Introduction

Clubroot, caused by Plasmodiophora brassicae, has been reported as the most devastating disease of cruciferous crops worldwide (Dixon, 2009; Karling, 1968; Zahr et al., 2022; Zheng et al., 2020). Yield losses from 30 to 100% have been reported in canola (Faggian & Strelkov, 2009) and, reviewing the Brassica species from the works cited in Saharan et al. (2021a) in the list of crop losses caused by clubroot, the incidence in cruciferous vegetables worldwide ranges from 3 to 86%. Broccoli (Brassica oleracea var. italica), a vegetable from the Brassicaceae family is grown by small farmers in the Andean region of Colombia and is one of the crops produced in the peasant and family farming systems, along with other Brassicaceae species commonly consumed in the country, such as cabbage and cauliflower. Colombia produces about 15,000 tons of broccoli on around 850 ha, with an average yield of 18.0 ton ha−1 (GOV.CO, 2020 on line), which is lower than the 20.0 ton ha−1 that is reported for the largest producers of broccoli, China and India (FAOSTAT, 2021). The production of broccoli and the other cruciferous in Colombia is threatened by the incidence of clubroot disease, which often causes growers to migrate to areas free of the disease. Yield losses in cabbage, cauliflower and broccoli crops in Colombia are close to 90% (Castillo & Guerrero, 2008), while in Chinese cabbage, this disease is responsible for 20 to 50% yield losses (Jaramillo & Díaz, 2006).

P. brassicae is a soil-borne obligate phytopathogen that causes clubbed roots, which are root tumors commonly named as galls, on cruciferous plant species. Infections by P. brassicae disrupt water and nutrient uptake in affected plants, resulting in wilting, stunting and premature ripening of the above-ground organs. The pathogen can survive as resting spores (RS) in the soil, and initially infects the root hairs, followed by cortical cell infection (Kageyama & Asano, 2009). Infections on the root hairs produce motile zoospores which invade the cortical tissue, where secondary plasmodia are formed, and the formation of galls as a hypertrophic growth of the root tissues is started by triggering unusually high production of plant growth regulators (Hwang et al., 2012). The secondary plasmodia are cleaved into millions of RS and the root galls disintegrate, releasing long-lived RS into the soil (Friberg et al., 2008).

Continuous cruciferous crops in infested soils increase the concentration of RS which is directly related to the disease severity, following a dose–response relationship (Tsushima et al., 2010). Thus, primary measures for disease management should be aimed at reducing the initial population of P. brassicae, while future activities should focus on reducing primary root hair infection and secondary cortical infections (Saharan et al., 2021b). Cultural practices such as removal of diseased plants, bait crops, a long crop rotation, effective drainage, management of susceptible weeds in the absence of the crop, use of soil amendments for increasing soil pH, and sanitization plans for tillage equipment, tools, footwear, plant material, irrigation water, and so on, are very important strategies to reduce P. brassicae inoculum in the soil (Saharan et al., 2021b). Application of fungicides such as fluazinam, cyazofamid, and flusulfamide have been reported as effective against Plasmodiophora (Eom et al., 2011; Irokawa et al., 2020; Wang et al., 2017) through inhibition of spore germination and suppression of both root hairs and cortical infection. Nevertheless, the efficacy of fungicides is limited and, in some cases, could have negative impacts on the environment and human health (Regueiro et al., 2015; Wightwick et al., 2012). Amendments, such as limestone and gypsum to increase soil pH higher than 5.7, favor root development and reduce the crop losses. However, this strategy is often neglected due to lack of knowledge by the farmers or to practical difficulties (dos Santos et al., 2020), which is characteristic in the tropics with acidic soils and frequent application of raw manures (Bhering et al., 2020).

Biological control of clubroot with antagonistic microorganisms is an additional management strategy (Struck et al., 2022), and represents a useful practice in a sustainable agriculture framework (Auer & Ludwig-Müller, 2023). Evaluations of biological control agents against clubroot are relatively recent, and consequently reports under field conditions are limited, and there are no commercial bioproducts recommended against P. brassicae (Saharan et al., 2021b). Nevertheless, bacteria such as Bacillus amyloliquefaciens, B. megaterium, B. subtilis and Pseudomonas fluorescens are well known as plant growth promoters in cruciferous species such as Arabidopsis, cabbage and mustard (Asari et al., 2016; Park et al., 2009; Turan et al., 2014; Zaidi et al., 2006). Bacillus strains have also been tested for their ability to protect plants by stimulating plant growth (Zhu et al., 2020).

