Development of a bacterial mixed culture preventing bacterial storage diseases of root vegetables

Abstract

The storage of root vegetables is a critical aspect of agricultural production, significantly impacting both the quality and quantity of produce available for consumption. This study aimed to develop a bacterial consortium capable of preventing storage bacterial diseases in root vegetables. A total of 225 bacterial strains were isolated from various ecological niches, including crops, peri-root soils, coastal dune soils, sewage sludge, and symbiotic bacteria of common insect pests. These strains were characterized for their ability to grow at storage temperatures (4 °C) and produce cellulase, lipase, protease, and chitinase enzymes. From 225 isolates, 24 strains exhibited the highest enzymatic activities were selected for mutual inhibition assays, resulting in the selection of four strains (SP65, Dv004, Dv008, and Dv024a) that demonstrated high growth rates and enzymatic activities. The selected strains grew at 4 °C (OD₆₀₀ range: 0.061–0,190) and they exhibited the high intensity of production of the following enzymes measured as a halo diameter (cm) on bacterial medium: cellulase (from 1,4 to 1,8 cm), lipase (0,7) − 1,5), proteinase 1,2 − 1,6) and chitinase (0,3 − 0,4). These strains, identified as Stenotrophomonas maltophilia (SP65) and Serratia liquefaciens, were combined to form a bacterial consortium. The efficacy of this consortium was tested on potato (Solanum tuberosum L.), carrot (Daucus carota L.), celeriac (Apium graveolens L.), parsley (Petroselinum crispum Mill.), and onion (Allium cepa L.). The mixed culture was tested against bacterial soft rot pathogens including Pectobacterium carotovorum, Kosakonia cowanii, Enterobacter cloacae subsp. dissolvens and Burkholderia cepacia. For celeriac and parsley, the effect of inhibition of Pectobacterium carotovorum, Burkholderia cepacia, E. cloaceae subsp. dissolvens, Kosakonia cowanii, and the mixture of those pathogens were observed when pathogens were applied at concentration of 1,3 × 10⁸ CFU mL⁻¹. In case of onion the same effect of inhibition of growth of pathogens were observed only for P. carotovorum, and for B. cepacia at concentration of 1,3 × 10⁸ CFU mL⁻¹. The results demonstrated the potential of using a bacterial consortium as a biological control method to enhance the storage life and quality of root vegetables.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s11274-026-05279-0.

Introduction

Poland is one of the leading producers of root vegetables in Europe (Eurostat 2024) with total area of cultivation of over 191,50 thousand hectares and the harvest volume of 4,1 million tons (Eurostat 2025; Trading Economics 2025) and is the third-largest producer of carrots in the European Union, accounting for about 12% of total EU production (Eurostat 2025). The latest Polish research on average annual consumption of fresh fruit and vegetables (Goryńska-Goldmann 2024), show that in the years 2019–2021, in Poland it was 200.1 kg per capita. It was also stated, that more than 22% of root vegetables, available for consumption in Poland, comes from long-term storage (Grzegorzewska et al. 2014).

To obtain a stable income and ensure a constant supply of vegetables to the market, food producers need an appropriate storage base that enables quick delivery of high-quality goods, increasing the flexibility of companies in response to changing market demands. Currently, the burden of storing vegetables rests mainly on producers. Exceeding recommended storage times for root and bulb vegetables such as potatoes, carrots, parsley, celeriac, and onions leads to significant quality deterioration including water loss, tissue softening, discoloration, sprouting (notably in onions), and increased susceptibility to fungal and bacterial diseases. These changes cause weight loss, reduced firmness, rot, and overall product decay, resulting in postharvest losses that can range from a few percent under optimal conditions to over 30–40% in poor storage environments, severely impacting commercial value and profitability (Ilić et al. 2016; Teshome and Leweye 2020). Proper postharvest handling techniques like curing (especially for onions), optimal harvest timing, washing treatments (e.g., with H₂O₂ or NaOCl), and maintaining low temperatures with high relative humidity are effective in preserving quality and reducing losses during storage (Ilić et al. 2016; Kiura et al. 2021).

In addition to unavoidable natural losses, very large losses in the mass of vegetable crops can occur as a result of the action and development of storage diseases, which, in addition to quantitative losses, also cause a deterioration in the quality of the crop, which directly translates into the profitability of production. A significant part of storage diseases is caused by bacteria, which are a big problem in the cultivation of onions (Allium spp.), iceberg lettuce (Lactuca spp.), celeriac (Apium spp.), parsley (Petroselinum spp.), carrots (Daucus spp.) and potatoes (Solanum spp.). Storage diseases of vegetables can cause significant yield losses, especially as vegetable varieties grown in Poland are highly variable in their resistance to these diseases. Storage diseases cannot be eliminated or cured, but they can be prevented. Therefore, prevention and control of storage diseases should be one of the main elements of the comprehensive cultivation, harvesting and overall handling of vegetables (Sosnowska 2011).

