing environment. Implementable actions to improve product quality could include: i) mid-week sanitation or targeted interim cleaning of the air chiller tunnel to reduce accumulation; ii) condensation management (dehumidification, ventilation adjustments) to limit psychrotroph growth; and iii) targeted environmental microbiome monitoring of the air chiller surfaces using incubation in MAP like atmospheres to detect CO₂ tolerant spoilage organisms (e.g., Carnobacterium spp.).

Although the study was limited to a single processing plant and to sampling the air chiller, our observations are consistent with prior work identifying the processing environment, particularly the air chilling zones, as a reservoir for spoilage bacteria that can transfer to products. Since we only sampled the air chiller, we cannot exclude contributions from other sites upstream or downstream of the air chiller. Ongoing work includes environmental sampling from bird transport cages through to packed products in two processing facilities (Norway and Portugal) and long-term shelf-life studies using different modified atmosphere packaging (MAP) conditions. The aim is to enable shelf-life prediction by integrating environmental metadata, product microbiomes and operational covariates.

Supporting information

S1 File. Additional figures and tables.

Includes S1 Fig (Alpha diversity (Shannon index) across sample categories), S1 Table (Group significance (PERMANOVA) for the different sample categories and distance metrics), and S2 Table (Pairwise PERMANOVA on Bray Curtis dissimilarities comparing sampling times within each processing/storage subset).

https://doi.org/10.1371/journal.pone.0344423.s001

(DOCX)

Acknowledgments

We sincerely thank the poultry processing company for contributing to this study.

