Characterizing the spoilage microbiota of Belgian market fresh-cut iceberg lettuce in controlled atmosphere packaging

Characterizing the spoilage microbiota of Belgian market fresh-cut iceberg lettuce in controlled atmosphere packaging Irene Ortega-Sanz * , Soma Ishihara, Frank Devlieghere , Andreja Rajkovic Department of Food Technology, Safety and Health, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, 9000 Ghent, Belgium ARTICLE INFO Keywords: Microbiota Bacterial population dynamics Microbial spoilage Microbial biomarkers Lactococcus Culture-based 16S rRNA gene sequencing ABSTRACT Fresh-cut iceberg lettuce is gaining popularity for its convenience, but it is highly susceptible to microbial spoilage. This study aimed to investigate the microbial community dynamics in commercially available fresh-cut iceberg lettuce packaged under controlled atmosphere in Belgium during storage at 7 ◦C for up to 12 days. To achieve a comprehensive understanding of the spoilage microorganisms and their interactions, a culturecomplemented metataxonomic approach was conducted, including plating on non-selective and selective culture media for enumeration of total psychrotrophic counts (TPC), enterobacteria, Pseudomonas spp., psychrotrophic lactic acid bacteria (LAB) and yeasts. Throughout storage, TPC remained the predominant group, followed by Pseudomonas spp. until day 2, and by enterobacteria thereafter. The initial oxygen present in the commercial lettuce bags (1.5–2.0 %) was rapidly consumed within the first 24 h of storage (<0.1 %), resulting in a decrease in the dominance of Pseudomonas spp. (38.9 %) since then. By the end of storage, Pseudomonas spp. represented a minor group (<2 %), yet retained the capacity to grow. In contrast, Lactococcus spp. increased in abundance as CO 2 rose (1.1–25.8 %), eventually dominating the bacterial community (>44 %) together with Serratia (21 %) and Rahnella (9.5–11 %) species. In fact, LAB showed the highest increase in population throughout storage, with levels rising from ~1.8 log on day 0 to ~7.5 log colony forming units per gram on day 12. These findings offer new insights into the deterioration mechanisms of fresh-cut lettuce stored under controlled atmosphere packaging, providing a basis for developing new strategies to prevent lettuce spoilage and thereby help the industry and retailers reduce food waste and associated economic losses. 1. Introduction Driven by their convenience, diverse culinary offerings, minimal preparation requirements, and substantial time savings, ready-to-eat (RTE) food products have gained considerable market share in recent years. Parallel to this trend, consumer demand for natural and healthoriented options has increased, leading to a marked rise in the consumption of fresh-cut RTE vegetables (Bhatia et al., 2024). Iceberg lettuce, in particular, is among the most widely consumed leafy greens, especially during the summer months in Belgium (Colruyt Group, 2025; FAO, 2023). However, preserving the sensory quality and freshness of these products remains a challenge due to their inherently short shelf life, which contributes to substantial food waste and associated economic losses. The Food and Agriculture Organization of the United Nations (FAO) estimates that up to 50 % of fruits and vegetables are lost post-harvest (Gustavsson et al., 2011). The development of enzymatic browning, loss in texture qualities, production of off-odors, and microbiological spoilage are factors causing lettuce deterioration that contribute to customer rejection (Lorente-Mento et al., 2022; Peng & Simko, 2023; Sun et al., 2022). Furthermore, fresh-cut preparations causes substantial wounding to vegetables that significantly shortens their shelf life compared to raw unprocessed products due to increased respiration rates and ethylene production (Choi et al., 2021). To prolong their shelf-life, fresh-cut leafy vegetables are packaged under equilibrium modified atmosphere (EMAP), which is combined with low storage temperatures and, as such, helps preserve their quality and minimizes spoilage (Alemu & Oanh, 2025). In this context, Oliveira et al. (2015) recommended storing fresh-cut iceberg lettuce under low oxygen conditions (0.5–3 %) combined with elevated carbon dioxide levels (10–15 %) and temperatures between 0 ◦C and 5 ◦C. This reduced oxygen atmosphere is applied to limit enzymatic browning, that constitutes the main cause of consumer rejection (Ioannidis et al., 2018). Nonetheless, MAP-stored lettuce under these conditions still remain susceptible to microbial spoilage, primarily caused by bacterial groups including * Corresponding author. E-mail addresses: [email protected], [email protected] (I. Ortega-Sanz). Contents lists available at ScienceDirect Food Research International journal homepage: www.elsevier.com/locate/foodres https://doi.org/10.1016/j.foodres.2025.117694 Received 25 August 2025; Received in revised form 2 October 2025; Accepted 25 October 2025 Food Research International 222 (2025) 117694 Available online 26 October 2025 0963-9969/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). Pseudomonas species like Pseudomonas fluorescens, enterobacteria such as Pantoea agglomerans, and lactic acid bacteria (LAB) like Leuconostoc mesenteroides and Lactococcus spp. (Alegbeleye et al., 2025; De Bock et al., 2024; Ioannidis et al., 2018; Manthou et al., 2022). Under aerobiosis, Pseudomonas species and enterobacteria produce pectinolytic enzymes that disrupt plant tissue resulting in browning and discoloration, while LAB are known for the production of volatile organic compounds (VOCs) that cause unpleasant off-odors under anaerobic conditions (Ragaert et al., 2007). Additionally, yeasts and psychrotrophic strains of LAB have been associated with cases of spoilage in minimally processed vegetable salads, particularly in the Belgian market (Pothakos et al., 2014; Ragaert et al., 2007). However, the intricate changes in the microbial profile of RTE iceberg lettuce during MAP storage is still not fully understood. Next Generation Sequencing (NGS) has become a powerful tool to explore microbial communities involved in a wide variety of ecosystems, making it indispensable for food safety surveillance on RTE vegetable salad products (Alegbeleye et al., 2025; Peng et al., 2025).Through methodologies such as amplicon and shotgun sequencing, researchers can gain deeper insights into the diversity and composition of the microbiota present on vegetables, and their implications for shelf life and food safety. These approaches not only facilitate the identification of known spoilage microorganisms and foodborne pathogens, but also reveal rare or non-culturable microbes that are often missed by conventional culturing methods (Cao et al., 2017; Delikanli-Kiyak et al., 2025; Mira Miralles et al., 2019). Targeting specific genetic markers, such as the 16S rRNA gene for bacteria, provides a cost-effective means to make taxonomic classification of microbial communities on fresh-cut lettuce (metataxonomics), while sequencing all genetic material through shotgun sequencing offers a broader and more comprehensive view enabling the analysis of functional genes associated with spoilage, such as those involved in enzymatic degradation of plant tissues (metagenomics). However, careful study design, including sampling strategy and contextual factors, is