In the case of bacterial storage diseases, it is difficult to determine the primary pathogenic factor because almost immediately, in addition to the primary infectious factor (cause of the disease), secondary pathogenic factors also begin to act, such as fungi, causing so-called mixed infections, and opportunistic bacterial plant pathogens. At the Institute of Plant Protection - National Research Institute, preliminary studies were conducted to better understand the factors contributing to the development of wet rots caused by bacteria, occurring in storage. The results suggest that, in addition to the obvious role of already known pathogens, bacterial species belonging to opportunistic pathogens play a significant role in the development of storage diseases. This means that in natural conditions these bacteria do not cause plant diseases, but in conditions favorable to them, such as poorly managed vegetable storage, or access to damaged crops, these bacteria can cause the development of wet rots, as a secondary pathogen, a component of mixed rots, thus increasing crop losses.

Therefore, the aim of the work is to: (1) identify the cause of bacterial storage diseases by determining the species composition of the population of bacteria causing wet rot of vegetables, (2) verify the research hypothesis assuming that, in addition to known plant pathogens, bacterial species that are not plant pathogens in field conditions, but may contribute to the development and development of wet rot in storage conditions, play a significant role in the development of storage diseases, (3) select bacterial strains exhibiting specific biochemical and physiological features that meet the theoretical assumptions of the consortium being developed, (4) attempt to reduce the occurrence of wet rot by developing the composition of a bacterial consortium that prevents the occurrence of storage diseases and experimentally verify the operation of the developed consortium.

A novelty for science will be a better understanding of the process of development of storage diseases, which may contribute to the improvement of existing principles and recommendations of integrated plant protection for storage, which in practice should contribute to a significant reduction in growers’ losses and improvement of food safety.

Materials and methods

Plant materials

The experimental plant material consisted of potato (Solanum tuberosum L.), carrot (Daucus carota L.), parsley (Petroselinum crispum Mill), celeriac (Apium graveolens L. var. rapaceum), and onion (Allium cepa L.). The choice of these species was based on their relevance to long-term storage practices in Poland and their economic significance. Vegetables used in the study were purchased from local retail stores in Poland to reflect typical consumer supply conditions in the region. Before experiments, the plants were cut into slices of 4–5 mm thick, and disinfected by immersing for 1 min, in solutions of 5% sodium hypochlorite (NaClO) followed by 70% ethyl alcohol (C₂H₅OH). Residual disinfectants were removed by rinsing three times in sterile distilled water (SDW), and the water from the final, third washing was screened for the presence of bacteria (Srivastava et al. 2024).

Obtaining bacterial strains for the study

Bacterial strains used in the study were obtained from several ecological niches including tissues of crop plants, peri-root soil of these plants, insect tissues of selected crop pests and sewage sludge selected as environments rich in microbial diversity and potential antagonistic bacteria. Sewage sludge is a microbially complex substrate containing high organic load and diverse microbial communities. Previous studies have shown that bacteria isolated from such environments often possess broad extracellular enzymatic activities, including proteolytic, cellulolytic and oxidative enzymes, reflecting adaptation to high competition and diverse substrates (Robledo-Mahón et al. 2020). The control strains, meaning the strains causing decay, were obtained from crop plants showing disease symptoms indicating the possibility of bacterial disease development (e.g. soft rot). The detailed description of bacteria isolation and identification is given below.

Plant-related endophytic bacterial strains were isolated from the tested plants in the following way: surface of the leaves, stems and roots was disinfected by immersing for 1 min, in solutions of 5% sodium hypochlorite (NaClO) followed by 70% ethyl alcohol (C₂H₅OH). Residual disinfectants were removed by rinsing three times in sterile distilled water (SDW) (Srivastava et al. 2024). Plant tissues were then homogenized in sterile homogenization pouches (Bioreba, Reinach, Switzerland) containing 3 ml of sterile saline (0.9% NaCl w/v). The resulting homogenates were diluted with SDW (1:10 v: v) and aliquots of 0.1 ml were inoculated into Petri dishes containing non-selective Tryptic Soy Agar (TSA) (Sigma Aldrich LCC, Darmstadt, Germany), spread evenly with bacteria spreader and incubated at 27℃ for 48 h. The effectiveness of the sterilization process was verified by inoculating 0.1 ml of SWD from last wash onto the TSA medium. No bacterial growth from this step was considered as proof of efficient sterilization process.