The effectiveness of Trichoderma spp. and Streptomyces sp. against clubroot in Chinese cabbage (Brassica chinensis L.) was demonstrated early (Cheah et al., 2000). Strains of Pseudomonas fluorescens and Trichoderma viride have been proven to be efficient against P. brassicae in cabbage with 50% reduction in incidence of clubroot (Loganathan et al., 2001). More recently, it has been reported that treatment of individual strains of B. velezensis and B. amyloliquefaciens can reduce the severity of the disease by 40% in canola (Zhu et al., 2020). Although some biocontrol agents have also shown antagonistic potential against Plasmodiophorida phytopathogens such as Spongospora subterranea (Mesa et al., 2017) and P. brassicae (Zhu et al., 2020), inconsistent as well as negative results of biocontrol also have been published. For instance, Bhattacharya and Pramanik (1998) found increased root hair infection by P. brassicae under in vitro conditions, also the disease index was increased in Chinese cabbage inoculated with B. subtilis T99. Additionally, the review made by Auer and Ludwig-Müller (2023) shows some examples of negative effect of biocontrol against clubroot by strains of Trichoderma and Streptomyces.

In Colombia, only a few works on biological control of clubroot have been reported, despite the disease being present in Colombia since 1972 (Torres, 1972). In contrast to the aforementioned reports, the first published local study reported that neither application of Trichoderma spp. every two weeks, nor a handmade preparation of a mixture of bacteria and yeasts (called as effective microorganisms – EM) reduced the clubroot disease in broccoli under field conditions, over three consecutive crop cycles (Castillo & Guerrero, 2008). A field experimentation carried out by Botero et al. (2015) showed some potential of biocontrol against clubroot in cabbage using the commercial formulation Tricotec® WP (200 g ha−1) based on Trichoderma koningiopsis Th003 (1 × 109 conidia g−1). However, the biological control of the disease was reduced when biological treatment was combined with 2.0 t ha−1 of lime but not under the combination with 1 t ha−1. Later, the same research group reported a differential response to biocontrol, depending on the density of P. brassicae in the soil, the Trichoderma species, the kind of the bioassay (controlled vs. field conditions), and the rate of applications of biocontrol treatments (Botero, 2016).

As living organisms, beneficial microorganisms applied in agriculture are subjected to a wide range of environmental biotic and abiotic factors. Thus, their compatibility with the native microbiota and factors such as light intensity, temperature, pH, physic and chemical soil characteristics, and water content, affect the performance of biocontrol agents (Mitter et al., 2019). These factors cause high variability in biocontrol efficiency, which has been one of the reasons why the implementation of large-scale biological control has been delayed, which, although stated previously (Andrews, 1992; Hjeljord et al., 2000; Guetsky et al., 2001; Shtienberg & Elad, 2002) is still a matter of concern. The use of microbial consortia, including biocontrol agents with different modes of action and different environmental requirements, to overcome this issue, improving the effectiveness and increasing the stability of populations of the introduced agents, was proposed years ago (Elad et al., 1998; Guetsky et al., 2001).

Nowadays, investigations on the use of combinations of antagonists – so-called synthetic microbial consortia or synthetic microbial communities (SynComs) – has been resumed with the aim of assuring consistent control effectiveness against multiple pathogens, and over a wide range of environmental conditions (Johns et al., 2016). According to the review published by Auer and Ludwig-Müller (2023) there are few reported studies evaluating microbial consortia against clubroot disease (e.g. Gao & Xu, 2014; Kurowski et al., 2009; Loit et al., 2020; Yu et al., 2015; Zhang et al., 2022), and considering that using microbial consortia in the field against clubroot is a promising approach to control this disease in cruciferous crops, the aim of this work was to demonstrate whether microbial consortia are more efficient controlling clubroot disease than treatments based on single beneficial native strains of filamentous fungi and rhizobacteria.

Materials and methods

Plant material

Broccoli seeds (Brassica oleracea var. italica) cv. Calabrese susceptible to P. brassicae were sown on peat, in 72-cell seeding trays (one seed per cell). Seedlings were grown for 30 days in a commercial seedling nursery, with a standard program of irrigation and nutrition as followed by the seedling grower. The seedlings were then transplanted to black plastic bags containing 600 g of substrate composed of soil and rice husk (5:1), hereafter referred to as soil (pH: 5.5), with one plant per bag. Plants were placed on elevated metal benches under greenhouse conditions, with average, minimum, and maximum temperatures of 17, 8, and 36 °C, respectively and average, minimum, and maximum relative humidity (RH) of 75, 25, and 96%, respectively, in the Tibaitatá Research Center of Agrosavia (Mosquera, Colombia).