essential to accurately capture microbial diversity and abundance, as an overabundance of plant DNA can compromise the reliability of bacterial profiling (Ioannidis et al., 2018). This challenge is particularly relevant in 16S rRNA gene sequencing, due to the sequence homology between chloroplast 16S rRNA genes, mitochondrial 18S rRNA genes and bacterial primers (Hanshew et al., 2013). Hence, primer selection plays a critical role in efficient target amplification, with primers like 799F/1391R and 799F/1193R, which amplify the V5–V7 hypervariable regions of the 16S rRNA gene, significantly minimizing chloroplast and mitochondrial DNA reads (Beckers et al., 2016; Hanshew et al., 2013; Wang et al., 2018). Indeed, previous studies failed to capture the true microbial diversity in MAP-stored iceberg lettuce due to the high abundance of chloroplast sequences after amplifying the V3–V4 regions, which led to low bacterial read counts and hindered accurate assessment of microbial community changes during storage (Ioannidis et al., 2018). Despite the microbiological quality and microbial population dynamics of RTE lettuce have been mainly studied in the Spanish and Dutch markets (Hernandez et al., 2020; Paillart et al., 2017; Woltering & Paillart, 2024), the study by Ioannidis et al. (2018) represents the only available research evaluating RTE iceberg lettuce obtained from a Belgian processing plant. Therefore, there is a need to effectively characterize the dynamics of the microbial community of Belgian market fresh-cut iceberg lettuce to understand the spoilage mechanisms, identify key spoilage microorganisms, and develop strategies to extend its shelf life. In this study, a culture-complemented metataxonomic approach is conducted on Belgian market fresh-cut iceberg lettuce spoilage, offering insights into the complex interactions underlying lettuce degradation. Through conventional plating and 16S rRNA gene sequencing with primers 799F/1193R, the detected spoilage microorganisms may serve as effective biomarkers for monitoring the quality of fresh-cut iceberg lettuce in intelligent packaging sensor-based systems. To the best of our knowledge, this is the first study evaluating the microbiological quality and dynamics of RTE iceberg lettuce available to consumers in the Belgian market. 2. Materials and methods 2.1. Sample preparation and characterization Fresh-cut iceberg lettuce (Lactuca sativa L. var. capitata) was produced and packaged in bags (25 ×25 ×6 cm) containing 500 g produce under controlled atmosphere packaging (1.5–2.5 % O 2 balanced with N 2 ) by a local processing plant in Belgium (ALLGRO NV, Sint-LievensHoutem, Belgium). The packaging film was made of coextruded polypropylene (PPCX, 30 μ m film thickness, 27.3 g/m 2 film weight) with a theoretical oxygen transmission rate (OTR) of 1600 cm 3 O 2 /m 2 /d/atm, measured at 23 ◦C and 0 % RH in accordance with standard ASTM D3985 as per the film manufacturer (Carton Pack S.p.A., Rutigliano, Italy). After packaging, the lettuce bags were transported in coolers (7 ± 1 ◦C) to the laboratory on the same day within 3 h (day 0). On their arrival, the lettuce bags were stored in the cold room at 7 ◦C for a maximum period of 12 days and were immediately analysed at the first sampling point (day 0). The storage temperature was selected to simulate retail sale conditions (Jacxsens et al., 2002; Paillart et al., 2017). According to the manufacturer, the commercial shelf-life was seven days after production. 2.2. Gas and pH measurements and microbial analysis Three packages from two different production batches produced on the 5th (batch 1) and 26th (batch 2) of March of 2025 were sampled under aseptic conditions on days 0, 1, 2, 5, 7, 9 and 12. At each sampling time, the CO 2 /O 2 gas analyser Checkmate 3 (Dansensor A/S, Ringsted, Denmark) was used to measure the composition of the gas mixture in the headspace of the packages. The pH was measured from the homogenate obtained by blending fresh-cut lettuce samples using the Edge® pH meter (Hanna Instruments, Belgium). The homogenate was prepared with the Braun hand blender type 4192 (600 W; De’Longhi Deutschland GmbH, Neu-Isenburg, Germany) for 60 s at room temperature. To determine microbial counts, 10 g of iceberg lettuce were transferred in a sterile blender bag with lateral filter (Novolab, Geraardsbergen, Belgium) and 90 g of peptone saline solution (PPS) were added (1 g/L neutralized bacteriological peptone, LP0034, Oxoid, Basingstoke, UK and 8.5 g/L NaCl CL00.1429.5000-5KG, AnalytiChem Belgium NV, Zedelgem, Belgium). Then, the sample was homogenized using a masticator homogenizer (IUL, Barcelona, Spain) for 60 s. Subsequently, a decimal dilution series was prepared in PPS, and specific volume aliquots were spread onto agar (0.1 mL) or pour plated (1 mL) for the enumeration of target microorganisms (Table 1). The pour plates were prepared using the double-layer method. The bags were disposed after each sampling day. 2.3. DNA isolation and 16S rRNA gene sequencing 16S rRNA gene amplicon sequencing analysis of the V5–V7 regions was conducted for the characterization of the iceberg lettuce microbiota over storage time. Samples were collected on days 0, 2, 7, and 12 to represent beginning, intermediate, “use-by” date, and final storage stages. On each sampling day, one replicate was prepared from each of the three lettuce bags analysed per batch, resulting in a total of six replicates per sampling time. DNA was extracted according to the method described in Krupka and Piotrowicz-Cie´ slak (2024). At each sampling time and after microbial sampling, 50 g of lettuce were aseptically mixed with 200 mL of Phosphate Buffered Saline (PBS) solution at pH 7.3 (BR0014G, Oxoid) containing 0.1 % Tween 80 (CL00.2086.1000, AnalytiChem Belgium NV) in a sterile blender bag with lateral filter. Thereafter, the mixture was sonicated with the Elmasonic S 30 (H) ultrasonic bath (Elma Schmidbauer GmbH, Singen, Germany) at an I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 2 ultrasonic frequency of 37 kHz for 7 min at room temperature to detach bacterial cells from the plant matrix. The suspension was subsequently vacuum filtered using a sterile cellulose nitrate (CN) membrane filter of 0.2 μ m pore size (Sartorius, Woluwe-Saint-Lambert, Belgium). Filters were stored at −20 ◦C for a maximum of 15 days before DNA extraction. DNA extraction was conducted using the DNeasy PowerSoil Pro kit (Qiagen, Antwerp, Belgium) following the manufacturer’s instructions with minor modifications. Briefly, the membrane filters were previously cut into small pieces using a sterile scalpel blade before transferring them to the PowerBead Pro Tube to increase the DNA yield (Krupka & Piotrowicz-Cie´ slak, 2024). The concentration and quality of the extracted DNA were assessed using the BioDrop μ Lite+spectrophotometer (Biochrom, Gründau, Germany). Library preparation and sequencing was carried out by Novogene (Cambridge, UK) (Novogene, 2025). Amplification of the V5–V7 region of the 16S rRNA gene was performed using the primers 799F (5’- AACMGGATTAGATACCCKG-3 ′ ) and 1193R (5’-ACGTCATCCCCACCTTCC-3 ′ ) with specific barcodes. These primers were selected to minimize amplification of chloroplast and mitochondrial sequences originating from host cells, thereby facilitating deeper sequencing and a more comprehensive identification