References

1. Yimenu SM, Koo J, Kim BS, Kim JH, Kim JY. Freshness-based real-time shelf-life estimation of packaged chicken meat under dynamic storage conditions. Poult Sci. 2019;98(12):6921–30. pmid:31504907
- View Article
- PubMed/NCBI
- Google Scholar
2. Lytou AE, Renieri CT, Doulgeraki AI, Nychas G-JE, Panagou EZ. Assessment of the microbiological quality and safety of marinated chicken products from Greek retail outlets. Int J Food Microbiol. 2020;320:108506. pmid:31981852
- View Article
- PubMed/NCBI
- Google Scholar
3. Rouger A, Tresse O, Zagorec M. Bacterial contaminants of poultry meat: sources, species, and dynamics. Microorganisms. 2017;5(3):50. pmid:28841156
- View Article
- PubMed/NCBI
- Google Scholar
4. EU. European Commission: Directorate-General for Agriculture and Rural Development. EU agricultural outlook 2025-2035. 2025. Available from: https://doi.org/10.2762/4491268
5. Sequino G, Cobo-Diaz JF, Valentino V, Tassou C, Volpe S, Torrieri E, et al. Environmental microbiome mapping in poultry processing chain and assessment of microbial dynamics in response to different storage conditions. Food Microbiol. 2025;128:104734. pmid:39952751
- View Article
- PubMed/NCBI
- Google Scholar
6. Chen SH, Fegan N, Kocharunchitt C, Bowman JP, Duffy LL. Impact of poultry processing operating parameters on bacterial transmission and persistence on chicken carcasses and their shelf life. Appl Environ Microbiol. 2020;86(12):e00594–20. pmid:32276979
- View Article
- PubMed/NCBI
- Google Scholar
7. Park J, Bae D, Kim SA. Microbial trace investigation throughout the entire chicken supply chain based on metagenomic high-throughput sequencing. Food Res Int. 2023;169:112775. pmid:37254378
- View Article
- PubMed/NCBI
- Google Scholar
8. Gaonkar PP, Santana-Pereira ALR, Golden R, Lambert A, Higgins C, Adhikari Y, et al. Microbiome and resistome dynamics in different stages of commercial broiler production with restricted antimicrobial use. BMC Microbiol. 2026;26(1):126. pmid:41507780
- View Article
- PubMed/NCBI
- Google Scholar
9. Rukchon C, Nopwinyuwong A, Trevanich S, Jinkarn T, Suppakul P. Development of a food spoilage indicator for monitoring freshness of skinless chicken breast. Talanta. 2014;130:547–54. pmid:25159445
- View Article
- PubMed/NCBI
- Google Scholar
10. Gram L, Ravn L, Rasch M, Bruhn JB, Christensen AB, Givskov M. Food spoilage--interactions between food spoilage bacteria. Int J Food Microbiol. 2002;78(1–2):79–97. pmid:12222639
- View Article
- PubMed/NCBI
- Google Scholar
11. Lauritsen CV, Kjeldgaard J, Ingmer H, Bisgaard M, Christensen H. Microbiota encompassing putative spoilage bacteria in retail packaged broiler meat and commercial broiler abattoir. Int J Food Microbiol. 2019;300:14–21. pmid:30991234
- View Article
- PubMed/NCBI
- Google Scholar
12. Herbert U, Albrecht A, Kreyenschmidt J. Definition of predictor variables for MAP poultry filets stored under different temperature conditions. Poult Sci. 2015;94(3):424–32. pmid:25638474
- View Article
- PubMed/NCBI
- Google Scholar
13. Morales PA, Aguirre JS, Troncoso MR, Figueroa GO. Phenotypic and genotypic characterization of Pseudomonas spp. present in spoiled poultry fillets sold in retail settings. LWT-Food Sci Technol. 2016;73:609–14.
- View Article
- Google Scholar
14. Lee HS, Kwon M, Heo S, Kim MG, Kim G-B. Characterization of the biodiversity of the spoilage microbiota in chicken meat using next generation sequencing and culture dependent approach. Korean J Food Sci Anim Resour. 2017;37(4):535–41. pmid:28943766
- View Article
- PubMed/NCBI
- Google Scholar
15. Hilgarth M, Fuertes S, Ehrmann M, Vogel RF. Photobacterium carnosum sp. nov., isolated from spoiled modified atmosphere packaged poultry meat. Syst Appl Microbiol. 2018;41(1):44–50. pmid:29279139
- View Article
- PubMed/NCBI
- Google Scholar
16. Marmion M, Ferone MT, Whyte P, Scannell AGM. The changing microbiome of poultry meat; from farm to fridge. Food Microbiol. 2021;99:103823. pmid:34119108
- View Article