Microorganisms

The strains of biocontrol agents belonging to the Trichoderma spp., Pseudomonas spp. and Bacillus spp. genera were supplied by the Agrosavia’s Microorganisms Germplasm Bank (Section of working collection strains) (Supplementary material Table S1). These strains were included in this work according to the prior knowledge of their potential biocontrol activity or plant growth promotion effects (Cotes et al., 2012; Gámez et al., 2019; Izquierdo-García et al., 2021; Ruíz et al., 2013; Sastoque, 2010). All the strains were stored at -80 °C in a sterile solution of glycerol (20%) and peptone (0.1%). Trichoderma strains were reactivated on Potato-Dextrose-Agar medium (PDA), Bacillus strains on Luria–Bertani-Agar medium (LBA), and Pseudomonas strains on King B-Agar medium (KBA).

P. brassicae was obtained from broccoli clubbed roots in a crop in rural areas of the municipality of Granada, Cundinamarca, Colombia. Mature galls were washed with tap water and a soft detergent solution was used to eliminate the excess of soil, then they were blended for two minutes in saccharose solution (10%) using the proportion 1:3 (galls: saccharose solution vol/vol). The obtained suspension was filtered through muslin cloth (0.5 mm mesh). The collected filtrate was centrifugated (5000 rpm, 10 min), and the pellet containing the spores of P. brassicae was stored at –20 °C in glycerol sterile solution (20%). Preserved spores were used to infest soil in pots (20 ml suspension per pot at 5 × 106 RS ml−1), in which broccoli plants were grown for two months, thus producing fresh spores of P. brassicae, which were used as pathogenic inoculum for protection bioassays.

Effect of microorganisms on plant growth

After microorganism reactivation, the second subculture (24 h and 7 days-old for bacteria and Trichoderma strains, respectively) was used as pre-inoculum of the liquid cultures for bacteria and solid-state fermentation for Trichoderma strains. Pseudomonas strains were grown in KB broth, and the other bacterial strains were grown in LB broth. Briefly, the biomass harvested from solid media was used to prepare concentrated cell suspensions in the respective liquid media. These suspensions were used to inoculate new sterile liquid media (300 ml in 1L-Erlenmeyers), adjusting the initial concentration to 1 × 106 cells ml−1, then incubated for 48 h at 30 °C with continuous agitation at 125 rpm. After incubation, the fermented broth was centrifuged (15,000 rpm for 15 min) and cells and supernatant were harvested separately. The supernatant was centrifugated again and filtered through sterile filters 0.22 µm (Sartorius®); the obtained cells were washed twice with sterile distilled water (sdw), centrifugating and discarding the liquid between washes. Suspensions of cells in sdw, supernatant solutions in sdw, and the combination of cells suspension in supernatant solution were prepared as treatments for the bioassays. Trichoderma strains were grown on autoclaved wet cereal grains (15 PSI for 20 min at 125 °C). Briefly, a conidial suspension from the PDA culture (7 days old) was prepared at 1 × 107 conidia ml−1 in Tween 80 sterile solution (0.1%). 24 g of sterile cereal-based substrate placed in aluminum trays was inoculated with 3 ml of conidial suspension, then incubated for seven days at 25 °C under lighting. Trichoderma conidia from sporulated substrate were harvested by washing with Tween 80 sterile solution (0.1%), vortexed for two minutes, then filtrated through sterile muslin cloth (0.5 mm mesh) to obtain the conidial suspensions.

The treatments were applied to broccoli seedlings (14 days old) in the nursery (4 ml per seedling), added to the soil in plastic bags by drenching 60 ml per bag at seven days before transplant and the day of transplant. Conidial suspensions were applied at 1 × 106 conidia.ml−1, while bacterial suspensions were applied at 5 × 107 cfu.ml−1. To adjust the supernatant solution harvested from bacterial liquid cultures, the respective volume of broth required to adjust the required concentration of cells was used. Treated seedlings (30 days old) were transplanted to the treated soil (one plant per bag), placed on elevated metal benches under greenhouse conditions with the environmental conditions described above, and irrigation and fertilizer were supplied as appropriate. The plants were carefully uprooted 30 days after transplant and the root was washed with tap water. The shoot and the root were separated and dried at 60 °C for three days, then the dry weight was measured in an analytical balance (Kern®).

