Selection of Promising Rhizobia for the Inoculation of Canavalia ensiformis (L.) DC. (Fabaceae) in Chromic Eutric Cambisol Soils

Abstract

Canavalia ensiformis (L.) DC. (Fabaceae) is used in Cuba in soils dedicated to coffee cultivation, contributing to soil nutrition and crop productivity. However, no rhizobial isolates are currently available for inoculating this legume in Chromic Eutric Cambisol soils. The aim of this study was to select rhizobial strains that promote the growth of C. ensiformis in Chromic Eutric Cambisol soils. Nodules were collected from C. ensiformis plants, surface-sterilized, and macerated to isolate potential rhizobia. The isolates were characterized based on cultural, morphological, and biochemical traits, and their symbiotic effectiveness was evaluated through in vitro inoculation assays in Macroptilium atropurpureum (siratro) plants. Inoculation trials were conducted under semi-controlled conditions and in the field between coffee rows. The number and dry weight of effective nodules, number of trifoliate leaves, and shoot dry biomass were measured. Nine bacterial isolates were obtained, grouped into four morphotypes, and assigned as possible members of the families Phyllobacteriaceae, Methylobacteriaceae, or Nitrobacteraceae. Under semi-controlled conditions, inoculation with three isolates increased the number of nodules (by 56–80%), the number of trifoliate leaves (by 20–45%), and shoot biomass (by 10–40%) compared to the non-inoculated treatment. Additionally, one of the isolates increased nodule dry weight by 27%. In the field between coffee row, increases were also observed in the number of trifoliate leaves (by 18–26%) and shoot biomass (by 15–24%). This study supports the selection of efficient rhizobia adapted to the edaphoclimatic conditions of Cuban coffee agroecosystems.

1. Introduction

Green manures are plants grown and incorporated into the soil in situ with the aim of preserving, improving, and restoring its physical, chemical, and biological properties. These crops provide organic matter and stimulate microbial activity, contributing to nutrient cycling in the soil. In addition, they reduce erosion and nutrient leaching, complement the nutrition of the main crop, and help control pests and diseases [1,2]. Leguminous plants have been widely used as green manures mainly due to their ability to supply nitrogen to the soil through the symbiotic association they establish with bacteria known as rhizobia [3]. This association results in the formation of a specialized organ in the roots, the nodule, where rhizobia reside and perform biological nitrogen fixation (BNF). Through this process, molecular nitrogen (N₂) from the atmosphere, which is not directly assimilable by plants, is reduced to ammonium (NH₄⁺) for protein synthesis within the plant [4]. The presence of rhizobia within a functional nodule indicates that the bacteria have successfully passed a multi-stage selection process involving host recognition, infection and entry into plant tissues, survival within the nodule environment, and a measurable physiological contribution to the host. Consequently, isolating rhizobia directly from nodules increases the likelihood of obtaining competitive and effective symbionts that are specifically adapted to the target legume.

Canavalia ensiformis (L.) DC. (Fabaceae) is a legume used as green manure in various cropping systems because it accumulates abundant biomass with high protein and organic nitrogen content that is incorporated into the soil (by cutting or in situ), decomposed by microbial activity, and absorbed by the main crop [5,6,7]. In Cuba, for example, C. ensiformis contributes more than 150 kg ha⁻¹ of nitrogen and up to 5 t ha⁻¹ of dry matter to the soil, resulting in yield increases in several economically important crops such as maize (Zea mays L.) and coffee (Coffea sp.) [8,9]. Coffee, in particular, requires between 100 and 150 kg N ha⁻¹ year⁻¹ to achieve high yields and optimal bean quality [10]. It is a traditional crop in Cuba, a highly demanded export commodity, and the economic backbone of the country’s mountainous regions, where 90% of the plantations are located (70,000 ha) [11,12]. Coffea arabica L. is the most widely cultivated species, occupying 65% of the area devoted to this crop. The cultivar C. arabica L. Isla 6–11 stands out for its high productivity and resistance to coffee rust (Hemileia vastatrix) [13].

In major coffee-producing countries of the region, such as Brazil and Colombia, chemical fertilizers, organic amendments, inoculants based on nitrogen-fixing microorganisms, as well as C. ensiformis cultivation, are commonly used as strategies to incorporate nitrogen into coffee production systems [14,15,16]. In Cuba, considering the scarcity of synthetic fertilizers and their negative environmental impact, C. ensiformis is increasingly used as a green manure complemented with microbial inoculants. Recent studies in production systems in Cuba report that C. ensiformis inoculated with rhizobia and arbuscular mycorrhizal fungi (AMF) contributes approximately 200 kg N ha⁻¹ and increases coffee bean yield by about 15% [8].