Bacterial strains from imago of maize beetle (Chrysomelidae: Diabrotica virgifera), were obtained using the following procedure. The specimens of maize beetle were immersed in a 70% ethanol solution for 3 min, for disinfection. Next, the residual ethanol was washed away by triple rinsing in SDW. The effectiveness of the sterilization process was verified by inoculating 0.1 ml of SDW from the last wash onto TSA medium. No bacterial growth was observed, confirming efficient surface sterilization. One sample consisted of 10 specimens pulled together and homogenized in a sterile porcelain mortar containing 9 ml of sterile saline. Subsequently, from each sample, aliquots of 0.1 ml of the homogenate were inoculated on Petri dishes with solid TSA medium, spread with a cell spreader, and incubated at 27 °C for 48 h. Pure bacterial cultures were obtained, by a series of reductional streaks on TSA medium.

Rhizosphere strains were isolated by sampling soil from the zone up to 1 cm away from the roots, directly influenced by root exudates. An aliquot of 10 g of soil, from each collected sample, was placed in a flask containing 90 ml of sterile saline with 0.1% Tween 80 (Bioshop, Burlington, Canada), shaken for 5 min at 100 rpm to wash the bacterial cells out of the soil particles. After shaking, the flasks were left for 15 min, allowing larger soil particles to settle to the bottom. The resulting supernatant was treated as a starting (10 ⁰) solution from which the decimal dilutions up to 10^− 9 were prepared. The aliquots of 0,1 ml of decimal dilutions ranged from 10^− 4 to 10^− 9, were streaked in triplicate, into Petri dishes with TSA medium, and incubated at 27 °C for 48 h.

For all tested samples, after incubation, morphotypes were distinguished from the grown bacterial colonies. Next, a series of reductional strikes were performed until the pure culture of each morphotype, henceforth strain, was obtained. The purity of culture of each strain was confirmed using Gram staining procedure.

Plant pathogenic bacterial strains causing bacterial soft rot symptoms on the tested plants were obtained from the collection of the Department of Virology and Bacteriology and were used as experimental controls. Strain were obtained from potato: Kosakonia cowanii (CH51), from carrot: Enterobacter cloacae subsp. dissolvens (Edi01) and Pectobacterium carotovorum (Pc), and one from onion: Burkholderia cepacia (B3218).

Pathogenicity of bacterial strains on plants

All obtained strains were submitted to pathogenicity test on slices of potatoes, carrots, parsley, celeriac and onion, using a “slice test” technique (Łojkowska 1991), to test their ability to cause soft rot symptoms. The 5 mm thick slices of vegetables were disinfected by soaking them twice, for 10 minutes, in 0.5% sodium hypochlorite solution, rinsing once with SDW, spraying with 70% ethanol, and rinsing three times with sterile distilled water to remove ethanol residues. After sterilization, the vegetables were dried and inoculated in the center of each slice, with 25 µl of a bacterial pathogen suspension at concentrations of 5.1 × 10⁶, 5 × 10⁷ and 5 × 10⁸ CFU mL⁻¹. Inoculated slices were placed in Petri dishes containing sterile Whatman filter paper discs soaked in SDW to maintain humidity during incubation conducted at 22℃ for 3 days. After incubation, the percentage of infested areas in each slice was determined. Inoculation of slices was performed in a series of repetitions, with slices in each repetition. The soft-rot-causing bacterial strains used in the experiments were obtained from internal collection of the Department of Virology and Bacteriology of the Institute of Plant Protection-National Research Institute. Those strains were used as an experiment control to cause soft rot symptoms on vegetable slices.

Criteria for bacterial strains selection

All plant endophytic bacterial strains obtained in this study were screened for the presence of five biochemical properties considered relevant for their survival and potential application: ability to grow under storage conditions (4 °C), and production of cellulase, lipase, protease, and chitinase enzymes. Chitinase activity was included in the screening not because of its direct relation to bacterial storage diseases, but due to its possible role in future biocontrol applications against fungal pathogens.

The ability to colonize vegetables during storage was evaluated by testing cellulase production on carboxymethylcellulose medium (CMC; (Sreedevi et al. 2013); TSB 30 g/l, agar 12 g/l, CMC 10 g/l, pH 7.0), lipase on lipoprotein medium (LP; peptone 10 g/l, NaCl 5 g/l, CaCl₂ 0.1 g/l, agar 15 g/l, Tween 20 10 ml/l, pH 7.0), and protease on skim milk agar (SMA; skim milk 15 g/l, yeast extract 0.5 g/l, agar 9 g/l, pH 7.0)(Ghodsalavi et al. 2013). Antifungal potential was assessed through chitinase activity on chitin agar medium (CHA; colloidal chitin 2 g/l, 0.07 g K₂HPO₄ 0,07 g\l, KH₂PO₄ 0.03 g\l, MgSO₄.7H₂O 0.05 g \l, FeSO₄ × 7H₂O 0.01 g\l, ZnSO₄ 0.001 g\l, MnCL₂ 0.001 g\l, agar 15 g\l, (pH = 7.0) with enzyme production quantified as the ratio of the halo zone diameter to the colony diameter (Fadhil et al. 2014).