of bacterial amplicon sequencing variants (ASVs) (Anguita-Maeso et al., 2022). All PCR reactions were carried out with 15 μ L of Phusion® High-Fidelity PCR Master Mix (New England Biolabs, Massachusetts, United States), 0.2 μ M of forward and reverse primers, and 10 ng of template DNA. Thermal cycling consisted of initial denaturation at 98 ◦C for 1 min, followed by 30 cycles of denaturation at 98 ◦C for 10 s, annealing at 50 ◦C for 30 s, and elongation at 72 ◦C for 30 s and 72 ◦C for 5 min. An equal volume of 1× loading buffer (containing SYBR Green) was mixed with the PCR products, and electrophoresis was performed on a 2 % agarose gel for DNA detection. Sequencing libraries were prepared from purified PCR products corresponding to the expected 394 bp amplicon using the NEBNext® Ultra™ IIDNA Library Prep Kit (Cat No. E7645) (New England Biolabs) for paired-end sequencing (2 ×250 bp) on an Illumina NovaSeq 6000 platform (Illumina, California, USA). The sequencing depth of each sample was at least 100,000 sequences. Sample IDs followed the format B{Batch}D{Day}{Replicate}, with batches labelled as 1 or 2, days labelled as 0, 2, 7 or 12, and replicates labelled as A, B, or C. 2.4. 16S rRNA gene sequencing data analysis Paired-end reads were assigned to samples based on their unique barcode and truncated by cutting off the barcode and primer sequence. The paired-end reads were merged using FLASH v1.2.11 (Magoˇ c & Salzberg, 2011) with a minimum overlap of 10 bp (default setting), a maximum overlap of 220 bp, and a maximum mismatch density of the overlap region equal to 0.2. The merged reads were quality filtered using Fastp v0.23.1 (Chen et al., 2018) to obtain high-quality clean tags. PolyG tail trimming was enforced, a minimum Phred quality threshold of 19 was applied, and up to 15 % of bases were allowed to below the quality threshold to keep the read. Then, the clean tags were compared with the Silva v138.1 database (Quast et al., 2013) to detect chimera sequences using vsearch v2.16.0 (Rognes et al., 2016) and the UCHIME algorithm (Edgar et al., 2011). After removing the chimeric sequences, the effective tags were obtained. Microbiota bioinformatics were performed with QIIME 2 v2024.10.1 (Bolyen et al., 2019). Denoising of the effective tags was performed using DADA2 v1.32.0 (Callahan et al., 2016) to generate initial amplicon sequence variants (ASVs). The ASVs were taxonomically classified using an in-house trained naïve Bayes classifier based on the SILVA v138 reference database (Silva 138 SSURef NR99 full-length sequences) available at https://docs.qiime2.org/2024.10/data-resources/ (Bokulich et al., 2018; Ii et al., 2021; Quast et al., 2013). The naïve Bayes classifier was trained on the V5–V7 region that was sequenced to improve the accuracy of taxonomic classification (Werner et al., 2012). ASVs with total abundances below 5 across all samples were subsequently removed to minimize ecologically uninformative noise (Li et al., 2020). Additionally, ASVs assigned to unknown domains, chloroplasts or mitochondria were excluded from the dataset. The phylogenetic tree was constructed by aligning ASV sequences using MAFFT (Katoh et al., 2002), followed by masking phylogenetically uninformative and ambiguously aligned columns. FastTree (Price et al., 2009) was then used to infer the phylogenies, which were subsequently midpointrooted. The absolute abundance of ASVs was normalized using a standard of sequence number corresponding to the sample with the least sequences in order to avoid bias due to different sampling depths and obtain a comparable dataset. Alpha rarefaction curves were then generated using the frequency data of representative ASVs from each sample to evaluate whether the sequencing depth was sufficient to capture the microbial diversity. All subsequent analysis of alpha diversity and beta diversity were performed based on the normalized dataset. Alpha diversity analyses were performed using core-metricsphylogenetic workflow of QIIME 2, producing several alpha diversity measures like Faith’s Phylogenetic Diversity (PD), observed ASVs, Pielou’s evenness, and Shannon index. Beta diversity was calculated based on weighted and weighted UniFrac distances. 2.5. Statistical analysis Statistical analyses were conducted to evaluate differences across sampling times and production batches. For variables requiring parametric testing, two-way ANOVA was applied after verifying assumptions of normality (Shapiro-Wilk test) and homogeneity of variance (Levene’s test). When normality and/or homoscedasticity was violated, the nonparametric Aligned Rank Transform (ART) ANOVA was applied, followed by post-hoc pairwise comparisons of estimated marginal means with Tukey correction. These analyses were performed in R v4.4.0 (R Core Team, 2024). Microbial community diversity was assessed in QIIME 2 using the Kruskal-Wallis test for the alpha-diversity metrics and PERMANOVA for the beta-diversity metrics using the default number of permutations (999) for p-value computation. In addition, differential abundance analysis was conducted at the genus level using ANCOM-BC2 v2.6.0 (Lin & Peddada, 2020) implemented in R. Statistical significance was set at p ≤0.05. Table 1 The target microorganisms and their corresponding agar, culture method, and incubation condition used. Target Agar medium Culture method Incubation temperature (◦C) Incubation time (h) Total psychrotrophic counts (TPC) Plate count agar (PCA) (CM0325, Oxoid) Spread plate 22 120 Enterobacteria Rapid’ Enterobacteriaceae agar (3564004, BIO-RAD, Marnes-La-Coquette, France) Pour plate 37 24 Pseudomonas spp. Pseudomonas agar (CM0559B, Oxoid) with C-F-C supplement (SR0103E, Oxoid) Spread plate 25 48 Psychrotrophic lactic acid bacteria (LAB) De Man, Rogosa and Sharpe agar (MRS) agar (CM0361, Oxoid) supplemented with 0.14 % sorbic acid (CL00.1961.0100, AnalytiChem Belgium NV) Pour plate 22 120 Yeasts Yeast glucose chloramphenicol (YGC) agar (3564104, BIO-RAD) Spread plate 25 120 I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 3 3. Results 3.1. Headspace gas composition (% O 2 and % CO 2 ), pH and appearance Variations in headspace gas composition (% O 2 and % CO 2 ) and pH during storage were significantly influenced by sampling time, production batch and their interaction (p ≤0.05) (Supplementary Table 1). Oxygen levels decreased significantly within the first 24 h of storage, remaining at residual concentrations thereafter (0.002–0.018 %) (Fig. 1A; Supplementary Table 2). Moreover, the initial O 2 content of batch 1 on day 0 (2.047 ±0.032 %) was significantly higher than that of batch 2 (1.510 ±0.242). In contrast, carbon dioxide was continuously produced during storage, increasing from below 2 % to over 23 % by day 12 in both batches (Fig. 1A, Supplementary Table 2). Batch-related differences in CO 2 levels during storage became evident mainly towards the end of storage. After days 5 and 7, only batch 2 showed significant increases, whereas after day 9, a significant increase was observed exclusively in batch 1. Moreover, both batches exhibited statistically significant CO 2 levels on day 9, when batch 2 exhibited a significantly higher CO 2 concentration than batch 1 (22.9 ±2.1 % CO 2 vs 18.2 ±0.4 % CO 2 ). With regard to the pH, fluctuations were observed during storage (5.75–6.28), with only a significant decrease to 5.20 in batch 