Two independent experiments were carried out for each group of microorganisms, i.e., bacterial based treatments in one experiment and Trichoderma strains in the other. 60 ml of water were applied per plant in experimental units of an untreated control each time the treatments were applied. The bioassay with bacteria was arranged in a randomized complete blocks design with factorial structure 10 (strain) * 3 (type of inoculum) + 1 (control). There were five replicates of each treatment. On the other hand, a randomized complete blocks design with four replications per treatment was used for the bioassay with Trichoderma strains. The size of each experimental unit was six plants for both experiments. Five and six plants were sampled for the bacterial and Trichoderma bioassays, respectively.

Biocontrol assay with single strains

This assay was carried out in pots under greenhouse conditions. An artificial inoculation of P. brassicae into the soil was made to reproduce the disease, and 30 day-old broccoli seedlings were transplanted. The suspension of RS was adjusted to 5 × 106 spores ml−1 by counting in a Neubauer chamber and 20 ml per bag were applied by drench at 14 days before transplant, i.e., 1.67 × 105 RS g−1 of soil. Two independent experiments were carried out for each group of microorganisms, i.e., one experiment to test the effect of bacteria on clubroot development and another experiment to test the effect of Trichoderma strains. The suspensions of bacteria adjusted to 5 × 107 cfu ml−1 and the suspensions of Trichoderma strains adjusted to 1 × 106 conidia ml−1 were applied as described before, 4 ml per plant at nursery, 60 ml per pot at 7 days before transplant and the day of transplant.

The experimental units (10 pots / one plant per pot placed in a plastic tray) were arranged in a randomized complete blocks design with four replicates. The incidence and severity of the clubroot were recorded at 60 days after transplant (dat). For the severity rating, a scale with five levels of symptoms intensity was used, as described in Cubeta et al. (2004). The disease severity index (DSI) was calculated with the equation DSI = [(ΣSi ∗ Ni) / (4 ∗ N)] ∗ 100, where Si represents the severity grade of the symptoms, Ni is the number of plants in each severity grade, 4 represents the number of severity levels minus 1, and N is the number of plants in the experimental unit (Deora et al., 2012). The efficacy (E) of the treatments for reduction of the DSI was calculated with the equation E = [(A-B)/A]*100, where A represents the value of the disease in the negative control and B represents the disease in the treatment under analysis. With the aim of evaluating the consistency of the results the entire experiment was replicated once.

The four treatments showing the highest efficacy of the disease reduction were selected to form combinations of strains and to evaluate them as consortia in a further experiment, their endophytic ability and survival in the substrate were also determined.

Estimation of populations of selected strains

Cell suspensions of the four strains that showed the best plant growth promotion or disease reduction effects were applied by drench to 15 day-old broccoli seedlings growing on peat in 72-cell seeding trays. The suspensions were adjusted to 5 × 107 cells ml−1 for bacteria and 1 × 106 conidia ml−1 for Trichoderma strains and 4 ml of each treatment were applied per plant. After 15 days of the application of the treatments, nine seedlings were sampled per treatment (three plants per experimental unit), and the stem, root and substrate were separated. The fresh weight of the roots and stem were recorded and then surface sterilization was performed by immersing for 2 min in ethanol (70%), then 10 min in NaOCl (3%), and finally four washes with sterile distilled water. Leaf and root tissue imprints were made on the surface of solid culture media to verify the efficacy of disinfection. Samples of plant tissues were macerated in a mortar with 10 ml of sterile distilled water. 100 µl-samples of this macerate were plated on solid LB, King B, and PDA supplemented with Triton X-100 (0.1% v/v) and chloramphenicol (250 mg L−1) in triplicate. Additionally, 5 g of sampled substrate was suspended in 45 ml of Tween 80 sterile solution (0.1% v/v) and stirred for two min. Serial dilutions were then made, and 100 µl-samples were spread on solid culture media in triplicate. Before serial dilutions in case of monitoring Br042 and Bs006, no diluted suspensions of both macerated tissues and washed substrate were incubated at 90° in thermostatic bath for 15 min. The inoculated Petri dishes were incubated at 30 °C in darkness for four days. Then, the number of colonies of biocontrol agents was counted and the population density (CFU g−1 of tissue and substrate) was calculated. Target colonies were distinguished by their similarity in shape, texture, color and size to the original strains, and compared with colonies obtained in the untreated control. This bioassay was arranged in a randomized complete blocks design with three replicates. The experimental unit consisted of 30 seedlings.

Biocontrol assay with consortia

Combinations of two, three and four selected strains were designed. The concentration of the biocontrol agents in the mixture was 5 × 107 cells ml−1 for bacterial strains and 1 × 106 conidia ml−1 for Trichoderma. The bioassay was performed according to the procedure described above, under a randomized complete blocks design with four replicates, and 10 plants per experimental unit. The entire experiment was replicated once.