For all biochemical tests, 100 µl aliquots of bacterial suspensions (10⁶ CFU mL⁻¹) were inoculated on the respective media. Growth at 4 °C was tested in Tryptic Soy Broth (TSB, Sigma Aldrich, Poland) for 72 h, with bacterial proliferation evaluated spectrophotometrically (OD₆₀₀) and by colony counts on TSA. Strains demonstrating rapid growth (OD₆₀₀ increase of ≥ 0,15 after 24 h and ≥ 100 cfu per plate) were considered suitable for further testing.

Additionally, antagonistic interactions between strains were evaluated using the double culture technique. Only strains fulfilling all the above criteria were selected for the bacterial consortium.

Development of bacterial consortium and testing its effectiveness

The bacterial consortium was formed from strains selected as described above. The consortium was formed by combining aliquots of 0,1 ml of aqueous suspensions of 10⁷ CFU mL⁻¹ of each strain. Performance of the consortium was evaluated on slices of vegetables prepared as described above. The efficiency of the consortium created to prevent storage diseases was tested in two experiments. In first, both consortium and plant pathogenic bacteria, were applied together onto the same vegetable slices. In the second experiment, the consortium suspension was applied on the vegetable slices and incubated for 24 h, next the suspension of pathogenic bacteria was applied.

Identification of the tested bacterial strains

Bacterial strains expressing all five biochemical and physiological traits simultaneously were selected as components of the bacterial consortium. These bacterial strains were identified using the Biolog Gen III system (Biolog Inc., Hayward, CA, USA) (v 2.8.0), according to manufacturer instructions, and by sequencing of the 16 S rRNA. Bacterial genomic DNA was extracted using a modified CTAB protocol (Doyle 1991). One microliter of the DNA was used as the template in each PCR reaction, performed with GoTaq^® Green Master Mix (Promega, Madison, WI, USA). The 16SA1/16SB1 primer pair (Kikuchi 2009) was used to obtain a product of approximately 1500 bp The obtained nucleotide sequences were aligned using BioEdit software (v7.2)(Hall 1999) to generate consensus sequences, each derived from three sequencing reads. These consensus sequences were subsequently identified using the BLAST tool (https://blast.ncbi.nlm.nih.gov/) and deposited in the GenBank database (http://www.ncbi.nlm.nih.gov/).

Statistical analysis

All experimental measurements (optical density at 600 nm for bacterial growth and halo zone diameters for enzymatic activities) were performed in triplicate. The biological data were processed, analyzed, and visualized using Statistica software (v. 12, StatSoft, Tulsa, OK, USA). Non-parametric descriptive statistics, including medians, interquartile ranges (IQR), and minimum/maximum values, were utilized to characterize and compare the distribution of enzymatic activities and growth rates of the isolated bacterial strains across different ecological niches. Outliers were determined using standard software criteria and plotted as individual points. For the pathogen-inhibition assays on vegetable slices, results were recorded qualitatively as the presence (Y) or absence (N) of soft rot symptoms based on three independent experimental replicates.

Results

Research material

In the study, a total of 225 bacterial strains were obtained from a variety of ecological niches, such as crops: winter wheat (Triticum sp. L.) (12 strains), winter oilseed rape (Brassica napus L.) (18 strains); peri-root soil of plants: knotweed (Reynoutria Houtt ) (11 strains), nettle (Urtica dioica L.) (13 strains), sorrel (Rumex sp. L.) (4 strains), plantain (Plantago major L.) (3 strains), grasses (Poaceae (R. Br.) Barnh, Gramineae Juss.) (11 strains); rapeseed extra-root soil (6 strains), rhizosphere soil (3 strains), rhizoplane soil (5 strains), coastal dune soil (8 strains); sewage sludge from sewage treatment plants (37 strains); symbiotic bacteria of common insect pests of plants: maize beetle (Diabrotica virgifera) (39 strains), horsetail beetle (Oulema sp.) (39 strains) and European corn borer (Ostrinia nubilalis) (16 strains) (Table S1).

Bacterial strains characterization

All obtained bacterial strains were examined for the presence of: (1) ability to grow in storage temperatures (4℃), (2) ability to produce cellulase, (3) lipase, (4) protease and (5) chitinase enzymes (Fig. 1, 2, 3, and 4, respectively). For each test, the group of strains with the 10% highest value of results, either OD₆₀₀ or the halo diameter in cm, was considered as strains highly expressing the tested feature. All results with the selection of 10% of the highest results are presented in supplemental file (Table S1), and in Figs. 1, 2, 3, and 4, while the final strains selected for the consortium are summarized in Table 1.

Fig. 1.