2 on day 12 (Fig. 1B, Supplementary Table 2). Thus, all differences observed in the temporal evolution of headspace gas composition and pH between batches appears limited during the commercial shelf-life. 3.2. Enumeration of microorganisms from lettuce Microbial counts were also significantly influenced by storage time, production batch and their interaction (p ≤0.05) (Supplementary Table 1). Although the effect of storage time on the growth of microbes depended on the production batch, the overall behaviour of the target microorganisms was comparable across batches, with no indication of meaningful biological impact. Nevertheless, microbial counts in batch 2 were slightly higher than those in batch 1, particularly at the beginning of the storage period (Table 2, Supplementary Fig. 1). All tested microorganisms, including TPC, enterobacteria, Pseudomonas spp., psychrotrophic LAB and yeasts, showed a progressive proliferation in the fresh-cut iceberg lettuce during storage (p ≤0.05) (Table 2). Initial contamination levels were primarily driven by TPC and Pseudomonas populations, with counts around 4.6 log CFU/g, whereas LAB were the least abundant microorganisms (≤2.5 log CFU/g). Throughout storage, TPC remained the predominant group, followed by Pseudomonas spp. until day 2, and by enterobacteria thereafter. As a result, bacterial concentrations reached between 6.8 and 7.8 log CFU/g by the “use-by” date (day 7), exceeding yeast levels of 5.3 log CFU/g. Moreover, during the first 24 h of storage, TPC, Pseudomonas spp. and yeasts showed a slight decrease, opposite to enterobacteria and LAB that were able to grow during the same period. Notably, only the increase in LAB during the first 24 h was statistically significant (p ≤0.05). LAB also exhibited the highest growth rate during storage, increasing by 5.7 log CFU/g from day 0 to day 12, with 4.5 log CFU/g of this increase occurring during the first five days. In contrast, the growth of TPC, enterobacteria, and LAB slowed considerably after day 5, with only 1.0–1.2 log CFU/g increases observed by day 12. 3.3. 16S rRNA gene sequencing analysis An average of 193,186 raw paired-end reads per sample were obtained across all samples, with read counts ranging from 113,262 to 218,564 (Supplementary Table 3). After quality filtering and removal of chimeric sequences, an average of 126,661 reads were retained (range: 61,364–158,674). These denoised reads resulted in the initial identification of 3470 unique ASVs across all samples, that were reduced to 2218 ASVs after removing low-abundance variants (36.1 %). Sequences corresponding to chloroplasts and mitochondria accounted then for a small proportion, with 65 (2.9 %) and 2 (<0.1 %) ASVs, respectively. These sequences were only detected at the beginning of the storage period (days 0 and 2), showing particularly higher relative abundance in the first batch (13.4 ±12.3 %) compared to the second batch (4.6 ±4.2 %). Moreover, in two replicates from the first batch (B1D0B from day 0 and B1D2A from day 2), these sequences accounted for 27.6 % and 25.3 % of total variants, respectively. Despite the significant reduction in dataset size for these two samples following removal of chloroplast and mitochondrial sequences, all maintained a sequencing depth above the determined minimum threshold of 47,398 sequences (Supplementary Table 3). Furthermore, at this sequencing depth, rarefaction curves plateaued across all samples (Supplementary Fig. 2). After rarefaction, a total of 2146 unique ASVs were obtained across all samples. Among these, 81 ASVs were consistently detected on all sampling days, constituting the core microbiota of the fresh-cut lettuce stored in MAP at 7 ◦C (Fig. 2A). Most of the ASVs (75.4 %) were only found during the early stages of storage (days 0 and 2), showing a marked decrease in diversity over storage, with ASVs dropping from 1396 on day 0 to 349 by day 12. Nonetheless, despite the progressive decline in bacterial richness during storage, the bacterial community remained relatively stable from the end of the commercial shelf-life (day 7) onward. In fact, a significant effect of storage time on bacterial richness was observed based on both Faith’s PD and Observed ASVs (p ≤ 0.05), with higher community richness at early storage stages (days 0 and 2) compared to later stages (days 7 and 12) (Fig. 2B). However, the Pielou’s evenness index did not show statistically significant differences in relative abundance (p >0.05). Similarly, the Shannon index, which integrates both richness and evenness, remained stable across both storage time and batches (p >0.05). Indeed, alpha diversity did not differ significantly between batches (p >0.05), although the first batch Fig. 1. Changes in headspace gas composition (% O 2 and % CO 2 ) (A) and pH (B) of fresh-cut lettuce during storage in MAP at 7 ◦C. Each data point represents the mean value, and error bars indicate the standard deviation from six independent samples analysed across two production batches. I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 4 exhibited greater within-batch richness (Fig. 2B). Beta diversity analysis further revealed significant changes in microbial community over time (p ≤0.05) based on both unweighted and weighted UniFrac distance matrices (Fig. 2C). While no significant differences in bacterial diversity between batches were observed (p >0.05), samples from days 0 and 2 clustered together according to unweighted UniFrac distances and were significantly different from those collected on days 7 and 12 (p ≤0.05), which also formed a distinct cluster. Moreover, weighted UniFrac distances indicated a more gradual transition in community structure, with day 2 representing an intermediate state between the initial microbiota (day 0) and later storage stages (days 7 and 12), which were significantly different (p ≤0.05). The relative distribution of the microbial composition in the freshcut lettuce showed increased intraand inter-variability at the beginning of storage on days 0 and 2, remaining stable from the end of the commercial shelf-life (day 7) thereafter (Fig. 3A; Supplementary Fig. 3). The initial microbiota (day 0) was generally diverse, with Pseudomonas being the most abundant genus (38.9 %), followed by enterobacteria, including Pantoea (22.2 %) and Yersinia (4.6 %), among others. However, their abundance declined over the storage period, particularly Pseudomonas, which decreased to 15.1 % by day 2, and fell below 2 % from day 7 onward (p ≤0.05, Fig. 3B). In contrast, Lactococcus emerged as the dominant genus from day 2 onward, increasing from 2.6 % on day 0 to 36.7 % on day 2, and exceeding 44 % on days 7 and 12 (p ≤0.05). A similar pattern was observed for Serratia, which increased from just 1.7 % on day 0 to become the second most abundant genus by day 7, maintaining a relative abundance of 21.3 % thereafter (p ≤0.05). Additionally, the abundance of Rahnella increased throughout storage, rising from 0.3 % on day 0 to 9.5 % by day 12, peaking as the third most abundant genus on day 7 at 11 % (p ≤0.05). In contrast, the relative abundance of Leuconostoc was 0.01 % on day 0, increased to 0.06 % on day 2, and declined to 0.03 % on days 7 and 12. 