- PubMed/NCBI
- Google Scholar
17. Holck AL, Pettersen MK, Moen MH, Sørheim O. Prolonged shelf life and reduced drip loss of chicken filets by the use of carbon dioxide emitters and modified atmosphere packaging. J Food Prot. 2014;77(7):1133–41. pmid:24988019
- View Article
- PubMed/NCBI
- Google Scholar
18. Kandeepan G, Tahseen A. Modified atmosphere packaging (MAP) of meat and meat products: a review. J Package Technol Res. 2022;6(3):137–48.
- View Article
- Google Scholar
19. Hansen AÅ, Langsrud S, Carlehög M, Haugen J-E, Moen B. CO₂ packaging increases shelf life through reduction of off-odor production by CO₂ tolerant bacteria in addition to growth inhibition of the spoilage bacteriota. Food Control. 2023;144:109390.
- View Article
- Google Scholar
20. Shao L, Chen S, Wang H, Zhang J, Xu X, Wang H. Advances in understanding the predominance, phenotypes, and mechanisms of bacteria related to meat spoilage. Trends Food Sci Technol. 2021;118:822–32.
- View Article
- Google Scholar
21. Snyder AB, Martin N, Wiedmann M. Microbial food spoilage: impact, causative agents and control strategies. Nat Rev Microbiol. 2024;22(9):528–42. pmid:38570695
- View Article
- PubMed/NCBI
- Google Scholar
22. Mead GC, Allen VM, Burton CH, Corry JE. Microbial cross-contamination during air chilling of poultry. Br Poult Sci. 2000;41(2):158–62. pmid:10890210
- View Article
- PubMed/NCBI
- Google Scholar
23. Chen SH, Fegan N, Kocharunchitt C, Bowman JP, Duffy LL. Changes of the bacterial community diversity on chicken carcasses through an Australian poultry processing line. Food Microbiol. 2020;86:103350. pmid:31703868
- View Article
- PubMed/NCBI
- Google Scholar
24. Vihavainen E, Lundström H-S, Susiluoto T, Koort J, Paulin L, Auvinen P, et al. Role of broiler carcasses and processing plant air in contamination of modified-atmosphere-packaged broiler products with psychrotrophic lactic acid bacteria. Appl Environ Microbiol. 2007;73(4):1136–45. pmid:17142357
- View Article
- PubMed/NCBI
- Google Scholar
25. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J. 2012;6(8):1621–4. pmid:22402401
- View Article
- PubMed/NCBI
- Google Scholar
26. Møretrø T, Moen B, Almli VL, Teixeira P, Ferreira VB, Åsli AW, et al. Dishwashing sponges and brushes: Consumer practices and bacterial growth and survival. Int J Food Microbiol. 2021;337:108928. pmid:33152572
- View Article
- PubMed/NCBI
- Google Scholar
27. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37(8):852–7. pmid:31341288
- View Article
- PubMed/NCBI
- Google Scholar
28. Bokulich NA, Subramanian S, Faith JJ, Gevers D, Gordon JI, Knight R, et al. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods. 2013;10(1):57–9. pmid:23202435
- View Article
- PubMed/NCBI
- Google Scholar
29. Amir A, McDonald D, Navas-Molina JA, Kopylova E, Morton JT, Xu ZZ. Deblur rapidly resolves single-nucleotide community sequence patterns. mSystems. 2017;2(2). pmid:WOS:000408195200009
- View Article
- PubMed/NCBI
- Google Scholar
30. McDonald D, Clemente JC, Kuczynski J, Rideout JR, Stombaugh J, Wendel D, et al. The biological observation matrix (BIOM) format or: how I learned to stop worrying and love the ome-ome. Gigascience. 2012;1(1):7. pmid:23587224
- View Article
- PubMed/NCBI
- Google Scholar
31. Rognes T, Flouri T, Nichols B, Quince C, Mahé F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016;4:e2584. pmid:27781170
- View Article
- PubMed/NCBI
- Google Scholar
32. Hamady M, Walker JJ, Harris JK, Gold NJ, Knight R. Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nat Methods. 2008;5(3):235–7. pmid:18264105
- View Article
- PubMed/NCBI
- Google Scholar
33. Hamady M, Knight R. Microbial community profiling for human microbiome projects: Tools, techniques, and challenges. Genome Res. 2009;19(7):1141–52. pmid:19383763
- View Article
- PubMed/NCBI
- Google Scholar
34. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80. pmid:23329690
- View Article
- PubMed/NCBI
- Google Scholar