The activity of cellulase enzyme of the tested bacterial strains obtained from various biological sources. The cellulase activity was measured on CMC medium. The y-axis shows the halo diameter in centimetres around bacterial colonies, indicating cellulase activity. The x-axis represents the origin of the bacterial strains. The boxplots depict the median, interquartile range, and minimum/maximum values, with outliers shown as individual points

Fig. 2.

The activity of lipase enzyme of the tested bacterial strains obtained from various biological sources. lipase activity was measured on LP medium. The y-axis shows the halo diameter in centimetres around bacterial colonies, indicating lipase activity. The x-axis represents the source of the bacterial isolates. The boxplots display the median, interquartile range, and minimum/maximum values, with outliers indicated as individual points

Fig. 3.

The activity of protease enzyme of the tested bacterial strains obtained from various biological sources. The protease activity was measured on SMA medium. The y-axis shows the halo diameter in centimetres around bacterial colonies, indicating protease activity. The x-axis represents the source of the bacterial strains. The boxplots display the median, interquartile range, and minimum/maximum values, with outliers indicated as individual points

Fig. 4.

The activity of chitinase enzyme of the tested bacterial strains obtained from various biological sources. The chitinase activity was measured on CHA medium. The y-axis represents the halo diameter in centimetres around bacterial colonies, indicating chitinase activity. The x-axis shows the source of the bacterial strains, which include Diabrotica virgifera and sewage sludge. The boxplots display the median, interquartile range, and minimum/maximum values, with outliers shown as individual points

Table 1.

Four strains (SP65, Dv004, Dv008, and Dv024a), exhibiting simultaneously a high rate of growth at 4℃ and the high activity of cellulase, lipase, protease and chitinase enzymes, were selected comprise the bacterial consortium. Taxonomic identity of the strains was confirmed as Stenotrophomonas maltophilia (SP65) and Serratia liquefaciens (Dv004, Dv008, Dv024a), (PV037668, PX460028, PX460019, PX460015)

No	Strain identification (16 S rRNA)	Strain origin	Strain code	Growth at 4℃ (OD600)	Halo diameter on bacterial medium [cm] *
					CMC	LP	SMA	CHA
33	Stenotrophomonas maltophilia	Sewage sludge	SP 65	0.114	1.7	0.7	1.5	0.5
38	Serratia liquefaciens	Diabrotica virgifera	Dv 004	0.148	1.8	1.5	1.6	0.3
43	Serratia liquefaciens	Diabrotica virgifera	Dv 008	0.061	1.8	1.2	1.2	0.4
44	Serratia liquefaciens	Diabrotica virgifera	Dv 024	0.190	1.4	0.7	1.5	0

Within the bacteria with the highest growth rate at 4℃, measured by the OD₆₀₀ value, 67 strains were recorded with OD₆₀₀ value ranging from 0,298 to 0,061. Of those strains 32 were isolated from D. virgifera, 11 were isolated from Ostrinia nubilalis, 4 were isolated from Oulema sp., and 20 from sewage sludge (Fig. S1).

70 strains with the highest cellulase activity, measured by the size of halo diameter on CMC medium, were recorded. The halo diameter range was from 1,2 cm to 3,4 cm. In this group, 12 strains were isolated from B. napus (both endophytes and soil), 7 strains from D. virgifera, 14 strains from O. nubilalis, 4 strains from Oulema sp., 5 strains from Poaceae (both endophytes and soil), 3 strains from Reynoutria sp., 1 strain from Rumex sp. 15 strains from sewage sludge, 6 strains from T. aestivum, and 2 strains from Urtica sp. Notably, only one strain isolated from Rumex sp. exhibited cellulase activity under the applied assay conditions (Fig. 1). The absence of whiskers in some box plots (Poaceae, Reynoutria sp., Urtica sp.) (Fig. 1) is due to the low variability of halo diameters within these groups, where most values were close to zero and only a few outliners exceeded the upper quartile.

The highest lipase activity, measured on LP medium, was recorded for 68 strains. The halo diameter range was from 0,4 cm to 1,6 cm. In this group, 4 strains were isolated from B. napus (both endophytes and soil), 19 strains from D. virgifera, 15 strains from O. nubilalis, 5 strains from Oulema sp., 1 strain from Poaceae (endophyte), 22 strains from sewage sludge, and 2 strains from Urtica sp. Notably, only one strain derived from Poaceae exhibited lipase activity under the applied assay conditions (Fig. 2).

The highest protease activity, measured on SMA medium, was observed for 75 strains, with the halo diameter on medium ranging from 2,7 to 1,2. In this group, 13 strains were isolated from B. napus (both endophytes and soil), 15 strains from D. virgifera, 6 strains from O. nubilalis, 2 strains from Oulema sp., 4 strains from Poaceae (both endophyte and soil), 22 strains from sewage sludge, 9 strains from T. aestivum, and 4 strains from Urtica sp (Fig. 3).