According to the World Reference Base for Soil Resources, the Guamuhaya massif in Cuba is one of the country’s most important coffee-producing areas, characterized by the prevalence of Reddish Brown Fersialitic soils, corresponding to Chromic Eutric Cambisols [17]. These soils occupy nearly 40% (60,000 ha) of the area dedicated to coffee cultivation and have served as a key site for research on rhizobial inoculation in C. ensiformis intercropped with coffee plants [18]. However, most of these studies have relied on rhizobial isolates obtained from other soil types, including Ferrallitic and Vertisols, Acidic Ferruginous Nodular Gley (pH 5.1), and Calcareous Sialitic Brown soils, located in the western, central, and eastern regions of the country, respectively [6,19,20].

Soil pH, organic matter content, and mineral composition are key abiotic factors influencing the rhizobia-Canavalia symbiosis [21,22,23]. In addition, the low rhizobia-legume compatibility and limited persistence of bacterial inoculants in soil highlight the need to identify rhizobial strains with high symbiotic efficiency and strong adaptation to the edaphoclimatic conditions of the target environments [24,25]. Based on these considerations, the aim of this study was to select rhizobial strains that promote the growth of C. ensiformis in Chromic Eutric Cambisol soils.

2. Materials and Methods

2.1. Sampling

The study area is located in eastern Cuba, within the Guamuhaya Mountain Massif, at the Agro-Forestry Experimental Station of Jibacoa, Villa Clara, Cuba (22°00′00″ N, 79°50′00″ W; 340 m a.s.l.), an important center for coffee cultivation in the region. The climate is humid montane, with two well-defined seasons: a rainy season (May–November) and a dry season (December–April). Annual precipitation ranges from 1600 to 2000 mm, with a mean annual temperature of approximately 23 °C, maximum temperatures of 31–33 °C, minimum temperatures of 22–24 °C, and relative humidity generally exceeding 80%. The vegetation is mainly composed of agroforestry systems and associated crops, predominantly coffee plantations and pastures interspersed with tree species and Poaceae ground cover [26].

The sampling site consisted of a monoculture of Canavalia ensiformis (L.) DC. (Fabaceae). Five collection sites were established, spatially distributed across the sampling area. Each site included a healthy C. ensiformis plant at the flowering stage, 90 days after sowing. Plants were carefully uprooted to preserve the root system, and all nodules were collected from the upper portion of the main root. The nodules were stored at 4 °C in sealed tubes containing calcium chloride to prevent hydration until laboratory processing [27].

2.2. Soil Chemical Analysis

The collected soil corresponded to a Chromic Eutric Cambisol from the Agro-Forest Experimental Station of Jibacoa, Villa Clara, Cuba, belonging to the Reddish Brown Fersialitic soil family. These soils are clayey, friable, and well-drained. Soil samples (500 g) were collected from the surface layer (0–20 cm), air-dried, sieved through a 2 mm mesh, and characterized according to the methods described by González-Fernández et al. [28]. Briefly, organic matter content was determined using the Walkley-Black method, and soil pH was measured potentiometrically in soil suspensions with KCl (1:2.5 w:v). Available P₂O₅ and K₂O (mg 100 g⁻¹ soil) were extracted using 0.1 N sulfuric acid at a soil-to-solution ratio of 1:2.5 for 3 min following the Oniani method; phosphorus was determined colorimetrically and potassium by flame photometry.