35. Anderson MJ, Walsh DCI. PERMANOVA, ANOSIM, and the Mantel test in the face of heterogeneous dispersions: What null hypothesis are you testing? Ecol Monogr. 2013;83(4):557–74.
- View Article
- Google Scholar
36. Anderson MJ. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 2001;26(1):32–46.
- View Article
- Google Scholar
37. Zhang QQ, Han YQ, Cao JX, Xu XL, Zhou GH, Zhang WY. The spoilage of air-packaged broiler meat during storage at normal and fluctuating storage temperatures. Poult Sci. 2012;91(1):208–14. pmid:22184446
- View Article
- PubMed/NCBI
- Google Scholar
38. Höll L, Behr J, Vogel RF. Identification and growth dynamics of meat spoilage microorganisms in modified atmosphere packaged poultry meat by MALDI-TOF MS. Food Microbiol. 2016;60:84–91. pmid:27554149
- View Article
- PubMed/NCBI
- Google Scholar
39. Rouger A, Moriceau N, Prévost H, Remenant B, Zagorec M. Diversity of bacterial communities in French chicken cuts stored under modified atmosphere packaging. Food Microbiol. 2018;70:7–16. pmid:29173642
- View Article
- PubMed/NCBI
- Google Scholar
40. Wang J, Zhuang H, Hinton A Jr, Zhang J. Influence of in-package cold plasma treatment on microbiological shelf life and appearance of fresh chicken breast fillets. Food Microbiol. 2016;60:142–6. pmid:27554156
- View Article
- PubMed/NCBI
- Google Scholar
41. Naveena BM, Khansole PS, Shashi Kumar M, Krishnaiah N, Kulkarni VV, Deepak SJ. Effect of sous vide processing on physicochemical, ultrastructural, microbial and sensory changes in vacuum packaged chicken sausages. Food Sci Technol Int. 2017;23(1):75–85. pmid:27386881
- View Article
- PubMed/NCBI
- Google Scholar
42. Li S, Mann DA, Zhang S, Qi Y, Meinersmann RJ, Deng X. Microbiome-informed food safety and quality: longitudinal consistency and cross-sectional distinctiveness of retail chicken breast microbiomes. mSystems. 2020;5(5):e00589–20. pmid:32900871
- View Article
- PubMed/NCBI
- Google Scholar
43. Dourou D, Spyrelli ED, Doulgeraki AI, Argyri AA, Grounta A, Nychas G-JE, et al. Microbiota of chicken breast and thigh fillets stored under different refrigeration temperatures assessed by next-generation sequencing. Foods. 2021;10(4):765. pmid:33916748
- View Article
- PubMed/NCBI
- Google Scholar
44. Carvalheira A, Silva J, Teixeira P. Acinetobacter spp. in food and drinking water - A review. Food Microbiol. 2021;95:103675. pmid:33397609
- View Article
- PubMed/NCBI
- Google Scholar
45. Møretrø T, Langsrud S, Heir E. Bacteria on meat abattoir process surfaces after sanitation: characterisation of survival properties of Listeria monocytogenes and the commensal bacterial flora. Adv Microbiol. 2013;03(03):255–64.
- View Article
- Google Scholar
46. Møretrø T, Langsrud S. Residential bacteria on surfaces in the food industry and their implications for food safety and quality. Compr Rev Food Sci Food Saf. 2017;16(5):1022–41. pmid:33371605
- View Article
- PubMed/NCBI
- Google Scholar
47. Kunert-Filho HC, Furian TQ, Sesterhenn R, Chitolina GZ, Willsmann DE, Borges KA, et al. Bacterial community identification in poultry carcasses using high-throughput next generation sequencing. Int J Food Microbiol. 2022;364:109533. pmid:35066330
- View Article
- PubMed/NCBI
- Google Scholar
48. Leisner JJ, Laursen BG, Prévost H, Drider D, Dalgaard P. Carnobacterium: positive and negative effects in the environment and in foods. FEMS Microbiol Rev. 2007;31(5):592–613. pmid:17696886
- View Article
- PubMed/NCBI
- Google Scholar
49. Vargas DA, De Villena JF, Larios V, Bueno López R, Chávez-Velado DR, Casas DE, et al. Data-mining poultry processing bio-mapping counts of pathogens and indicator organisms for food safety management decision making. Foods. 2023;12(4):898. pmid:36832973
- View Article
- PubMed/NCBI
- Google Scholar
50. De Filippis F, Valentino V, Alvarez-Ordóñez A, Cotter PD, Ercolini D. Environmental microbiome mapping as a strategy to improve quality and safety in the food industry. Curr Opin Food Sci. 2021;38:168–76.
- View Article
- Google Scholar