For the strains with the highest chitinase activity, measured on CHA medium, 29 strains were recorded, with the halo diameter on CHA medium ranging from 0,6 to 0,2. In this group, 18 strains were isolated from D. virgifera, and 9 strains from sewage sludge (Fig. 4).

In the next step, from the total collection of isolates, we selected those representing the top 10% of results for each tested biochemical parameter (i.e., isolates showing the highest activity values within the 90th percentile range; Figs. 1, 2, 3 and 4) we selected those that exhibit a minimum 4 out of the tested 5 biochemical qualities and recorded their ability to lack of mutual inhibition. Ten of these were found to exhibit antagonistic properties towards other strains (Fig. S2). Bacterial strains from sewage sludge inhibited the growth of strains isolated from maize beetle. Strains showing antagonistic properties towards other tested strains were excluded from further testing.

Development of the consortium composition and verification of its performance

Based on growth at 4 °C and pronounced extracellular enzymatic activities (cellulase, lipase, protease, and chitinase), four bacterial strains (SP65, Dv004, Dv008, and Dv024a) were selected for consortium development (Table 2) (Table S1).

Table 2.

Verification of the consortium ability to inhibit activity of tested plant pathogenic bacteria. The table shows ability of the developed consortium (10⁷ CFU mL⁻¹ ) to inhibit the growth of individual plant pathogenic bacterial strains and their mixture. Taxonomic identity of the pathogenic strains was confirmed as Pectobacterium carotovorum (Pc), Burkholderia cepacia (B3217), Enterobacter cloacae subsp. dissolvens (Edi), Kosakonia cowanii (CH51) (PZ315353, PZ315355, PZ315356, PZ315354)

Tested plant	Pathogen inoculum: 0,1 ml	Pectobacterium carotovorum (Pc)		Burkholderia cepacia (B3217)		Enterobacter cloacae subsp. dissolvens (Edi)		Kosakonia cowanii (CH51)		Mix of the pathogens (Mix)
		Time since inoculation with the consortium
		0	24	0	24	0	24	0	24	0	24
Potato	Pathogen inoculum 1,3 × 108 CFU mL⁻¹	N	Y	Y	Y	N	Y	N	Y	Y	Y
	Pathogen inoculation 1,2 × 107 CFU mL⁻¹	N	Y	N	Y	Y	Y	N	Y	Y	Y
Carrot	Pathogen inoculum 1,3 × 108 CFU mL⁻¹	Y	Y	Y	N	Y	Y	Y	N	N	Y
	Pathogen inoculation 1,2 × 107 CFU mL⁻¹	Y	N	Y	Y	Y	N	Y	N	N	N
Celeriac	Pathogen inoculum 1,3 × 108 CFU mL⁻¹	N	N	N	N	N	N	N	N	N	N
	Pathogen inoculation 1,2 × 107 CFU mL⁻¹	N	N	N	N	N	N	N	N	N	N
Parsley	Pathogen inoculum 1,3 × 108 CFU mL⁻¹	N	N	N	N	N	N	N	N	N	N
	Pathogen inoculation 1,2 × 107 CFU mL⁻¹	N	N	N	N	N	N	N	N	N	Y
Onion	Pathogen inoculum 1,3 × 108 CFU mL⁻¹	N	N	N	N	N	Y	N	Y	N	Y
	Pathogen inoculation 1,2 × 107 CFU mL⁻¹	N	Y	N	N	N	Y	N	Y	N	Y

*Y  inhibition observed (no disease symptoms), N no inhibition observed (disease symptoms present)

The efficacy of the developed bacterial consortium was tested on potato (Solanum tuberosum L.), carrot (Daucus carota L.), celeriac (Apium graveolens L.), parsley (Petroselinum crispum Mill.) and onion (Allium cepa L.) (Table 1).

The best activity in inhibition of development of storage diseases was observed for celeriac and parsley (Table 1). The development of rotting symptoms has stopped. A simultaneous application of pathogen suspension and the consortium, and application of the pathogen suspension 24 h after the consortium suspension resulted in complete inhibition of disease symptoms (Figs. 5 and 6).

Fig. 5.

Example of inhibition of storage disease symptoms, by the tested bacterial consortium, on celeriac. A – Rot symptoms caused by the presence of mixture (10⁸ CFU mL⁻¹) of the control pathogens (Pectobacterium carotovorum (Pc), Burkholderia cepacia (B3217), , Enterobacter cloacae subsp. dissolvens (Edi 01), Kosakonia cowanii (CH51) ; B – Symptoms caused by the presence of mixture of pathogens (10⁸ CFU mL⁻¹), mixed with developed consortium (10⁷ CFU mL⁻¹); C – The lack of symptoms caused by the presence of mixture of pathogens (10⁷ CFU mL⁻¹), mixed with developed consortium (10⁷ CFU mL⁻¹)

Fig. 6.