2.3. Isolation and Characterization of Rhizobial Strain

The isolation and preliminary characterization of putative rhizobia from C. ensiformis nodules were conducted at the Soil Microbiology Laboratory, Department of Physiology and Biochemistry, National Institute of Agricultural Sciences (Cuba). Nodules were first hydrated in tap water for 3 h, then surface-disinfected with 95% ethanol for 1 min, followed by immersion in 0.1% sodium hypochlorite for 4 min, and finally rinsed ten times with sterile distilled water. The efficiency of the surface disinfection process was verified by culturing 100 µL of water from the last rinse on the rich media Tryptic Soy (TS) medium (per liter: tryptone 15 g, soy peptone 5 g, NaCl 5 g; pH 7.3 ± 0.2) and R2A medium (per liter: yeast extract 0.5 g, proteose peptone 0.5 g, casamino acids 0.5 g, glucose 0.5 g, soluble starch 0.5 g, K₂HPO₄ 0.3 g, MgSO₄·7H₂O 0.024 g, sodium pyruvate 0.3 g; pH 7.2 ± 0.2). The disinfected nodules were macerated in 0.5 mL of sterile saline solution (8.5 g L⁻¹ NaCl). Serial tenfold dilutions were prepared, and 0.1 mL aliquots were plated on yeast mannitol agar (YM) medium (per liter: mannitol 10 g, yeast extract 0.5 g, K₂HPO₄ 0.5 g, MgSO₄·7H₂O 0.2 g, NaCl 0.1 g, and CaCO₃ 0.15 g; pH 6.8) with Congo Red (25 mg L⁻¹). The dye increases the contrast of typical rhizobial colonies, which are usually white or cream-colored and mucilaginous, against the medium, thereby facilitating the observation of their typical morphologies. Rhizobial isolates do not absorb this dye, so its use helps in selecting typical colonies [29]. Plates were incubated at 29 ± 1 °C for 10 days. Bacterial colonies exhibiting morphological characteristics typical of rhizobia were selected, purified, and maintained on YM slants at 4 °C [30].

Cultural, morphological, and physiological characterization of the isolates was conducted to identify traits consistent with rhizobial bacteria. For the evaluation of cultural characteristics, the isolates were grown on YM agar supplemented with Congo Red (pH 6.8) and incubated at 28 °C for 10 days. Colony diameter (mm), color, and mucus production were recorded. Colonies with diameters of 1–2 mm were classified as small, whereas those larger than 2 mm were considered large. For morphological characterization, Gram staining was performed, and cell morphology, Gram reaction, and the presence or absence of spores were determined.

The physiological characteristics evaluated included growth rate and acid or base production. Growth rate was determined by monitoring the appearance of visible colonies on YM medium with Congo Red every 24 h for ten days. The isolates were classified as fast-growing (1–2 days), moderately growing (3–5 days), or slow-growing (more than five days). Acid or base production was assessed on YM medium (pH 6.8) supplemented with bromothymol blue (0.5% in 0.016 N NaOH). A color change from green to yellow indicated acid production, whereas a shift from green to blue indicated base production [30].

Symbiotic effectiveness was assessed under controlled conditions using siratro [Macroptilium atropurpureum (Moc. & Sessé ex DC.) Urb.] as part of the characterization. This plant is widely used as a trapping species and to authenticate the nodulation capacity of rhizobial strains from different legumes due to its ability to establish symbiosis with many rhizobial genera [31,32]. The isolates were grown in YM medium for 18 h at 30 °C and 150 rpm. Siratro seeds were provided by the Experimental Station of Pastures and Forages of Cascajal, Villa Clara, Cuba. The seeds were surface-disinfected with 70% ethanol for 5 min and rinsed once with sterile distilled water. To thin the waterproof testa of the siratro seeds, a previously described methodology using concentrated sulfuric acid was followed [33], immersing the seeds in sulfuric acid for 10 min. Subsequently, they were placed in a 25% (v/v) sodium hypochlorite solution for 15 min and washed ten times with sterile distilled water. Finally, the seeds were pre-germinated on water agar (0.75% w/v) at 30 °C in the dark for 24 h.

Seedlings were transferred to 70 mL bottles containing 50 mL of Norris and Date semi-solid medium [34] (one seedling per bottle) and inoculated with 0.5 mL of bacterial suspension (5 × 10⁸ CFU mL⁻¹ in YM medium). Seedlings inoculated with 0.5 mL of sterile YM medium served as the negative control. Five seedlings were used per treatment. Plants were grown under a 16 h light / 8 h dark photoperiod at 70% relative humidity for 30 days, and nodule effectiveness was determined by the presence of a reddish internal color inside the nodules. This indicates the presence of leghemoglobin, a protein synthesized in legume nodules during effective interaction with rhizobia, which serves as indirect evidence of a functional symbiosis and of nitrogen fixation being carried out by the bacteria [35,36]. The isolates that formed effective nodules were subsequently selected for nodulation experiments on C. ensiformis plants under both greenhouse and field conditions.