[ref1] 1. Yimenu SM, Koo J, Kim BS, Kim JH, Kim JY. Freshness-based real-time shelf-life estimation of packaged chicken meat under dynamic storage conditions. Poult Sci. 2019;98(12):6921–30. pmid:31504907
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Lytou AE, Renieri CT, Doulgeraki AI, Nychas G-JE, Panagou EZ. Assessment of the microbiological quality and safety of marinated chicken products from Greek retail outlets. Int J Food Microbiol. 2020;320:108506. pmid:31981852
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Rouger A, Tresse O, Zagorec M. Bacterial contaminants of poultry meat: sources, species, and dynamics. Microorganisms. 2017;5(3):50. pmid:28841156
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. EU. European Commission: Directorate-General for Agriculture and Rural Development. EU agricultural outlook 2025-2035. 2025. Available from: https://doi.org/10.2762/4491268

[ref5] 5. Sequino G, Cobo-Diaz JF, Valentino V, Tassou C, Volpe S, Torrieri E, et al. Environmental microbiome mapping in poultry processing chain and assessment of microbial dynamics in response to different storage conditions. Food Microbiol. 2025;128:104734. pmid:39952751
View Article
PubMed/NCBI
Google Scholar

[15] View Article

[16] PubMed/NCBI

[17] Google Scholar

[ref6] 6. Chen SH, Fegan N, Kocharunchitt C, Bowman JP, Duffy LL. Impact of poultry processing operating parameters on bacterial transmission and persistence on chicken carcasses and their shelf life. Appl Environ Microbiol. 2020;86(12):e00594–20. pmid:32276979
View Article
PubMed/NCBI
Google Scholar

[19] View Article

[20] PubMed/NCBI

[21] Google Scholar

[ref7] 7. Park J, Bae D, Kim SA. Microbial trace investigation throughout the entire chicken supply chain based on metagenomic high-throughput sequencing. Food Res Int. 2023;169:112775. pmid:37254378
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref8] 8. Gaonkar PP, Santana-Pereira ALR, Golden R, Lambert A, Higgins C, Adhikari Y, et al. Microbiome and resistome dynamics in different stages of commercial broiler production with restricted antimicrobial use. BMC Microbiol. 2026;26(1):126. pmid:41507780
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref9] 9. Rukchon C, Nopwinyuwong A, Trevanich S, Jinkarn T, Suppakul P. Development of a food spoilage indicator for monitoring freshness of skinless chicken breast. Talanta. 2014;130:547–54. pmid:25159445
View Article
PubMed/NCBI
Google Scholar

[31] View Article

[32] PubMed/NCBI

[33] Google Scholar

[ref10] 10. Gram L, Ravn L, Rasch M, Bruhn JB, Christensen AB, Givskov M. Food spoilage--interactions between food spoilage bacteria. Int J Food Microbiol. 2002;78(1–2):79–97. pmid:12222639
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref11] 11. Lauritsen CV, Kjeldgaard J, Ingmer H, Bisgaard M, Christensen H. Microbiota encompassing putative spoilage bacteria in retail packaged broiler meat and commercial broiler abattoir. Int J Food Microbiol. 2019;300:14–21. pmid:30991234
View Article
PubMed/NCBI
Google Scholar

[39] View Article

[40] PubMed/NCBI

[41] Google Scholar

[ref12] 12. Herbert U, Albrecht A, Kreyenschmidt J. Definition of predictor variables for MAP poultry filets stored under different temperature conditions. Poult Sci. 2015;94(3):424–32. pmid:25638474
View Article
PubMed/NCBI
Google Scholar

[43] View Article

[44] PubMed/NCBI

[45] Google Scholar

[ref13] 13. Morales PA, Aguirre JS, Troncoso MR, Figueroa GO. Phenotypic and genotypic characterization of Pseudomonas spp. present in spoiled poultry fillets sold in retail settings. LWT-Food Sci Technol. 2016;73:609–14.
View Article
Google Scholar

