In the experiment, we investigated three crops with different salt tolerance properties: sunflower (H. annuus), lettuce (L. sativa), and barley (H. vulgare). Commercial seeds of sunflower (H. annuus) and butterhead lettuce (L. sativa) were purchased from Torseed (Toruń, Poland) and Plantico Company (Stare Babice, Poland), while winter barley (H. vulgare) seeds were obtained from the Plant Breeding and Acclimatization Institute (IHAR) (National Research Institute, Strzelce, Poland). Healthy and uniformly sized seeds were surface sterilized using a 30% hydrogen peroxide and 70% ethanol mixture (1:1, final concentration 15% hydrogen peroxide and 35% ethanol) for 11 (sunflower), 7 (lettuce), and 3.5 min (barley). Then, seeds were rinsed six times with sterile distilled water. The residual water after the last washing was used for the sterility test for bacteria and fungi, by plating on R2A (Difco™ and BBL™) and PDA (Difco™ and BBL™) plates, respectively. Only successfully sterilized seeds were used in the pot experiment. For each investigated plant species, three inoculation variations were used: control (non-inoculated, NI) and inoculated with one of two endophytic bacterial (EB) strains (B1—P. stutzeri ISE12 and B2—K. marisflavi CSE9). Plants were cultivated in one of 4 salinity treatments: 0, 50, 150, or 300 mM. Substrate (sterile mixture of sand and vermiculite, 1: 1) in plastic tray (each pot: 30 × 46 cm, 0.077 L) was salted by adding to the litter box 2 l of sterile Hoagland’s medium enriched with the appropriate concentration of NaCl (Figure 1). Seeds were sown into growth substrate and cultivated under controlled conditions (temperature: 24 ± 2°C and light/dark: 16/8 h) and were irrigated with Hoagland medium (once per week) and distilled water (according to plant requirements) (Hoagland and Arnon, 1950). After 6 weeks, plants were collected and analyzed (Figure 1).

Two halotolerant endophytic bacterial strains, P. stutzeri ISE12 (B1) and K. marisflavi CSE9 (B2), were used for the inoculation of plants in the pot experiment. Both bacterial strains were originally isolated from the roots of S. europaea and characterized for plant growth promoting (PGP) properties (Szymańska et al., 2016, 2020). P. stutzeri ISE12 (NCBI Acc. No. KX686983) was characterized as having a wider range of plant growth-promoting potential (e.g., nitrogen fixation, siderophores, IAA synthesis) and lower tolerance to salinity than K. marisflavi CSE9 (NCBI Acc. No. KX027360) (Szymańska et al., 2016, 2020).

Bacterial inocula were prepared using cells suspended in 2% NaCl solution, which was then diluted to OD = 0.5 (OD—optical density, measured at 600 nm, equivalent to 1.5 × 10⁸ cells/ml) after 72 h of culture incubation on R2A medium (Difco™) supplemented with 2% NaCl at 24°C. Plants cultivated in the pot experiment were inoculated with bacterial strains twice: 1st inoculation—surface sterilized seeds were incubated in bacterial inocula for 45 min at 24°C with shaking (control variant—non-inoculated seeds were incubated in sterile 0.9% NaCl solution); 2nd inoculation—1 ml of bacterial inocula was added to the growth substrate close to the root zone of 2-week-old seedlings (control variant—2% NaCl solution was used).

The number of germinated seeds was determined daily (at 10 a.m.) for 10 days.

The germination percentage (GP) was calculated according to the International Seed Testing Association (ISTA) method:

GP = number of normally germinated seeds/total number of seeds sown × 100

The germination index (GI) was calculated using the following formula:

where Gt is the number of seeds germinated on day t, and Tt is the number of days (Ruan et al., 2002).

Chlorophyll levels in leaves were detected by a hand-held Chl meter CCM-200 (Opti-Sciences, Tyngsboro, Massachusetts, United States) to assess the chlorophyll content index (CCI) after 6 weeks of plant cultivation (Richardson et al., 2002). Then, the plants were carefully removed from the pots, and the roots were washed using tap water to eliminate the residue from the growth substrate. The number of leaves and root and shoot lengths of 42-day-old plants were measured. Leaf area was scanned and measured by digiShape software.