2.4. Effect of Rhizobial Inoculation on Nodulation and Growth of C. ensiformis

The experiment was conducted between June and October 2021 at the Jibacoa Agro-Forestry Experimental Station (22°00′00.0″ N, 79°50′00.0″ W; 340 m a.s.l.), located in Manicaragua Municipality, Villa Clara, Cuba. Two experiments were performed: one under semi-controlled conditions and the other in the field between coffee rows, both established on Chromic Eutric Cambisol soil. C. ensiformis seeds were provided by the Seed Bank of the Jibacoa Agro-Forestry Experimental Station, Villa Clara, Cuba. Seeds were inoculated with the previously selected putative rhizobial isolates using the seed-coating method, which involved applying the bacterial inoculant to the seeds, mixing thoroughly, and allowing them to air-dry in the shade before sowing. A dose of 4 mL kg⁻¹ of seeds (5 × 10⁸ CFU mL⁻¹ of inoculant) was used. The control treatment consisted of non-inoculated seeds.

2.4.1. Semi-Controlled Conditions

Inoculated C. ensiformis seeds were sown in black polyethylene bags (30 cm height × 20 cm width) containing 13.67 kg of non-sterile Chromic Eutric Cambisol soil, and the bags were placed outdoors at a spacing of 0.40 m × 0.30 m. Three seeds were sown per bag, and after the emergence of the first trifoliate leaf, thinning was carried out to maintain one plant per bag. A total of nine plants per treatment were used, following a randomized block design with three replicates. At 90 days after inoculation, when approximately 80% of the plants had reached the flowering stage, nine plants per treatment were collected to assess nodulation and growth variables. Nodulation parameters included the number of effective nodules, identified by the presence of a reddish internal coloration, and the dry weight of effective nodules (mg), determined after oven-drying at 65 °C. Growth variables included the number of trifoliate leaves and the total plant dry weight (g), both measured after oven-drying at 65 °C.

2.4.2. Field Conditions Between Coffee Rows

C. ensiformis seeds were inoculated using the same bacterial isolates, concentrations, and inoculation procedure described for the semi-controlled experiment. The control treatment consisted of non-inoculated seeds. Seeds were sown between the rows of a two-year-old Coffea arabica L. cv. Isla 6–11 plantation, established at a spacing of 2 m × 1 m (5000 plants ha⁻¹). Two rows of C. ensiformis were planted at a spacing of 0.40 m × 0.20 m (25,000 plants ha⁻¹), within plots of 5 m². A randomized block design with three replicates (blocks) was used. At 90 days after inoculation, when approximately 80% of the C. ensiformis plants had reached the flowering stage, 18 plants (six per block) were randomly selected from each treatment. Growth parameters, including the number of trifoliate leaves and shoot dry weight (g), were recorded after oven-drying at 65 °C. The bacterial isolate that produced the greatest improvement in the evaluated variables was subsequently selected and taxonomically identified using molecular methods.

2.5. Molecular Identification of the Selected Strain

The 16S rRNA gene of the selected bacterial isolate was PCR-amplified and sequenced following the methodology described by Mareque et al. [37]. Universal primers Eub27f (5′-AGAGTTTGATCMTGGCTCAG-3′) and Eub1492r (5′-TACGGYTACCTTGTTACGACTT-3′) were used for amplification. Sequence quality was evaluated using FinchTV v1.4.0 (Geospiza, Inc., Seattle, WA, USA), applying a minimum quality threshold of 20 per base. Up-to-date 16S rRNA gene sequences were retrieved from the National Center for Biotechnology Information (NCBI) nucleotide database [38]. Multiple sequence alignment was performed with MAFFT v7.505 [39], and the best-fitting nucleotide substitution model was determined prior to maximum-likelihood phylogenetic inference using IQ-TREE v2.4.0 with 1000 bootstrap replicates [40]. The resulting phylogenetic tree was visualized and annotated using iTOL [41].

2.6. Statistical Analysis

The datasets obtained from the inoculation assays under semi-controlled conditions (number and dry weight of effective nodules, number of trifoliate leaves, and shoot dry weight), as well as from the field trials conducted between coffee rows (number of trifoliate leaves and shoot dry weight), were evaluated for homogeneity of variance and normality. When these assumptions were not met, mean ranks were compared using the non-parametric Kruskal–Wallis test (p ≤ 0.05). All statistical analyses were performed using InfoStat, and data visualization was conducted in Microsoft Excel 2016.