[47] View Article

[48] Google Scholar

[ref14] 14. Lee HS, Kwon M, Heo S, Kim MG, Kim G-B. Characterization of the biodiversity of the spoilage microbiota in chicken meat using next generation sequencing and culture dependent approach. Korean J Food Sci Anim Resour. 2017;37(4):535–41. pmid:28943766
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref15] 15. Hilgarth M, Fuertes S, Ehrmann M, Vogel RF. Photobacterium carnosum sp. nov., isolated from spoiled modified atmosphere packaged poultry meat. Syst Appl Microbiol. 2018;41(1):44–50. pmid:29279139
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref16] 16. Marmion M, Ferone MT, Whyte P, Scannell AGM. The changing microbiome of poultry meat; from farm to fridge. Food Microbiol. 2021;99:103823. pmid:34119108
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref17] 17. Holck AL, Pettersen MK, Moen MH, Sørheim O. Prolonged shelf life and reduced drip loss of chicken filets by the use of carbon dioxide emitters and modified atmosphere packaging. J Food Prot. 2014;77(7):1133–41. pmid:24988019
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref18] 18. Kandeepan G, Tahseen A. Modified atmosphere packaging (MAP) of meat and meat products: a review. J Package Technol Res. 2022;6(3):137–48.
View Article
Google Scholar

[66] View Article

[67] Google Scholar

[ref19] 19. Hansen AÅ, Langsrud S, Carlehög M, Haugen J-E, Moen B. CO₂ packaging increases shelf life through reduction of off-odor production by CO₂ tolerant bacteria in addition to growth inhibition of the spoilage bacteriota. Food Control. 2023;144:109390.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref20] 20. Shao L, Chen S, Wang H, Zhang J, Xu X, Wang H. Advances in understanding the predominance, phenotypes, and mechanisms of bacteria related to meat spoilage. Trends Food Sci Technol. 2021;118:822–32.
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref21] 21. Snyder AB, Martin N, Wiedmann M. Microbial food spoilage: impact, causative agents and control strategies. Nat Rev Microbiol. 2024;22(9):528–42. pmid:38570695
View Article
PubMed/NCBI
Google Scholar

[75] View Article

[76] PubMed/NCBI

[77] Google Scholar

[ref22] 22. Mead GC, Allen VM, Burton CH, Corry JE. Microbial cross-contamination during air chilling of poultry. Br Poult Sci. 2000;41(2):158–62. pmid:10890210
View Article
PubMed/NCBI
Google Scholar

[79] View Article

[80] PubMed/NCBI

[81] Google Scholar

[ref23] 23. Chen SH, Fegan N, Kocharunchitt C, Bowman JP, Duffy LL. Changes of the bacterial community diversity on chicken carcasses through an Australian poultry processing line. Food Microbiol. 2020;86:103350. pmid:31703868
View Article
PubMed/NCBI
Google Scholar

[83] View Article

[84] PubMed/NCBI

[85] Google Scholar

[ref24] 24. Vihavainen E, Lundström H-S, Susiluoto T, Koort J, Paulin L, Auvinen P, et al. Role of broiler carcasses and processing plant air in contamination of modified-atmosphere-packaged broiler products with psychrotrophic lactic acid bacteria. Appl Environ Microbiol. 2007;73(4):1136–45. pmid:17142357
View Article
PubMed/NCBI
Google Scholar

[87] View Article

[88] PubMed/NCBI

[89] Google Scholar

[ref25] 25. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J. 2012;6(8):1621–4. pmid:22402401
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref26] 26. Møretrø T, Moen B, Almli VL, Teixeira P, Ferreira VB, Åsli AW, et al. Dishwashing sponges and brushes: Consumer practices and bacterial growth and survival. Int J Food Microbiol. 2021;337:108928. pmid:33152572
View Article
PubMed/NCBI
Google Scholar

[95] View Article

[96] PubMed/NCBI

[97] Google Scholar

[ref27] 27. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37(8):852–7. pmid:31341288
View Article
PubMed/NCBI
Google Scholar