The tissue water content (TWC) was calculated according to the formula of Black and Pritchard (2002):

Biomass accumulation was calculated as fresh (W_f) and dry weight (W_d). The dry weight (W_d) was assessed after 72 h of drying at 85°C. Moreover, the following growth indices were calculated: specific leaf area (SLA), leaf weight ratio (LWR), and root weight ratio (RWR) (Piernik et al., 2017).

Two-way ANOVA was used to check the effect of EB inoculation and NaCl concentration on growth parameters of tested plants. One-way ANOVA was used with Tukey’s post hoc test to compare differences between treatments. Salt resistance assessments were based on comparisons with NI conditions by one-way ANOVA with Tukey’s post hoc test. General assessment of the effect of inoculation (P. stutzeri ISE12—B1 and K. marisflavi CSE9—B2) on the growth of H. vulgare, L. sativa, and H. annuus was performed using multivariate statistical analysis. All above tests were done using the Statistica 10.0 software package (StatSoft, 2006). Discriminant analysis with a forward selection procedure was applied to test which measured features were the most affected in each solution and which discriminated among the NI, B1, and B2 treatments the best (CVA, Canoco 5 package; ter Braak and Šmilauer, 2012).

The analysis of the growth parameters of plants grown in different concentrations of NaCl allowed for the differentiation of the tested crops in terms of tolerance to salinity and indicated their significant species diversity in this respect.

Substrates with increasing salinity (0, 50, 150, 300 mM NaCl) were used for crop cultivation to determine the most significantly affected plant traits (Table 1 and Figures 2–4). H. vulgare plants grown under salt condition were smaller than those grown in the control variant (0 mM NaCl): W_f (was lower 25.2% in 50 mM compared to the control to 81.7% in 150 and 300 mM), W_d (from 15.5 to 73.2%, respectively), NoL (16.7–39.2%), and RL (25.7% to above 42.4%). The TWC was 1.7–3.4% lower under each salt treatment than under the control conditions (Supplementary Tables 1, 2). The germination parameters GP and GI decreased only at the highest salinity (300 mM) (Figure 2).

In general, the plant genotypes from the family Asteraceae (L. sativa and H. annuus) were similarly affected by salt stress. However, L. sativa was more sensitive (non-inoculated plants did not grow in 300 mM NaCl) (Figure 3). L. sativa cultivated in substrate supplemented with 150 and 300 mM NaCl had a lower germination ability, as expressed by GP and GI (Figure 3). The growth parameters were also reduced under salt treatment: W_f (28.7% lower than the control at 50 mM, 72.2% at 150 mM), W_d (25.0 and 68.7%, respectively), and RL (15.1 and 38.8%) (Figure 3 and Supplementary Table 3).

Slightly different strategies to cope with salt stress were demonstrated by H. annuus. Similarly, GP and GI decreased under high salinity (150 and 300 mM) (Figure 4), along with W_f (24.2% lower compared to the control at 150 mM and 58.4% lower at 300 mM), W_d (56.8% at 300 mM), and NoL (22.1% at 300 mM) (Supplementary Table 4). However, RL and RWR first increased significantly at 50 and 150 mM as an adaptation to water deficit and then decreased as a result of salt stress at 300 mM, but only by 34.1% (RL) and 12.6% (RWR). Thus, adaptation to cope with the highest salinity (300 mM NaCl) increased the LWR and CCI (Supplementary Table 4).

In summary, the effect of salinity on plant growth revealed the following order of salt tolerance levels in the investigated crops: H. annuus > H. vulgare > L. sativa.

The differences between the two tested strains (P. stutzeri ISE12—B1 and K. marisflavi CSE9—B2) were determined on the basis of their effect on the increase of plant growth parameters and more universal to all investigated crops strain was selected. The results of two-way ANOVA demonstrated that both investigated bacterial strains significantly affected most of the investigated plant traits (Table 1). In the case of H. vulgare, B1 stimulated NoL development, W_f and W_d, whereas B2 was responsible for better root development (RL and RWR), and water accumulation (TWC). A detailed comparison of salt treatments revealed that bioaugmentation of H. vulgare with both investigated strains had a positive effect on plants cultivated in 50 and 150 mM NaCl, as seen for the following parameters: GP, GI, NoL, RL, W_f, W_d, LWR, and TWC (Figure 2 and Supplementary Table 2). Under a lower salinity of 50 mM, inoculation of H. vulgare with B1 resulted in higher NoL and biomass accumulation (W_f and W_d) compared to that with B2 (Figure 2 and Supplementary Table 2). At 150 mM, both strains stimulated NoL, Wf, Wd, and TWC, but strain B2 affected root growth more strongly (higher RL).