4. Discussion The respiration rate in the commercial RTE iceberg lettuce followed a trend similar to previous studies on this and other lettuce types after processing (Hernandez et al., 2020; Paillart et al., 2017). Oxygen was consumed rapidly within the first 24 h of storage that resulted in an anaerobic-like atmosphere (<0.1 % O 2 ). This is consistent with earlier studies reporting that controlled atmosphere storage inhibits respiration in fresh-cut produce, although the timing of this effect depends on the initial oxygen concentration and the film oxygen transmission rate (OTR) (Ioannidis et al., 2018; Kim et al., 2005; Paillart et al., 2017). In the present study, the depletion of the oxygen was similar to that observed in packages composed of PP film with an OTR of 8 pmol/m 2 /s/ Pa measured at 5 ◦C in air (equivalent to ~1600 cm 3 O 2 /m 2 /d/atm), which also resulted in rapid depletion of O 2 to ~0 kPa within 1 day when flushed with ≤2.5 kPa initial O 2 (Kim et al., 2005). The Belgian company used a film with even lower permeability, which, according to the Arrhenius equation, is predicted to be 661 cm 3 O 2 /m 2 /d/atm at 7 ◦C (Yaptenco et al., 2007). This OTR is comparable to BOPP films used in anaerobic laboratory packaging applications (OTR =632 cm 3 O 2 /m 2 /d/ atm, measured at 7 ◦C and 90 % RH) (Ioannidis et al., 2018). Therefore, our findings indicate that the low initial oxygen concentration together with the low permeability of the packaging film contributed to the rapid oxygen depletion due to aerobic respiration. Low initial oxygen levels are commonly used in the industry to help preserve the freshness of the lettuce during its commercial shelf-life (Hernandez et al., 2020; Woltering & Paillart, 2024). However, the resulting rapid anaerobic-like atmosphere can lead to premature formation of undesirable sour offodors from anaerobic respiration of the lettuce and the microbiota (Ioannidis et al., 2018; Tudela et al., 2013). Hence, the gas mixture used in the lettuce bags may have contributed to the development of lettuce discoloration and potential flavor changes during storage, likely due to shifts in the microbiota. Despite oxygen depletion, residual levels of oxygen were still detected in the lettuce bags, in agreement with previous studies (Hernandez et al., 2020; Woltering & Paillart, 2024). The facultative anaerobic nature of LAB, enterobacteria and certain yeasts allows them to adapt quickly to low-oxygen environments by switching to anaerobic respiration or fermentation. At the same time, lettuce can also shift to anaerobic respiration under these conditions, all of which may have possibly contributed to the presence of residual oxygen. Meanwhile, Pseudomonas spp. and yeasts initially reduced their growth rates when the oxygen was rapidly depleted, yet they were able to persist and even Pseudomonas thrived under the apparently anaerobic conditions, a pattern also reported in RTE rocket salad by Manthou et al. (2022). This suggests that the film’s gas permeability may have supported the growth of aerobic microorganisms, particularly Pseudomonas spp., by providing oxygen, along with the metabolic activity of facultative anaerobes. In fact, Pseudomonas aeruginosa is capable of utilizing ethanol and organic acids (e.g., acetic acid) produced by LAB and yeasts as alternative substrates under oxygen-limiting conditions (Crocker et al., 2019) and at reduced pH levels (5.5–7) (Bushell et al., 2019), which might have supported the proliferation of the Pseudomonas population under nearzero oxygen conditions. At the same time, carbon dioxide levels increased steadily and exceeding 20 % at the end of the storage period, as also observed Paillart et al. (2017) and Woltering and Paillart (2024). This phenomenon is attributed to the combined respiration of the lettuce Table 2 Changes in microbial counts (log CFU/g) of the target microorganisms analysed in the fresh-cut lettuce samples stored up to 12 days in MAP at 7 ◦C. Target Batch Day 0 Day 1 Day 2 Day 5 Day 7 Day 9 Day 12 TPC 1 4.1 ±0.2 a 4.0 ±0.1 a 5.0 ±0.5 b 7.1 ±0.1 d 7.4 ±0.2 de 8.0 ±0.2 f 8.4 ±0.0 fg 2 3.1 ±0.3 b 3.4 ±0.0 b 4.4 ±0.4 c 6.4 ±0.3 ef 7.2 ±0.1 fg 7.5 ±0.2 fg 8.3 ±0.1 g mean 4.6 ±0.6 4.5 ±0.5 5.5 ±0.6 7.6 ±0.5 7.8 ±0.5 8.3 ±0.3 8.6 ±0.2 Enterobacteria 1 3.1 ±0.3 a 3.4 ±0.0 ab 4.4 ±0.4 c 6.4 ±0.3 e 7.2 ±0.1 ef 7.5 ±0.2 f 8.3 ±0.1 g 2 4.0 ±0.6 bc 4.6 ±0.2 c 5.5 ±0.2 d 7.4 ±0.1 f 7.6 ±0.1 fg 7.9 ±0.1 fg 7.9 ±0.4 fg mean 3.6 ±0.7 4.0 ±0.7 5.0 ±0.7 6.9 ±0.5 7.4 ±0.2 7.7 ±0.3 8.1 ±0.3 Pseudomonas spp. 1 4.0 ±0.1 a 4.0 ±0.0 a 4.8 ±0.3 b 6.1 ±0.0 d 6.4 ±0.1 de 7.4 ±0.2 fg 8.0 ±0.1 h 2 5.2 ±0.2 bc 4.7 ±0.1 b 5.4 ±0.1 c 6.9 ±0.2 ef 7.1 ±0.2 fg 7.4 ±0.1 fg 7.7 ±0.4 gh mean 4.6 ±0.6 4.3 ±0.4 5.1 ±0.4 6.5 ±0.5 6.8 ±0.4 7.4 ±0.2 7.9 ±0.4 LAB 1 1.6 ±0.3 a 2.0 ±0.2 a 3.2 ±0.1 b 6.1 ±0.2 d 6.7 ±0.2 de 7.0 ±0.2 ef 7.5 ±0.1 f 2 2.2 ±0.3 a 3.0 ±0.5 b 4.0 ±0.2 c 6.4 ±0.2 de 6.8 ±0.2 def 7.6 ±0.1 f 7.5 ±0.4 f mean 1.8 ±0.4 2.5 ±0.7 3.6 ±0.5 6.3 ±0.2 6.8 ±0.2 7.3 ±0.3 7.5 ±0.3 Yeasts 1 2.7 ±0.2 a 2.9 ±0.4 a 3.0 ±0.1 a 4.3 ±0.0 cd 4.5 ±0.1 cd 5.0 ±0.5 de 6.0 ±0.2 fg 2 4.0 ±0.7 bc 3.1 ±0.1 ab 4.0 ±0.4 bc 5.1 ±0.1 def 6.0 ±0.1 efg 5.9 ±0.4 fg 6.2 ±0.2 g mean 3.4 ±0.8 3.0 ±0.3 3.5 ±0.6 4.7 ±0.5 5.2 ±0.8 5.5 ±0.6 6.1 ±0.2 Results are presented as mean ±standard deviation from three independent samples analysed per production batch, including the overall average across both batches. Two-way ANOVA followed by Tukey’s HSD test was performed to assess differences in microbial counts. Different letters indicate statistically significant differences between batch ×day combinations (p ≤0.05). TPC, total psychrotrophic counts. LAB, psychrotrophic lactic acid bacteria. I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 5 Fig. 2. Alpha and beta diversity measures of the 16S rRNA gene sequencing data. (A) Venn diagram and upset plot showing the distribution of unique and shared ASVs across sampling days. (B) Comparisons of alpha diversity measures (Faith’s PD, observed ASVs, Pielou’s evenness, and Shannon) by batch (1: blue; 2: orange) and sampling day. The box represents the first (Q1) and third (Q3) quartiles of the distribution, and the line within the box marks the median. The whiskers extend from Q1 and Q3 to the last data points within 1.5 ×IQR and values beyond these whiskers are considered outliers (black border). Non-parametric KruskalWallis test followed by post-hoc pairwise Wilcoxon rank-sum tests with Benjamini-Hochberg False Discovery Rate correction for multiple comparisons was conducted. Different letters indicate statistically significant differences between sampling times (days) (p ≤0.05). (C). PCoA based on beta diversity measures (weighted and unweighted UniFrac distances of microbial communities). I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 6 and microbiota, along with the low permeability of the packaging film, that prevented the system from reaching equilibrium under MAP. Stable CO 2 concentrations have only been reported in packaging films with higher permeabilities (2200 cm 3 O 2 /m 2 /d/atm, measured at 7 ◦C and 90 % RH; 3250 cm 3 O 2 /m 2 /d/atm, measured at 5 ◦C in air) (Ioannidis et al., 2018; Kim et al., 2005). To achieve an equilibrium modified atmosphere within the package, the gas permeability of the film to O 2 and CO 2 has to align with the respiration rate of the packaged fresh-cut produce (Jacxsens et al., 2002). Here, the initial low microbial load suggests that early CO 2 production was primarily driven by aerobic respiration performed by the