To determine whether the performance of the consortia against the disease incidence and severity resulted from an additive, synergistic or antagonistic interaction, the expected efficacy (Eexp) was calculated with the equation described by Abbott (1925) Eexp = (a + b) − ((a × b)/100), where a represents the efficacy under the single given strain application, and b represents the efficacy under the other one strain application that composes the consortium. For more than two strains in the consortium, an additional factor has to be included in the equation to calculate the expected efficacy: Eexp = (a + b + c) − ((a × b × c)/100). The synergy factor (SF) was calculated using the equation SF = Eobs/Eexp, were Eobs corresponds to the efficacy calculated in the consortium under analysis. SF = 1 represents an additive interaction between the strains; SF < 1 means an antagonistic interaction; and SF > 1 indicates a synergistic interaction (Guetsky et al., 2001; Levy et al., 1986).

Statistical analysis

The data from all the experiments were submitted to both normality (Shapiro Wilks, α = 0.05) and homoscedasticity (Bartlett, α = 0.05) tests. To determine whether the treatments had significant effects on the response variables, a two-way ANOVA was performed for the plant growth promotion in the bacteria experiment, while a one-way ANOVA was performed for the other experiments. Atypical data from the biocontrol experiment with single strains were eliminated before the analysis, then a t-test (α = 0.05) for equality of variances for the two replicates of the entire experiment was performed. This analysis allowed us to pool six data from the two repetitions for the statistical analysis. The Tukey’s HSD test (α = 0.05) was performed to compare the means of treatments. The analysis was performed using the statistical software S.A.S Enterprise (8.3 SAS Institute, Cary, NC).

Results

Effect of microorganisms on plant growth

The addition of bacterial cells, supernatant from bacterial cultures, and a combination of bacteria with supernatant to the plants, caused differential effects on both shoot and root dry weight (Table 1). Overall, regardless of the type of inoculum, the application of biological treatments promoted the growth of the plant shoot when compared with the untreated control. In particular, the shoot dry weight in plants treated with supernatant from Br042 or Pf014 liquid cultures, the combination of supernatant with cells of Pf014, and the cells of Br019 and Bs006 was significantly higher than the shoot dry weight of the control plants. The other treatments did not promote the growth of the shoot at 30 days after transplant, but none of them caused negative effects on the shoot growth. On the other hand, it was observed that the root dry weight of plants treated with Bs003-supernatant was significantly higher than the root dry weight in the control, but no other treatment showed promotion of root growth. On the contrary, the roots of plants treated with several of the treatments, mainly under the effect of bacterial cells, showed a root dry weight significantly lower than the root weight of the control plants (Table 1).

Table 1 Effect of bacterial based treatments on shoot and root dry weight of broccoli
Full size table

Except for the application of Tricotec®, treatments based on Trichoderma strains did not promote the growth of the plants either by shoot or root dry weight at 30 days after transplant, compared with the control. However, plants treated with Tricotec®, based on T. koningiopsis Th003, showed a significantly higher root dry weight compared with the control plants and with most of the plants treated with Trichoderma strains. Interestingly, plants treated with the strains T. longibrachiatum Th032 and T. virens Gl006 showed a root dry weight significantly lower than the control plants, but none of the Trichoderma strains caused a reduction in aerial dry weight compared with the control plants (Table 2).

Table 2 Effect of Trichoderma strains on shoot and root dry weight of broccoli
Full size table

Screening of potential antagonists

Application of washed cells of most bacteria strains to the soil showed no reduction in clubroot disease incidence, which increased from 90 to 98% at two months after transplant. However, in the case of clubroot severity, plants treated with the strains of bacteria P. migulae Pf014 and P. fluorescens Ps006 showed a significantly lower disease severity index, compared with the negative control (Fig. 1A). Similarly, two strains of Trichoderma stood out for reducing the disease. Thus, plants treated with T. longibrachiatum Th032 and T. asperellum Th034 showed a significantly lower incidence (data no shown) and disease severity index compared to the negative control (Fig. 1B). In terms of the average efficacy from the two experiments, the bacteria strains Ps006 and Pf014 reduced the clubroot severity index by 24 and 14%, respectively, while Th032 and Th034 reduced the severity index by 28 and 23, respectively.