Both species from the Asteraceae family (L. sativa and H. annuus) were more susceptible to B2 inoculation (Table 1). In general, inoculation of L. sativa with B2 increased most of the growth parameters. Detailed analysis of the salt treatments demonstrated that under the highest salinity (300 mM), strain B2 stimulated RL, W_f, W_d, and CCI, whereas under lower salinity (50 and 150 mM), NoL and RWR were stimulated (Figure 3 and Supplementary Table 3). The role of B1 in growth stimulation was much smaller than that of B2. Moreover, some traits were negatively affected by this strain, e.g., RL and CCI (Figure 3 and Supplementary Table 3).

The growth of H. annuus was also better stimulated by B2 (Table 1), resulting in stronger root development (RL and RWR), photosynthesis ability (CCI and LWR), and fresh biomass accumulation (W_f), whereas strain B1 positively affected mainly water accumulation (TWC) and photosynthesis ability (LWR). In general, B2-inoculated plants were characterized by better seed germination (GP and GI) (Figure 4). In all salinity treatments, these plants had better developed roots (RL and RWR), higher biomass (W_f and W_d), and a 50 mM higher chlorophyll concentration in the leaves (CCI) (Figure 4 and Supplementary Table 4). Conversely, decreases in growth parameters, e.g., Wd, RWR, and CCI, were observed after inoculation with B1. The results clearly show a lack of compatibility between H. annuus and P. stutzeri ISE12; moreover, an adverse influence of that strain was observed.

The PCA results presented in Figure 5 show the percentage differences between plant traits for inoculated and control (NI) treatments, which included three plant species (H. vulgare, L. sativa, H. annuus) and four different levels of salinity (0, 50, 150, 300 mM NaCl). Positive correlations were found between L. sativa (all investigated variants: Ls_B2_0, Ls_B2_50, Ls_B2_150) and H. annuus (three of the four variants: Ha_B2_0, Ha_B2_150, Ha_B2_300) inoculated with B2 under most salinity treatments among the NoL, W_f, W_d, RL, and RWR traits. Bioaugmentation of plants from this family with B1 was associated only with SLA, TWC, and LWR. Inoculation of all investigated genotypes with B2 stimulated growth parameters according to the following trend: L. sativa > H. annuus > H. vulgare (Figure 5). Strain B1 revealed the opposite trend and had no evident effect.

Based on the obtained results, an analysis was carried out, which allowed for the selection of plant growth parameters most sensitive to increasing salinity and bacterial inoculation. The results of the discriminant analysis and forward selection procedure revealed that salinity modified the effect of EB in a quantitative way. In the case of H. vulgare, the most critical feature for discrimination between the NI, B1, and B2 groups in all salinity treatments was the water content in plants (TWC), which explained 28–14% of the total variation (Table 2). Moreover, under 0 mM NaCl, the most affected plant parameters were LWR (explaining 48.2% of the trait variance between the NI, B1, and B2 groups) and CCI (7.5%) (Supplementary Figure 1). At 50 mM, in addition to TWC, significant group separation was found for NoL (31.4%), and in 150 mM, the same was found for RL (40.5%).

The L. sativa analysis demonstrated that the most essential features for discrimination between groups were at 0 mM NaCl root development, indicated as RWR (explaining 38% of the variation), chlorophyll index (CCI, 16%), and number of leaves (NoL, 9%) (Table 2). At 50 mM NaCl, the most affected traits were W_d (explaining 48.2% of the variance between the NI, B1, and B2 groups) and SLA, whereas under 150 mM, W_d (27%) and RL (16%) were the most affected, and under 300 mM onlyRL was the most affected (63%, NI plants died) (Supplementary Figure 2).

The most essential trait for discrimination between the NI, B1, and B2 treatments for H. annuus was the chlorophyll index (CCI, explaining 9–35% of trait variance) across all NaCl concentrations, except for 150 mM (Table 2). Moreover, under 0 mM NaCl, the critical features for discrimination between the NI, B1, and B2 treatments were LWR (explaining 24% of the variation) and W_f, while under 50 mM, root development, expressed as RWR (24%), under 150 mM, RWR (31%) and TWC (18%) and under 300 mM NaCl, fresh weight (W_f, 8%) and NoL (14%) (Supplementary Figure 3).