Example of inhibition of storage disease symptoms, by the tested bacterial consortium, on celeriac. A – Rot symptoms caused by the presence of mixture (10⁸ CFU mL⁻¹) of the control pathogens (Pectobacterium carotovorum (Pc), Burkholderia cepacia (B3217), , Enterobacter cloacae subsp. dissolvens (Edi 01), Kosakonia cowanii (CH51); B – Symptoms caused by the presence of mixture of pathogens (10⁸ CFU mL⁻¹), mixed with developed consortium (10⁷ CFU mL⁻¹); C – The lack of symptoms caused by the presence of mixture of pathogens (10⁷ CFU mL⁻¹), mixed with developed consortium (10⁷ CFU mL⁻¹)

Discussion

In the climatic conditions of Poland, a vegetable storage plays an extremely important role in agriculture. It is essential for maintaining a regular and constant supply of fresh vegetables throughout the year. Achieving this objective requires maintaining high quality of vegetables after harvesting. This is accomplished by creating appropriate storage conditions to keep the vegetable tissue alive and to inhibit the progression of aging processes as much as possible (Grzegorzewska et al. 2014). Storage preserves the high quality and biological value of the harvested crop over a long period but involves the risk of quantitative and qualitative losses. These losses are associated with natural degradation and with the development of storage diseases, in which bacteria play a significant role. Bacterial diseases are a major problem in the cultivation of potatoes, carrots, celeriac, parsley, and onions. These diseases cannot be cured and must be prevented. Despite many measures and factors aimed at reducing the incidence of storage diseases, rot remains an important cause of yield losses (Sosnowska 2011).

Since stored vegetables are directly consumed by humans, often in their raw state, the use of aggressively acting chemicals to prevent storage diseases is severely limited. For this reason, integrated pest management techniques, particularly bacterial-based biological control methods, are gaining importance. In recent years, there has been an increase in scientific publications highlighting the significant role of microbiome plants, which directly supports the increased use of bacteria in agriculture.

One of the hypotheses arising from research on plant microbiome is that species diversity among endophytic bacteria, protects plant’s natural microbiome against unstable climatic conditions. This principle is known as the ‘insurance hypothesis’ and has been observed in various ecosystems. For example, an opportunistic pathogen Pseudomonas aeruginosa has been shown to exhibit extensive genetic changes when functioning as a component of a biofilm, also known as a biological membrane. These changes are induced by a genetic mechanism involving the recA gene and affect various bacterial traits, including bacterial function in the biofilm. Some bacterial variants induced by such changes show an increased ability to spread, while others appear to target biofilm formation and maintenance. The presence of these functionally diverse bacteria enhances the ability of biofilms to resist environmental stress (Boles et al. 2004). Such discoveries allow us to conclude that diversity, self-generated by biofilms, represents a form of biological ‘insurance,’ helping to protect the bacterial community in the face of adverse environmental conditions.

In the present study, the developed bacterial consortium comprises Stenotrophomonas maltophilia (strain SP65) and Serratia liquefaciens (strains Dv004, Dv008, and Dv024a). Both bacterial species are increasingly recognized as efficient biological control agents due to their antagonistic properties and ecological adaptability. Members of the genus Serratia, particularly S. liquefaciens, have demonstrated substantial efficacy in controlling post-harvest rots and plant pathogenic fungi under low-temperature storage conditions, primarily mediated by the robust secretion of extracellular hydrolytic enzymes and volatile organic compounds (Stanley et al. 1994). Similarly, S. maltophilia has been reported as a promising biocontrol candidate capable of suppressing destructive soil-borne bacterial pathogens of root crops, such as Ralstonia solanacearum in potato (Messiha et al. 2007).

Remarkably, the cooperative and synergistic application of these two specific taxa is supported by recent literature. Osman et al. (2023) demonstrated that the dual application of S. maltophilia and S. liquefaciens significantly suppressed white rot disease caused by Stromatinia cepivora in onion (Allium cepa L.) while simultaneously promoting plant growth, crop yield, and activating the plants’ antioxidant defense mechanisms. This is strongly congruent with our findings, where the combined application of S. maltophilia and S. liquefaciens exhibited exceptional synergistic protective effects, completely inhibiting soft rot symptoms in celeriac and parsley. The multi-enzymatic activity profile of our selected strains—spanning cellulases, lipases, proteases, and chitinases—likely underpins their ability to competitively colonize host tissues and actively restrict the expansion of aggressive soft-rot-causing pathogens like Pectobacterium carotovorum and Kosakonia cowanii under storage temperatures.