[99] View Article

[100] PubMed/NCBI

[101] Google Scholar

[ref28] 28. Bokulich NA, Subramanian S, Faith JJ, Gevers D, Gordon JI, Knight R, et al. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods. 2013;10(1):57–9. pmid:23202435
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref29] 29. Amir A, McDonald D, Navas-Molina JA, Kopylova E, Morton JT, Xu ZZ. Deblur rapidly resolves single-nucleotide community sequence patterns. mSystems. 2017;2(2). pmid:WOS:000408195200009
View Article
PubMed/NCBI
Google Scholar

[107] View Article

[108] PubMed/NCBI

[109] Google Scholar

[ref30] 30. McDonald D, Clemente JC, Kuczynski J, Rideout JR, Stombaugh J, Wendel D, et al. The biological observation matrix (BIOM) format or: how I learned to stop worrying and love the ome-ome. Gigascience. 2012;1(1):7. pmid:23587224
View Article
PubMed/NCBI
Google Scholar

[111] View Article

[112] PubMed/NCBI

[113] Google Scholar

[ref31] 31. Rognes T, Flouri T, Nichols B, Quince C, Mahé F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016;4:e2584. pmid:27781170
View Article
PubMed/NCBI
Google Scholar

[115] View Article

[116] PubMed/NCBI

[117] Google Scholar

[ref32] 32. Hamady M, Walker JJ, Harris JK, Gold NJ, Knight R. Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nat Methods. 2008;5(3):235–7. pmid:18264105
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref33] 33. Hamady M, Knight R. Microbial community profiling for human microbiome projects: Tools, techniques, and challenges. Genome Res. 2009;19(7):1141–52. pmid:19383763
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref34] 34. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80. pmid:23329690
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref35] 35. Anderson MJ, Walsh DCI. PERMANOVA, ANOSIM, and the Mantel test in the face of heterogeneous dispersions: What null hypothesis are you testing? Ecol Monogr. 2013;83(4):557–74.
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref36] 36. Anderson MJ. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 2001;26(1):32–46.
View Article
Google Scholar

[134] View Article

[135] Google Scholar

[ref37] 37. Zhang QQ, Han YQ, Cao JX, Xu XL, Zhou GH, Zhang WY. The spoilage of air-packaged broiler meat during storage at normal and fluctuating storage temperatures. Poult Sci. 2012;91(1):208–14. pmid:22184446
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref38] 38. Höll L, Behr J, Vogel RF. Identification and growth dynamics of meat spoilage microorganisms in modified atmosphere packaged poultry meat by MALDI-TOF MS. Food Microbiol. 2016;60:84–91. pmid:27554149
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref39] 39. Rouger A, Moriceau N, Prévost H, Remenant B, Zagorec M. Diversity of bacterial communities in French chicken cuts stored under modified atmosphere packaging. Food Microbiol. 2018;70:7–16. pmid:29173642
View Article
PubMed/NCBI
Google Scholar

[145] View Article

[146] PubMed/NCBI

[147] Google Scholar

[ref40] 40. Wang J, Zhuang H, Hinton A Jr, Zhang J. Influence of in-package cold plasma treatment on microbiological shelf life and appearance of fresh chicken breast fillets. Food Microbiol. 2016;60:142–6. pmid:27554156
View Article
PubMed/NCBI
Google Scholar

[149] View Article

[150] PubMed/NCBI

[151] Google Scholar

[ref41] 41. Naveena BM, Khansole PS, Shashi Kumar M, Krishnaiah N, Kulkarni VV, Deepak SJ. Effect of sous vide processing on physicochemical, ultrastructural, microbial and sensory changes in vacuum packaged chicken sausages. Food Sci Technol Int. 2017;23(1):75–85. pmid:27386881
View Article
PubMed/NCBI
Google Scholar

[153] View Article

[154] PubMed/NCBI

[155] Google Scholar

[ref42] 42. Li S, Mann DA, Zhang S, Qi Y, Meinersmann RJ, Deng X. Microbiome-informed food safety and quality: longitudinal consistency and cross-sectional distinctiveness of retail chicken breast microbiomes. mSystems. 2020;5(5):e00589–20. pmid:32900871
View Article
PubMed/NCBI
Google Scholar