Fig. 1
figure 1

Effect of beneficial rhizobacteria (A) and Trichoderma strains (B) on broccoli clubroot severity index (DSI) two months after transplant. Pb: negative control. TRCTC: Commercial biofungicide Tricotec® based on T. koningiopsis Th003. Values represent an average from 6 experimental units, 3 from first replicate of the entire experiment and 3 from the second one, i.e., after removing atypical data (n = 6). Bars on the columns represent standard deviation. Treatments with same letter are not significantly different according to the Tukey’s test (α = 0.05)

Full size image

As mentioned before, this experiment was carried out two times, and it was observed that some strains of bacteria and fungi showed consistent effects between the two experiments, while other strains did not. Thus, Ps006 was the bacterial strain with the consistently highest efficacy in reducing the club root severity (23 and 25% in experiments 1 and 2, respectively) (supplementary material—Figure S1), while Th032 (26 and 31% efficacy) and Th034 (28 and 31% efficacy) (supplementary material—Figure S1) were the highlighted strains in the group of fungi. Considering the biocontrol activity of Ps006 and Th034 and the plant growth promotion effect of Bs006 and Br042, these strains were selected for the next phase of consortia evaluation. Th032 was discarded due to its negative effect on root growth (Table 2).

Populations of selected strains

The aim of this experiment was to determine the ability of beneficial microorganisms to colonize the inner tissues of the root and stem of broccoli plants as well as to quantify their survival on the substrate used for plant rooting. Typical colonies of the four strains that showed the best plant growth promotion or disease reduction effects, Lysinibacillus xylanilyticus Br042, Bacillus velezensis Bs006, Pseudomonas fluorescens Ps006, and Trichoderma asperellum Th034 were obtained on solid culture media after plating the dilutions from both macerate of disinfected tissues and washed substrate. Ps006 and Th034 showed the highest population densities in the substrate, with 1.3 × 107 and 7.9 × 104 cfu g−1, respectively (Table 3). The analysis of the root samples showed that all three bacteria can colonize the root of broccoli, and these were found in a range of 9.0 × 101 to 1.4 × 102 cfu g−1. However, not all samples showed the target bacterial colonies, since colonies outgrew from just 22 to 44% of samples. Counts of bacterial colonies in Petri dishes from the macerate-stem samples demonstrated that only Bs006 colonies were recovered, and from just 22% of the samples. No colonies from the other strains were recovered. Colonies of fluorescent bacteria and Trichoderma were also found in both the stem tissue and substrate of the control seedlings. However, the morphology and the low density allowed us to distinguish between the applied microorganisms and the native ones. Bacteria that were tolerant of thermal shocks were not found in the control samples.

Table 3 Population density of selected microorganisms in plant tissue and substrate
Full size table

Effect of microbial consortium on clubroot

The aim of this experiment was to determine whether the mixtures of selected microorganisms improved the biocontrol activity against clubroot as compared with the effect of single strains. Although the entire experiment was performed twice following the same procedure, the level of the disease was very different between the two experiments, being higher in the first one than the second one (Fig. 2 and supplementary material – Figure S2). The level of the disease in the first replicate of the experiment was similar to that found in the screening bioassay described above, so the description of the results of this experiment is based on the observations made in the first replicate. Thus, consistent with observations made previously, the incidence of the disease was not controlled by the treatments based on single strains, nor with most of the tested consortia, except with the combination of Ps006 + Bs006 + Br042 which showed 54 ± 27% efficacy. This consortium also was the only one that showed a significantly lower severity index compared with the negative control and with the other treatments (Fig. 2). In terms of efficacy, this consortium reduced clubroot severity index by 69 ± 19%. Despite the mentioned performance of the disease in the second replicate of the experiment, this treatment kept the effects consistently, and showed 73 ± 35% efficacy in reducing the severity index (Supplementary material—Figure S3).

Fig. 2
figure 2

Effect of beneficial selected microorganisms applied as single strains or as consortia on broccoli club root severity index (DSI) two months after transplant. Pb: negative control, i.e., plants growing in soil inoculated with P. brassicae with no application of treatments. Strains Br042: L. xylanilyticus; Bs006: B. velezensis; Ps006: P. fluorescens; Th034: T. asperellum. Values represent analysis from 4 experimental units from the first replication of the experiment (n = 4)

Full size image

The determination of the type of interaction by the microbial members in the evaluated consortia showed a synergistic interaction in Ps006 + Br042 and Ps006 + Bs006 + Br042 combinations. The interactions in the other consortia were antagonistic, and no additive interactions were found (Table 4). Therefore, the results suggest that these combinations of rhizobacteria are more effective to control the disease than the application of single treatments into the soil. However, a greater observed efficacy under the tripartite consortia (69%) outlines this treatment as more promising against clubroot disease, compared with a consortium based on Ps006 + Br042 (21% efficacy).