Another concept emerging from plant bacterial research is the notion of a ‘holobiont.’ This term describes a cluster of species interacting with each other and associated with plants. A selective pressure acting on the components of the holobiont have likely formed plant-associated microbes, creating host-adapted microbial communities that influence plant fitness in many ways. High microbial densities detected in plant tissues, coupled with short microbial generation times and longer phylogenetic lineage compared to the host, suggest that microbial interactions are important selection force. These interactions shape complex microbial assemblages in the phyllosphere, rhizosphere, and plant endosphere. Reductionist approaches in laboratory experiments have been crucial for deciphering the strategies used by specific microorganisms to cooperate and compete within or outside plant tissues. However, our understanding of microbial interactions in the formation of more complex plant-associated microbial communities, and their relevance to host health in a more natural context, remains limited. This implies that interactions between members of the microbiota are fundamental for establishing and maintaining host-microbe homeostasis (Hassani et al. 2013).

Individual species of endophytic bacteria tend to colonize specific parts of plants. Some species colonize mainly roots, others stems and roots, while some can colonize all parts of the plant. Based on these findings, an experiment was conducted where a selected endophytic bacterium, known to colonize multiple plant hosts, was sprayed on plant flowers. This approach aimed to ensure that the bacterium, which has a protective effect on the plant, could colonize all parts of the plant and be present in the seeds (Santoyo et al. 2016). The importance of bacteria in maintaining homeostasis is evident from studies showing that, in addition to their well-known role in stimulating plant growth, numerous species of soil bacteria found in the plant rhizosphere or around plant tissues also reduce nematode populations through antagonistic behavior. Our results, congruent with the studies described above, suggest that the presence of S. marcescens in the consortium can have a significant meaning for colonization of vegetable tissues and for the protection against bacterial pathogens, on celiac and parsley.

Traditionally, plant diseases have been thought to be caused by a single species or even a specific strain of pathogen. However, recent insights emphasize that microorganisms in nature exist as part of complex communities, a concept known since the early days of microbiology but gaining increasing recognition. While classical microbiology often focuses on single microbial strains grown in isolation, it is now understood that many diseases in humans, animals, and plants result from synergistic interactions between multiple species. This complexity not only complicates disease diagnosis but also highlights the need to consider these interactions when developing more effective phytosanitary control measures. Although reports of synergistic pathogen-pathogen interactions in plant diseases are limited, and their mechanisms remain largely unknown, these interactions may be more common than previously thought. Understanding these underlying mechanisms is crucial for effective epidemiology and phytosanitary risk management. Recent advances in diagnostic technologies are beginning to shed light on the microbial communities associated with complex plant diseases, offering valuable insights into how these communities contribute to disease development (Du et al. 2025; Wassermann et al. 2022).

An additional important aspect concerns the interaction between the introduced inoculum and the indigenous microbial communities naturally present on vegetable surfaces. Unprocessed plant material harbors complex autochthonous microbiota that may compete with or modulate the activity of the applied bacteria (Jackson et al. 2015). The inoculum used in this study is therefore expected to function within an existing microbial ecosystem, potentially altering community structure in a transient manner.

The interaction between the applied inoculum and the indigenous microbial communities present on vegetables was not investigated in this study. Further research is required to assess potential shifts in native microbiota and to evaluate the persistence and efficacy of the inoculum under non-sterile, real storage conditions.

To sum up, our results offer significant applications and scientific potential. It could pave the way for developing new bacterial consortia with similar functionalities or for optimizing the performance of existing consortia. On the other hand, the novel aspect of our research lies in the discovery that bacteria not associated with storage environments, and even with vegetables, can be effectively used to protect stored vegetables against bacterial plant pathogens.

Conclusions

The present study demonstrated the successful development of bacterial consortium capable of effectively preventing storage bacterial diseases of root vegetables. Among 225 bacterial strains screened, four strains were selected based on their ability to grow at low temperatures and produce key hydrolytic enzymes. The resulting consortium exhibited strong antagonistic activity against major bacterial pathogens responsible for wet rot during storage.

Experiments confirmed the effectiveness of the developed consortium in significantly inhibiting disease development in all tested vegetables, with complete suppression of rotting symptoms in celeriac and parsley. These findings confirm that bacterial species not typically recognized as plant pathogens under field conditions can play a substantial role in storage- related diseases, and that their interactions can be effectively controlled through beneficial microbial consortia.

The results provide a new perspective on the microbiological mechanisms of storage diseases and offer a promising biological control strategy for reducing postharvest losses and improving the storage quality of root vegetables. The application of this consortium may contribute to the development of improved integrated plant protection systems for storage, enhancing both food safety and sustainability in agricultural production.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by A.H. with assistance from W.Z. The first draft of the manuscript was written by K.K. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

No funding was received to assist with the preparation of this manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declaration

Competing interests

The authors declare no competing interests.