This analysis allowed us to designate chlorophyll concentration (CCI), the development of leaves (NoL, LWR, SLA), water storage (TWC), root development (RL, RWR), and biomass accumulation (W_f, W_d) as crucial traits affected by salinity and bacterial inoculation. However, the effects reported here are species specific.

Determination of plant germination capacity under high salinity conditions is a crucial parameter for assessing plant tolerance to salt stress (Li et al., 2020). Salinity adversely affects the germination stage by limiting water availability and changing stored reserve mobilization and structural protein organization (Demir and Mavi, 2008; Nawaz et al., 2010). The negative effect on germination can be genotype- or cultivar-specific (Al-Ashkar et al., 2020; Li et al., 2020; Kadioǧlu, 2021). In the present study, we observed significant differences in salinity between the three investigated crop genotypes. Li et al. (2020) studied the different seed germination stage parameters (such as germination rate, germination index, germination energy, germination vigor index, and water content) of 552 sunflower germplasms with different levels of salt stress tolerance. The authors found that the germination index and the germination vigor index were the most reliable traits and were strongly associated with salt tolerance. In our work, increasing the concentration of NaCl in substrates inhibited seed germination; however, the effect was species-specific. The effect of salinity on the germination of H. annuus and L. sativa was particularly evident at higher salinities (150 and 300 mM NaCl). However, across all tested plant species, we did not observe decreased germination at 50 mM NaCl. This is because a low concentration of Na⁺ has a positive effect on plant growth, since Na⁺ participates in osmotic and nutritional mechanisms. This effect was also observed in our studies on red beet and canola (Szymańska et al., 2019, 2020), as well as in experiments with red beet, wheat and Zygophyllum xanthoxylum described by Wu et al. (2013). H. annuus and H. vulgare are crops with the highest levels of salt stress tolerance (Garthwaite et al., 2005; Li et al., 2020; Vasilakoglou et al., 2021), which we confirmed in the experiments presented in this work. In contrast, we observed relatively low tolerance of L. sativa to salinity, especially at the highest concentration of NaCl (300 mM), where we observed complete inhibition of germination and plant growth. Moreover, gradual plant death was observed over several weeks during the experiment. Ludwiczak et al. (2021) showed that osmotic stress rather than ionic effects is responsible for this process.

The compatibility between selected strains was plant specific. Both endophytes (P. stutzeri ISE12 and K. marisflavi CSE9) increased H. vulgare adaptation to salt stress, which was confirmed by the growth parameters, especially under supplementation with 50 and 150 mM NaCl. Similarly, a positive effect of the same strains was observed in our earlier studies with B. napus and B. vulgaris (Szymańska et al., 2019, 2020).

Our results suggest that crops from the same family growing under similar conditions (e.g., substrate, temperature, light) can be compatible with the same microorganisms. Our results show that H. annuus and L. sativa (family Asteraceae) inoculated with K. marisflavi CSE9 showed increases in the same growth parameters (germination percentage and index, number of leaves, fresh weight, root length, root weight ratio, chlorophyll content index). However, the positive effect of inoculation with a specific strain may also be related to the specific metabolic activities of bacteria used for inoculation. P. stutzeri ISE12 possesses a wider range of metabolic properties (siderophores and IAA synthesis, cellulolytic activity, and presence of nifH) than K. marisflavi CSE9 (siderophores and IAA synthesis) (Szymańska et al., 2020). Moreover, although both strains showed high salinity tolerance, we confirmed that K. marisflavi CSE9 had higher potential (Szymańska et al., 2020). Both strains had IAA synthesis ability, and multifunctional phytohormones play crucial roles in plant growth and development promotion as well as stress resistance (Numan et al., 2018; Khan et al., 2019b; Saleem et al., 2021). Egamberdieva (2009) observed higher crop productivity after the inoculation of wheat with IAA-producing strains under salinity stress than without salt (2008). Similarly, the mitigation of salt stress and stimulation of plant growth was noted for cotton, and the authors suggested that IAA synthesized by bacteria can be an extremely important factor in the salt stress tolerance of plants (Egamberdieva et al., 2015).