Table 4 Effect of microbial consortia on clubroot severity index and type of interaction exerted by the evaluated consortia
Full size table

Discussion

The plant growth promotion trait exerted by the beneficial microorganisms included in this study was considered a selection factor for its positive effects on plants, under the precept that more vigorous plants can tolerate the attack of phytopathogens to a greater extent (Gupta et al., 2017; Tripathi et al., 2022). The strains of beneficial microorganisms included in this study have been the subject of previous works in which plant growth promotion or biocontrol activity have been demonstrated (Mesa et al., 2017; Moreno-Velandia et al., 2019; Smith et al., 2013; Zapata-Narváez et al., 2020), but data for some of these strains had not been published. However, the demonstrated plant growth promoting traits, and the antimicrobial or biocontrol activity by these strains let us consider them as potential candidates to be tested against P. brassicae.

In this study it was observed that after five weeks of interaction with the plants, the treatments based on B. pumilus Br019, B. velezensis Bs006, P. migulae Pf014, T. virens Gl006, and Tricotec® showed significant plant growth promotion effects. However, the durability of the effect from applying the microorganisms just once is not yet clear, and deserves more attention since the response to that question could inform the requirements for establishment of a minimum active population of the beneficial microorganism in the soil/rhizosphere interface. Rather than using reported data, we designed an application program for the biological treatments intending to simulate what happen in real conditions, where the pathogen is already established in the soil. It would be ideal to produce seedlings primed with beneficial microorganisms to prevent P. brassicae in infested soils, agreeing with reports by He et al. (2019) and Lahlali et al. (2013) on the preventive incorporation of the bacterium B. subtilis XF-1 and the endophytic fungus Heteroconium chaetospira to the seedlings, respectively. The application of the beneficial strains into the soil before crop planting would allow early interactions between biological treatments and the pathogen, and additional supplementation of the treatments would be necessary once the seedlings arrive to maintain a minimal effective dose of the beneficial microorganisms.

We tested whether the harvested supernatant from the liquid cultures of bacteria could exert plant growth promoting effects when used as soil treatments. Supernatants from Br042 and Pf014 effectively enhanced the plant growth in this study, suggesting that these bacteria have the ability to produce compounds with plant growth stimulation effects. As recently reviewed by Gómez-Godínez et al., (2023), scientific literature has showed that plant growth promotion by rhizobacteria could happen as a result of direct mechanisms of action, such as phytohormone production, nitrogen fixation, nutrient solubilization, and reduction in ethylene levels in the plant; and by indirect mechanisms when there are antagonistic effects on phytopathogens through the production of cell wall degrading enzymes, antibiotics, siderophores, and molecular signatures for the induction of plant systemic resistance. Several of these compounds can be produced by the bacteria during fermentation at laboratory conditions, which could be applied to the soil as a supplementary treatment for beneficial effects to the plants (Köhl et al., 2019). Nevertheless, the positive as well as the possible negative impact on the crop should be tested in the future for the specific pathosystem of P. brassicae—cruciferous, as it was demonstrated in the Fusarium oxysporum f. sp. lycopersici – tomato – B. velezensis system (Másmela-Mendoza & Moreno-Velandia, 2022).

Although not the case in the present work, beneficial fungi can also display direct and indirect mechanisms of plant growth promotion, as it was reviewed in Köhl et al. (2019) in the case of Trichoderma spp. Potentially, the durability of the beneficial effects was different among the tested bacterial and Trichoderma strains under the experimental conditions of this work. No systematic observations have allowed us to suggest that plant growth promoting effects by beneficial microorganisms are transitory, with the need for additional applications of the biological treatment. However, to our knowledge no studies have been done to elucidate this hypothesis.

The plant growth promoting effect of the bacteria was related with a greater dry weight of shoot at the time of measurement i.e., 30 days after transplant, while in the case of Trichoderma strains only the bioproduct Tricotec® had a positive effect on the roots of plants, but not on the shoot. Interestingly, T. virens Gl006 and T. longibrachiatum Th032 caused a reduction of root dry weight, suggesting a negative effect on broccoli plants. However, this phenomenon deserves additional studies to determine whether this is a consistent or a transitory effect, and moreover whether the crop yield and the plant response to pathogens are affected too. In this regard, T. virens has been described as a good plant growth promoter through the production of indole-3-acetic acid (IAA) and IAA-related substances (Contreras-Cornejo et al., 2009). However, at the same time these authors reported a dose-dependent effect on biomass production, where high biomass was obtained under nM amounts of IAA, but plant growth repressing was observed at mM range.