vegetable itself. At later stages, microbial activity contributed more significantly as microbial populations increased, particularly facultative anaerobes such as LAB, enterobacteria and yeasts, as well as Pseudomonas spp., which can grow under anaerobic conditions via ethanol catabolism (Crocker et al., 2019). These findings suggest that evolution of headspace gas composition during storage of RTE lettuce are likely determined by the combined effects of film gas permeability, microbiota changes and respiration rates. The sequencing depth was adequate to capture most of the microbial diversity in the MAP-stored lettuce, as evidenced by the rarefaction curves. Moreover, alpha diversity metrics confirmed a reduction in bacterial richness during storage, contrary to Ioannidis et al. (2018), probably due to a higher proportion of unidentified bacterial genera at the beginning of storage (65–70 %) in their study. Nonetheless, opposing trends in bacterial richness during the storage of RTE lettuce are commonly observed across studies (De Bock et al., 2024; Hernandez et al., 2020). Despite the evenness of the bacterial community remained stable during storage, changes in the headspace gas composition contributed to shifts in the dominance of the microbial population in line with previous observations (Paillart et al., 2017; Woltering & Paillart, 2024). These changes were further supported by the beta diversity metrics, which revealed two distinct bacterial community groups corresponding to the early and later stages of storage. This indicates that, while the microbial communities maintained a similar balance in terms of abundance distribution, the specific bacterial taxa dominating the community changed over time. When the oxygen was abundant at the beginning of storage, both microbial counts and sequencing reads demonstrated a dominance of Gram-negative bacteria like Pseudomonas spp., followed by Pantoea species, in accordance with Gu et al. (2024) and De Bock et al. (2024). From the intermediate storage stage (day 5) onward, under anaerobic-like conditions, enterobacteria remained more viable than Pseudomonas spp. and LAB. Although the viability of Pseudomonas spp. and LAB remained similar, Pseudomonas species were present at very low abundance, while Lactococcus spp. proliferated extensively, ultimately becoming the predominant spoilage bacteria in the fresh-cut lettuce (Woltering & Paillart, 2024). This data contrasts with previous studies reporting higher LAB counts under anaerobic conditions when LAB populations prevail (Paillart et al., 2017), and suggests that a reduced proportion of the Lactococcus community was viable during the later stages of storage. Hence, the findings highlight the relevance of combining traditional culture-based methods with 16S rRNA gene sequencing for a more comprehensive understanding. In addition, 18S rRNA gene sequencing could have provided valuable insights into the abundance of fungi, which are well-known agents of microbial spoilage in lettuce (Alegbeleye et al., 2022). Other spoilage bacteria that were detected included Serratia and Rahnella species (Caldera & Franzetti, 2014; Ioannidis et al., 2018; Jacxsens et al., 2003), that together with Lactococcus accounted for 77 % of the bacteriome on the 7th day of storage, which is the “use-by” date given by the producer. Other studies have also reported the association of Leuconostoc species with the spoilage of fresh-cut lettuce in Belgium (Ioannidis et al., 2018; Pothakos et al., 2014; Woltering & Paillart, 2024). However, these LAB bacteria were consistently detected here at very low abundances (<1 %) throughout storage. The bacterial population dynamics resulted in bacterial counts above 7.5 log CFU/g by the end of the 12-day storage period, being the psychotropic bacteria the most viable group (Hernandez et al., 2020). While psychrotrophic bacteria remained below 8 log CFU/g during the commercial shelf-life, a level typically linked to sensory quality loss, yeast counts above 5 log CFU/g could nonetheless cause changes in lettuce flavor (Ragaert et al., 2007). Thus, a sensory evaluation would have been valuable to determine whether the fresh-cut lettuce remained fully acceptable at the “useby” date to correctly predict the shelf life and minimize food waste. In conclusion, the combined effects of film gas permeability, microbiota dynamics, and respiration rates led to a rapid depletion of the initial oxygen, accompanied by a progressive increase in carbon dioxide levels. Nonetheless, the Pseudomonas population was able to grow under the limited oxygen atmosphere, achieving levels as high as other bacteria by the end of storage, even though they ultimately represented a minor group in the bacterial community. However, the initial low oxygen levels did not suppress the excessive proliferation of 0 10 20 30 40 50 60 70 80 90 100 Relative abundance (%) Day 0 Day 2 Day 7 Day 12 Other g__ Rahnella g__ Erwinia g__ Serratia g__ Flavobacterium g__ Lactococcus g__ Yersinia f__ Enterobacteriaceae g__ Pantoea g__ Pseudomonas AA 0.97 0.91 0.62 −0.05 −0.35 −0.3 −1.18 −3.33** −4.1*** −2.14 −2.92* −0.77 −0.85 −3.01** −3.31*** −2.16 −2.46* −0.3 0.64 −1.08 0.92 −1.72 0.28 2.01 3.12* 3.17** 3.17** 0.05 0.05 0 1.43 0.7 0.72 −0.73 −0.71 0.02 1.54 −0.22 −0.56 −1.76 −2.09 −0.34 −0.39 −2.95*** −3.08*** −2.56* −2.69* −0.13 2.36 3.72*** 3.51*** 1.37 1.16 −0.21 2.05 3.13** 3.07** 1.08 1.03 −0.06 −0.26 −3.44** −3.53** −3.18** −3.27** −0.09 0.65 0.91 1.19 0.26 0.54 0.28 0.72 0.69 0.81 −0.03 0.08 0.12 Yersinia Weissella Stenotrophomonas Serratia Rahnella Pseudomonas Pantoea Leuconostoc Lactococcus Lactobacillus Janthinobacterium Flavobacterium Erwinia D0 − D2 D0 − D7 D0 − D12 D2 − D7 D2 − D12 D7 − D12 −4 −2 0 2 Log Fold Change BB Fig. 3. Taxonomic composition of the microbial communities at the genus level in fresh-cut lettuce stored in MAP at 7 ◦C based on the metataxonomic analysis of the 16S rRNA gene. (A) Cumulative bar charts of abundance of frequently detected genera – g – (or family – f – when genus was not specified) are shown, while all remaining genera are grouped together as “Other”. (B) Heatmap of differentially abundant genera between sampling days. The X-axis represents the specific pairwise comparisons between sampling days, while the Y-axis displays the significant genera identified by ANCOM-BC2. Each cell is colour-coded, with blue indicating decreased abundance and red indicating increased abundance relative to the comparison, and the numbers on each cell indicate the log fold-change. Statistical significance is denoted by asterisks (*, p ≤0.05; **, p ≤0.01; ***, p ≤0.001). P-values were adjusted for multiple testing using the HolmBonferroni method. I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 7 Lactococcus species, which became the most abundant spoilage bacteria together with enterobacteria such as Serratia and Rahnella species. Therefore, the detection of these bacteria is critical to prevent spoilage of fresh-cut iceberg lettuce and may serve as effective biomarkers for monitoring its quality in intelligent packaging sensor-based systems. CRediT authorship contribution statement Irene Ortega-Sanz: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Soma Ishihara: Writing – review & editing, Methodology. Frank Devlieghere: Writing – review & editing, Supervision, Resources, Funding acquisition. Andreja Rajkovic: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work was funded by the European Union’s Horizon Europe Research and Innovation programme under grant agreement No. 101136542 (FOODGUARD project). Soma Ishihara received a scholarship by the Heiwa Nakajima Foundation (Japan). The authors are grateful to ALLGRO for its collaboration in this study and to HPC-UGent supercomputing infrastructure for the HPC support. Trade names or commercial products mentioned in this publication are cited solely for the purpose of providing specific information and do not constitute an endorsement or recommendation for their use or consumption. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.foodres.2025.117694. Data availability I have shared the link to my data References Alegbeleye, O., Boas, D. M. V., & Sant’Ana, A. S.. (2025). Harnessing the microbiota of vegetables and ready-to-eat (RTE) vegetables for quality and safety. Food Research International, 214, Article 116667. https://doi.org/10.1016/j.foodres.2025.116667 Alegbeleye, O., Odeyemi, O. A., Strateva, M., & Stratev, D. (2022). Microbial spoilage of vegetables, fruits and cereals. Applied Food Research, 2(1), Article 100122. https:// doi.org/10.1016/j.afres.2022.100122 Alemu, T. T., & Oanh, V. T. K. (2025). Implications of developmental levels and packaging materials on the quality of iceberg lettuce for marketing system: A review. International Journal of Food Science, 2025(1), Article 5548305. https://doi.org/ 10.1155/ijfo/5548305 Anguita-Maeso, M., Haro, C., Navas-Cort´ es, J. A., & Landa, B. B. (2022). Primer choice and xylem-microbiome-extraction method are important determinants in assessing xylem bacterial Community in Olive Trees. Plants, 11(10), Article 10. https://doi. org/10.3390/plants11101320 Beckers, B., Op De Beeck, M., Thijs, S., Truyens, S., Weyens, N., Boerjan, W., & Vangronsveld, J. (2016). Performance of 16s rDNA primer pairs in the study of rhizosphere and Endosphere bacterial microbiomes in Metabarcoding studies. Frontiers in Microbiology, 7. https://doi.org/10.3389/fmicb.2016.00650 Bhatia, V., Nag, R., Burgess, C. M., Gaffney, M., Celayeta, J. M. F., & Cummins, E. (2024). Microbial risks associated with ready-to-eat fresh produce (RTEFP) – A focus on temperate climatic conditions. Postharvest Biology and Technology, 213, Article 112924. https://doi.org/10.1016/j.postharvbio.2024.112924 Bokulich, N. A., Kaehler, B. D., Rideout, J. R., Dillon, M., Bolyen, E., Knight, R., … Gregory Caporaso, J. (2018). Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2’s q2-feature-classifier plugin. Microbiome, 6(1), 90. https://doi.org/10.1186/s40168-018-0470-z Bolyen, E., Rideout, J. R., Dillon, M. R., Bokulich, N. A., Abnet, C. C., Al-Ghalith, G. A., … Caporaso, J. G. (2019). Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nature Biotechnology, 37(8), 852–857. https://doi.org/10.1038/s41587-019-0209-9 Bushell, F. M. L., Tonner, P. D., Jabbari, S., Schmid, A. K., & Lund, P. A. (2019). Synergistic impacts of organic acids and pH on growth of Pseudomonas aeruginosa: A comparison of parametric and Bayesian non-parametric methods to model growth. Frontiers in Microbiology, 9, 3196. https://doi.org/10.3389/fmicb.2018.03196 Caldera, L., & Franzetti, L. (2014). Effect of storage temperature on the microbial composition of ready-to-use vegetables. Current Microbiology, 68(2), 133–139. https://doi.org/10.1007/s00284-013-0430-6 Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods, 13(7), 581–583. https://doi.org/10.1038/ nmeth.3869 Cao, Y., Fanning, S., Proos, S., Jordan, K., & Srikumar, S. (2017). A review on the applications of next generation sequencing technologies as applied to food-related microbiome studies. Frontiers in Microbiology, 8. https://doi.org/10.3389/ fmicb.2017.01829 Chen, S., Zhou, Y., Chen, Y., & Gu, J. (2018). fastp: An ultra-fast all-in-one FASTQ preprocessor. Bioinformatics, 34(17), i884–i890. https://doi.org/10.1093/ bioinformatics/bty560 Choi, I.-L., Lee, J.-H., Choi, D.-H., Wang, L.-X., & Kang, H.-M. (2021). Evaluation of the storage characteristics in maintaining the overall quality of whole and fresh-cut Romaine lettuce during MA storage. Horticulturae, 7(11), 11. https://doi.org/ 10.3390/horticulturae7110461 Colruyt Group. (2025). 4 trends for a crackling Colruyt barbecue. Retrieved from https:// press.colruytgroup.com/4-trends-for-a-crackling-colruyt-barbecue. Accessed August 25, 2025. Crocker, A. W., Harty, C. E., Hammond, J. H., Willger, S. D., Salazar, P., Botelho, N. J., … Hogan, D. A. (2019). Pseudomonas aeruginosa ethanol oxidation by AdhA in lowoxygen environments. Journal of Bacteriology, 201(23). https://doi.org/10.1128/ JB.00393-19. e00393–19. De Bock, T., Wieme, A., Vandamme, P., Devlieghere, F., & Uyttendaele, M. (2024). Microbial dynamics in equilibrium modified-atmosphere packaged fresh-cut Romaine lettuce: A culturomics approach. 28th international ICFMH conference (FoodMicro24), abstract book (pp. 146–147). http://hdl.handle.net/1854/LU-01JHZJPVXFRN21PNV M89NJH3HG. Delikanli-Kiyak, B., Yilmaz, I., & Guldas, M. (2025). Can metagenomic analyses be used effectively in safe food production? Food Science & Nutrition, 13(8), Article e70772. https://doi.org/10.1002/fsn3.70772 Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C., & Knight, R. (2011). UCHIME improves sensitivity and speed of chimera detection. Bioinformatics, 27(16), 2194–2200. https://doi.org/10.1093/bioinformatics/btr381 FAO. (2023). FAOSTAT. Retrieved from https://www.fao.org/faostat/en/#home. Accessed August 25, 2025. Gu, G., Ding, Q., Redding, M., Yang, Y., O’Brien, R., Gu, T., … Nou, X. (2024). Differential microbiota shift on whole romaine lettuce subjected to source or forward processing and on fresh-cut products during cold storage. International Journal of Food Microbiology, 416, Article 110665. https://doi.org/10.1016/j. ijfoodmicro.2024.110665 Gustavsson, J., Cederberg, C., Sonesson, U., van Otterdijk, R., & Meybeck, A. (2011). Global food losses and food waste: Extent, causes and prevention. In International congress save food!, Rome. Food and Agriculture. Organization of the United Nations. Hanshew, A. S., Mason, C. J., Raffa, K. F., & Currie, C. R. (2013). Minimization of chloroplast contamination in 16S rRNA gene pyrosequencing of insect herbivore bacterial communities. Journal of Microbiological Methods, 95(2), 149–155. https:// doi.org/10.1016/j.mimet.2013.08.007 Hernandez, I., Alfaro, B., & Rodriguez-Ezpeleta, N. (2020). Enhancing high throughput sequencing unveils changes in bacterial communities during ready-to-eat lettuce spoilage. Journal of Horticulture and Postharvest Research, 3(Issue 2), 297–310. https://doi.org/10.22077/jhpr.2020.3029.1115 Ii, M. S. R., O’Rourke, D. R., Kaehler, B. D., Ziemski, M., Dillon, M. R., Foster, J. T., & Bokulich, N. A. (2021). RESCRIPt: Reproducible sequence taxonomy reference database management. PLoS Computational Biology, 17(11), Article e1009581. https://doi.org/10.1371/journal.pcbi.1009581 Ioannidis, A.-G., Kerckhof, F.