[157] View Article

[158] PubMed/NCBI

[159] Google Scholar

[ref43] 43. Dourou D, Spyrelli ED, Doulgeraki AI, Argyri AA, Grounta A, Nychas G-JE, et al. Microbiota of chicken breast and thigh fillets stored under different refrigeration temperatures assessed by next-generation sequencing. Foods. 2021;10(4):765. pmid:33916748
View Article
PubMed/NCBI
Google Scholar

[161] View Article

[162] PubMed/NCBI

[163] Google Scholar

[ref44] 44. Carvalheira A, Silva J, Teixeira P. Acinetobacter spp. in food and drinking water - A review. Food Microbiol. 2021;95:103675. pmid:33397609
View Article
PubMed/NCBI
Google Scholar

[165] View Article

[166] PubMed/NCBI

[167] Google Scholar

[ref45] 45. Møretrø T, Langsrud S, Heir E. Bacteria on meat abattoir process surfaces after sanitation: characterisation of survival properties of Listeria monocytogenes and the commensal bacterial flora. Adv Microbiol. 2013;03(03):255–64.
View Article
Google Scholar

[169] View Article

[170] Google Scholar

[ref46] 46. Møretrø T, Langsrud S. Residential bacteria on surfaces in the food industry and their implications for food safety and quality. Compr Rev Food Sci Food Saf. 2017;16(5):1022–41. pmid:33371605
View Article
PubMed/NCBI
Google Scholar

[172] View Article

[173] PubMed/NCBI

[174] Google Scholar

[ref47] 47. Kunert-Filho HC, Furian TQ, Sesterhenn R, Chitolina GZ, Willsmann DE, Borges KA, et al. Bacterial community identification in poultry carcasses using high-throughput next generation sequencing. Int J Food Microbiol. 2022;364:109533. pmid:35066330
View Article
PubMed/NCBI
Google Scholar

[176] View Article

[177] PubMed/NCBI

[178] Google Scholar

[ref48] 48. Leisner JJ, Laursen BG, Prévost H, Drider D, Dalgaard P. Carnobacterium: positive and negative effects in the environment and in foods. FEMS Microbiol Rev. 2007;31(5):592–613. pmid:17696886
View Article
PubMed/NCBI
Google Scholar

[180] View Article

[181] PubMed/NCBI

[182] Google Scholar

[ref49] 49. Vargas DA, De Villena JF, Larios V, Bueno López R, Chávez-Velado DR, Casas DE, et al. Data-mining poultry processing bio-mapping counts of pathogens and indicator organisms for food safety management decision making. Foods. 2023;12(4):898. pmid:36832973
View Article
PubMed/NCBI
Google Scholar

[184] View Article

[185] PubMed/NCBI

[186] Google Scholar

[ref50] 50. De Filippis F, Valentino V, Alvarez-Ordóñez A, Cotter PD, Ercolini D. Environmental microbiome mapping as a strategy to improve quality and safety in the food industry. Curr Opin Food Sci. 2021;38:168–76.
View Article
Google Scholar

[188] View Article

[189] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Experimental design

Sampling and microbial analysis

Identification of colonies

DNA extraction and Illumina partial 16S rRNA amplicon sequencing

Sequence analysis

Gas analysis

Statistical analysis

Results

Bacterial load

Accumulation of bacteria during the week in the air chiller.

Variation in the bacterial load of chicken products.

Microbial diversity in the air chiller and fresh products (culture dependent)

Microbial diversity in the air chiller and end-of-shelf-life products (culture independent)

Colony based bacterial composition of end-of-shelf-life products.

Differences in bacterial community.

Fecal indicators and potential pathogens

Discussion

Supporting information

S1 File. Additional figures and tables.

Acknowledgments

References