Based on the results obtained in the study, we also determined universal parameters that are the most affected by inoculation under salt stress. However, we recognized that the inoculation effect was dependent on the salinity level and was species specific. In the case of H. vulgare, we identified TWC as the most affected trait in all salt treatments. In general, bacteria increased the water content in plant tissues at 50 and 150 mM NaCl, but at 300 mM, the water content significantly decreased. This trait was also independently affected by salinity level. TWC decreased together with increased salinity. This result is in line with those by Ludwiczak et al. (2021), who identified the TWC of H. vulgare and A. sativa as one of the most affected traits under 150 mM NaCl. However, Garthwaite et al. (2005) investigated the salt tolerance of eight wild Hordeum species (including H. bogdanii, H. intercedens, H. jubatum, H. lechleri, H. marinum, H. murinum, H. patagonicum, and H. secalinum) and cultivated barley (H. vulgare) and observed the smallest reduction in this parameter for H. vulgare compared to the wild Hordeum species. On the other hand, stimulation of tissue water storage after inoculation by endophytes was also reported for draught stressed maize by Naveed et al. (2014) and lettuce by Molina-Montenegro et al. (2016). Additional crucial traits of H. vulgare were the number of leaves (NoL) under 50 mM NaCl and RL under 150 mM NaCl, which were the most stimulated by B2 bacteria.

The RL trait was also crucial for discriminating between L. sativa treatments and was stimulated by B2 bacteria, especially at 150 and 300 mM NaCl. The reduction in root length under salt stress may result from the direct contact of the roots with an unfavorable abiotic factor (Jamil et al., 2006; Vibhuti et al., 2015; Peng-cheng et al., 2021). The stimulation of root growth may be a solution for the improvement of water absorption under salt stress (Piernik et al., 2017).

Interestingly, in the case of lettuce, TWC was not a significant factor for discriminating the investigated treatments. Al-Maskri et al. (2010) noted a negative effect of salinity on several L. sativa growth parameters (including the number of leaves, plant fresh weight, shoot fresh and dry weight, shoot dry matter percentage, root fresh and dry weight, root dry weight percentage, leaf area and leaf area index), and similar to our work, they did not observe adverse effects on the root and shoot water content percentages. The opposite results were obtained by Ahmed et al. (2019), who observed a gradual reduction in the water content in L. sativa roots and shoots with increasing NaCl concentration (2019). Many differences in the growth parameters of L. sativa grown under salt stress have been observed by various researchers (e.g., Bar-Yosef et al., 2005; Ünlükara et al., 2008; Kurunc, 2021). Presumably, the varieties of L. sativa are characterized by large differences in salinity tolerance, and plant responses may be related to other morphological and physiological parameters.

Not only the root length but also the root weight ratio (RWR) was crucial for the discrimination between NI, B1, and B2 treatments and was positively stimulated by B2 bacteria, especially in the case of L. sativa and H. annuus. This result suggests changes in the plant growth strategy and a higher investment in root development with endophyte inoculation (Reynolds and D’Antonio, 1996).

For L. sativa and H. annuus, biomass accumulation was of crucial importance in the context of salinity and endophytic bacteria interactions. The dry weight (W_d) showed that the best salt levels were 50 and 150 mM NaCl for lettuce. This trait was significantly stimulated by B2 bacteria. The fresh weight (W_f) showed that the best salt levels were 0 and 150 mM for sunflower and was also stimulated by B2 (K. marisflavi).

The most interesting result is the impact of the investigated endophytes on the chlorophyll content, expressed here as CCI. Inoculation with B1 (P. stutzeri) increased the CCI of H. vulgare, whereas inoculation with B2 increased the CCI of L. sativa under the 0 mM control conditions. Moreover, the CCI parameter was crucial for discriminating across the investigated treatments of H. annuus at 0, 50, and 300 mM, where B2 bacteria were responsible for higher values. An increase in chlorophyll content after inoculation with endophytic bacteria has also been observed by other researchers (e.g., Zhao et al., 2015; Yuan et al., 2018; Tharek et al., 2021). In our previous research, we found the same effect of P. stutzeri on beetroot (Piernik et al., 2017; Szymańska et al., 2020). A higher CCI may result in a higher efficiency of photosynthesis under environmental stress (Zhao et al., 2020).