-M., Riahi Drif, Y., Vanderroost, M., Boon, N., Ragaert, P., De Meulenaer, B., & Devlieghere, F. (2018). Characterization of spoilage markers in modified atmosphere packaged iceberg lettuce. International Journal of Food Microbiology, 279, 1–13. https://doi.org/10.1016/j.ijfoodmicro.2018.04.034 Jacxsens, L., Devlieghere, F., & Debevere, J. (2002). Predictive modelling for packaging design: Equilibrium modified atmosphere packages of fresh-cut vegetables subjected to a simulated distribution chain. International Journal of Food Microbiology, 73(2), 331–341. https://doi.org/10.1016/S0168-1605(01)00669-9 Jacxsens, L., Devlieghere, F., Ragaert, P., Vanneste, E., & Debevere, J. (2003). Relation between microbiological quality, metabolite production and sensory quality of equilibrium modified atmosphere packaged fresh-cut produce. International Journal of Food Microbiology, 83(3), 263–280. https://doi.org/10.1016/S0168-1605(02) 00376-8 Katoh, K., Misawa, K., Kuma, K., & Miyata, T. (2002). MAFFT: A novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Research, 30(14), 3059–3066. https://doi.org/10.1093/nar/gkf436 Kim, J. G., Luo, Y., Tao, Y., Saftner, R. A., & Gross, K. C. (2005). Effect of initial oxygen concentration and film oxygen transmission rate on the quality of fresh-cut romaine lettuce. Journal of the Science of Food and Agriculture, 85(10), 1622–1630. https:// doi.org/10.1002/jsfa.2158 I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 8 Krupka, M., & Piotrowicz-Cie´ slak, A. I. (2024). Optimization of the method for isolating bacterial DNA from the aboveground part of lettuce. International Journal of Molecular Sciences, 25(15), 8513. https://doi.org/10.3390/ijms25158513 Li, M., Shao, D., Zhou, J., Gu, J., Qin, J., Chen, W., & Wei, W. (2020). Signatures within esophageal microbiota with progression of esophageal squamous cell carcinoma. Chinese Journal of Cancer Research, 32(6), 755. https://doi.org/10.21147/j. issn.1000-9604.2020.06.09 Lin, H., & Peddada, S. D. (2020). Analysis of compositions of microbiomes with bias correction. Nature Communications, 11(1), 3514. https://doi.org/10.1038/s41467020-17041-7 Lorente-Mento, J. M., Valverde, J. M., Serrano, M., & Pretel, M. T. (2022). Fresh-cut salads: Consumer acceptance and quality parameter evolution during storage in domestic refrigerators. Sustainability, 14(6), Article 6. https://doi.org/10.3390/ su14063473 Magoˇ c, T., & Salzberg, S. L. (2011). FLASH: Fast length adjustment of short reads to improve genome assemblies. Bioinformatics, 27(21), 2957–2963. https://doi.org/ 10.1093/bioinformatics/btr507 Manthou, E., Coeuret, G., Chaillou, S., & Nychas, G.-J. E. (2022). Metagenetic characterization of bacterial communities associated with ready-to-eat leafy vegetables and study of temperature effect on their composition during storage. Food Research International, 158, Article 111563. https://doi.org/10.1016/j. foodres.2022.111563 Mira Miralles, M., Maestre-Carballa, L., Lluesma-Gomez, M., & Martinez-Garcia, M. (2019). High-throughput 16S rRNA sequencing to assess potentially active Bacteria and foodborne pathogens: A case example in ready-to-eat food. Foods, 8(10), 480. https://doi.org/10.3390/foods8100480 Novogene. (2025). 16S/18S/ITS Amplicon Metagenomic Sequencing. Retrieved from: https://www.novogene.com/eu-en/services/research-services/metagenomesequencing/16s-18s-its-amplicon-metagenomic-sequencing/. Accessed September 21, 2025. Oliveira, M., Abadias, M., Usall, J., Torres, R., Teixid´ o, N., & Vi˜ nas, I. (2015). Application of modified atmosphere packaging as a safety approach to fresh-cut fruits and vegetables – A review. Trends in Food Science & Technology, 46(1), 13–26. https:// doi.org/10.1016/j.tifs.2015.07.017 Paillart, M. J. M., van der Vossen, J. M. B. M., Levin, E., Lommen, E., Otma, E. C., Snels, J. C. M. A., & Woltering, E. J. (2017). Bacterial population dynamics and sensorial quality loss in modified atmosphere packed fresh-cut iceberg lettuce. Postharvest Biology and Technology, 124, 91–99. https://doi.org/10.1016/j. postharvbio.2016.10.008 Peng, H., Lavelle, D. O., Truco, M. J., Michelmore, R. W., & Simko, I. (2025). Differential low oxygen response and transcriptomic shifts drive fresh-cut lettuce deterioration in modified atmosphere packaging. Postharvest Biology and Technology, 227, Article 113571. https://doi.org/10.1016/j.postharvbio.2025.113571 Peng, H., & Simko, I. (2023). Extending lettuce shelf life through integrated technologies. Current Opinion in Biotechnology, 81, Article 102951. https://doi.org/10.1016/j. copbio.2023.102951 Pothakos, V., Snauwaert, C., De Vos, P., Huys, G., & Devlieghere, F. (2014). Psychrotrophic members of Leuconostoc gasicomitatum, Leuconostoc gelidum and Lactococcus piscium dominate at the end of shelf-life in packaged and chilled-stored food products in Belgium. Food Microbiology, 39, 61–67. https://doi.org/10.1016/j. fm.2013.11.005 Price, M. N., Dehal, P. S., & Arkin, A. P. (2009). FastTree: Computing large minimum evolution trees with profiles instead of a distance matrix. Molecular Biology and Evolution, 26(7), 1641–1650. https://doi.org/10.1093/molbev/msp077 Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., … Gl¨ ockner, F. O. (2013). The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Research, 41(Database issue), D590–D596. https:// doi.org/10.1093/nar/gks1219 R Core Team. (2024). R: A language and environment for statistical computing [computer software]. R foundation for statistical. Computing. https://www.R-project.org/. Ragaert, P., Devlieghere, F., & Debevere, J. (2007). Role of microbiological and physiological spoilage mechanisms during storage of minimally processed vegetables. Postharvest Biology and Technology, 44(3), 185–194. https://doi.org/ 10.1016/j.postharvbio.2007.01.001 Rognes, T., Flouri, T., Nichols, B., Quince, C., & Mah´ e, F. (2016). VSEARCH: A versatile open source tool for metagenomics. PeerJ, 4, Article e2584. https://doi.org/ 10.7717/peerj.2584 Sun, Y., Zhao, X., Ma, Y., Ma, Z., He, Z., Zhao, W., … Wang, D. (2022). Investigation on the microbial diversity of fresh-cut lettuce during processing and storage using high throughput sequencing and their relationship with quality. Foods, 11(12), 12. https://doi.org/10.3390/foods11121683 Tudela, J. A., Marín, A., Martínez-S´ anchez, A., Luna, M. C., & Gil, M. I. (2013). Preharvest and postharvest factors related to off-odours of fresh-cut iceberg lettuce. Postharvest Biology and Technology, 86, 463–471. https://doi.org/10.1016/j. postharvbio.2013.07.028 Wang, F., Men, X., Zhang, G., Liang, K., Xin, Y., Wang, J., Li, A., Zhang, H., Liu, H., & Wu, L. (2018). Assessment of 16S rRNA gene primers for studying bacterial community structure and function of aging flue-cured tobaccos. AMB Express, 8, 182. https://doi.org/10.1186/s13568-018-0713-1 Werner, J. J., Koren, O., Hugenholtz, P., DeSantis, T. Z., Walters, W. A., Caporaso, J. G., … Ley, R. E. (2012). Impact of training sets on classification of high-throughput bacterial 16s rRNA gene surveys. The ISME Journal, 6(1), 94–103. https://doi.org/ 10.1038/ismej.2011.82 Woltering, E. J., & Paillart, M. J. M. (2024). Effect of low oxygen modifiedand controlled-atmospheres on quality attributes and microbial population dynamics in fresh-cut iceberg lettuce. Acta Horticulturae, 1386, 169–174. https://doi.org/ 10.17660/ActaHortic.2024.1386.23 Yaptenco, K. F., Kim, J. G., & Lim, B. S. (2007). Gas transmission rates of commercially available polyethylene and polypropylene films for modified atmosphere packaging. Philippine Agricultural Scientist, 90(1). https://www.ukdr.uplb.edu.ph/journal-arti cles/2809. I. Ortega-Sanz et al. Food Research International 222 (2025) 117694 9