In the present study, applications of single strains of beneficial rhizobacteria and Trichoderma into the soil artificially inoculated with P. brassicae did not reduce the incidence of clubroot disease in broccoli plants, which was very high. In fact, 18 to 24% of experimental units showed a disease incidence value between 75 to 89%, and the other experimental units showed incidence between 90 to 100%. P. brassica is one of the most difficult phytopathogens to control, and traits such as a high number of resting spores released into the soil and the high content of chitin (about 25% in the cell wall, Saharan et al., 2021c) can explain the low efficacy of chemical and biological measures to control it. Based on the information in the review of Saharan et al. (2021b) on clubroot disease management by biological control agents, most cases of satisfactory control by microbial agents show a reduction in disease severity but not in disease incidence.

Moreover, agreeing with the results in the present study, the review of Saharan et al (2021b) shows more events of biocontrol of clubroot by antagonistic bacteria than strains of Trichoderma spp. However, the influence of biotic and abiotic factors on the complexity of microbial interactions and variation in the outcome of biocontrol must be considered, as these authors stated. For instance, greater biocontrol has been achieved under low or moderate pressure of P. brassicae inoculum in the soil, like ≤ 1.5 × 105 RS g−1 compared with higher concentrations (Narisawa et al., 2005; Peng et al., 2011). Agreeing with the reports mentioned above, it is noteworthy the improving of the efficacy showed by the consortium of Ps006 + Bs006 + Br042 compared with the application of the individual strains, considering the high concentration of P. brassicae used in the present work (1.67 × 105 RS g−1).

A biocontrol efficacy of 24%, as demonstrated by solo application of P. fluorescens Ps006, is an important value considering how difficult it is to control P. brassicae. However, more interesting is the increase in efficacy by combining Ps006 with B. velezensis and L. xylanilyticus, which are also recognized plant beneficial bacteria (Rabbee et al., 2019; Tan et al., 2015), even though these were not effective individually. In this context, it has been stated that the use of microbial consortia to control plant disease instead of single strains of biocontrol agents could be consider a milestone in the new era of biological control of plant diseases, since it has been a strategy designed to increase the efficacy and improve the consistency of both the biocontrol and plant growth promotion effects (Santoyo et al., 2021).

In the group of fungi used in the present work, several species of Trichoderma are among those recognized as efficient biological control agents in a broad number of pathosystems (Alfiky & Weisskopf, 2021; Vinale et al., 2008). Although the six species of Trichoderma tested in this study showed some efficacy at reducing the clubroot severity in the first replication of the entire experiment, just two strains (T. longibrachiatum Th032 and T. asperellum Th034) showed consistent effects against clubroot disease in the two replications of the single treatments experiment. However, Th034 did not keep the high biocontrol effect in the experiment of consortia evaluation. Moreover, the consortia in which Th034 was included showed low efficacy against clubroot, and antagonistic interactions.

Struck et al. (2022) suggested that those biological control agents that efficiently reduce both the germination of resting spores of P. brassicae in the soil and the germination of the secondary spores in the cortex should be considered as good candidates to control clubroot, since they could reduce hair root infection and the development of the pathogen inside the host tissue. However, our rationale for this study was that the combination of Trichoderma strains with strains from the Bacillus subtilis species complex would be an efficient treatment to control soilborne phytopathogens. Thus, we considered that parasitism, as the most predominant biocontrol traits among Trichoderma species (Steyaert et al., 2003), in combination with the antibiosis, as the most prominent biocontrol trait among the antagonistic rhizobacteria (Gómez-Godínez et al., 2023) would be effective against P. brassicae. However, when applied together to the soil, interactions between them could affect their performance when compared with the efficacy of solo applications, as the results of this study suggested for this specific pathosystem, with the combinations in which T. asperellum Th034 participated. Nevertheless, there are cases of compatible interactions between Trichoderma and beneficial bacteria with positive effects of biocontrol (Izquierdo-García et al., 2020).

Overall, the findings from this study demonstrated great potential of biocontrol by some strains of bacteria as well as fungi as single treatments against clubroot disease, considering they were confronted with a high concentration of P. brassicae in the soil. Moreover, the performance showed by the consortium of the rhizobacteria P. fluorescens, B. velezensis and L. xylanilyticus against the disease in terms of both efficacy level and consistency under the aforementioned conditions of inoculum pressure, suggests it as a good biocontrol candidate against P. brassicae. Additional studies to characterize the modes of action of the consortium, as well as validations of the biocontrol effect in field should be carried out.