3. Results

The soil from which the Canavalia ensiformis nodule samples were collected was slightly acidic (pH 6.5) and characterized by low phosphorus and potassium contents. The soils used for the nodulation assays under both semi-controlled and field conditions were also acidic. However, the field soil showed higher pH, organic matter content, and available phosphorus and potassium levels compared to the soil used under semi-controlled condition (

Table 1. Chemical characteristics of Chromic Eutric Cambisol soils from the sampling site, from the soil used in the inoculation assays under semi-controlled conditions, and from field soils collected between rows of Coffea arabica L. cv. Isla 6–11.

	pH (KCl)	OM (%)	P2O5	K2O
			mg 100 g−1 soil
Sampling site	6.5 ± 0.1	2.6 ± 0.1	6.4 ± 0.4	19.4 ± 0.3
Semi-controlled	5.3 ± 0.1	2.1 ± 0.1	8.9 ± 0.8	12.7 ± 0.7
Field conditions	5.8 ± 0.5	3.5 ± 0.2	15.4 ± 3.5	18.5 ± 2.4

Mean values of three composite soil samples, each comprising six subsamples collected at a depth of 0–20 cm. pH was determined potentiometrically. Organic matter (OM, %) was quantified using the Walkley-Black method. Available P₂O₅ and K₂O (mg 100 g⁻¹ soil) were extracted with 0.1 N H₂SO₄ following Oniani’s method, with phosphorus measured by colorimetry and potassium by flame photometry.

).

3.1. Isolation of Putative Rhizobia

Nine colonies exhibiting cultural characteristics consistent with rhizobia on YM medium were selected and classified into four morphotypes based on cultural, cellular, and physiological traits (

Table 2. Cultural, morphological, physiological traits and putative taxonomic assignment of nodule-associated bacteria from Canavalia ensiformis (L.) DC. in Chromic Eutric Cambisol soil.

Isolate	Morphotype	Characterization
		Cultural a	Morphological	Growth (Days)	Acid or Base Production b	Nodule Formation c	Putative Identification d
1R	I	2–4 mm, semitranslucent, mucous	coccobacilli, Gram negative, not sporulated	2	base	−	Not determined
9R					acid	−
13R	II		bacilli, Gram negative, sporulated		base	Not determined
4R	III	1–2 mm, pale orange non-mucous	bacilli, Gram negative, not sporulated	3–5	base	+	Phyllobacteriaceae or Methylobacteriaceae
5R		1–2 mm, light pink, non-mucous		3–5	base	+
6R				3	base	+
7R			coccobacilli, Gram negative, not sporulated	3	base	+
IIB	IV	<1 mm, semitranslucent, non-mucous	bacilli, Gram negative, not sporulated	7–10	base	+	Nitrobacteraceae
IVB				7–10	base	−	Not determined

^a Cultural characterization on YM medium supplemented with Congo Red (pH 6.8). ^b Acid or base production on YM medium supplemented with bromothymol blue. ^c Presence of effective nodules (reddish internal coloration) on roots of M. atropurpureum at 30 days after inoculation under in vitro conditions. (−) no nodules; (+) nodules. ^d Putative taxonomic assignment (family level) based on cultural (colony diameter, color, mucus production), morphological (cell morphology, Gram reaction, presence or absence of spores), and physiological traits (growth rate, acid or base production) [29].

). The first morphotype (I) included two isolates (22.2%) that formed large, semi-translucent, mucous colonies. Gram staining revealed non-sporulated, Gram-negative coccobacilli. These isolates showed fast growth and produced either acid or base in the culture medium. The second morphotype (II) consisted of a single isolate (11.1%) that shared cultural and physiological traits with morphotype I; however, Gram-negative, sporulating bacilli were observed. The third morphotype (III) represented the majority of isolates (44.5%), forming small, non-mucous colonies and consisting of Gram-negative, non-sporulating bacilli or coccobacilli. These isolates exhibited moderate growth and base production in YM medium. The fourth morphotype (IV) included two isolates (22.2%) that produced very small, semi-translucent, non-mucous colonies. These were identified as Gram-negative, non-sporulating coccobacilli, characterized by slow growth and base production in YM medium (Table 2).

3.2. Nodulation Capability

To corroborate symbiotic effectiveness, all isolates were inoculated into M. atropurpureum plants under in vitro conditions, except for isolate 13R (morphotype II), whose morphological traits—particularly the presence of spores—were inconsistent with those typically described for rhizobia. Plants inoculated with isolates 1R and 9R (morphotype I) and IVB (morphotype IV) did not develop nodules on M. atropurpureum roots under in vitro conditions. In contrast, isolates 4R, 5R, 6R, and 7R (morphotype III), as well as isolate IIB (morphotype IV), formed nodules exhibiting a reddish internal coloration, indicative of active nitrogen fixation (Table 2). Based on their cultural, morphological, and physiological traits, the isolates were tentatively assigned to putative rhizobial families. Isolates 4R, 5R, 6R, and 7R were assigned to the families Phyllobacteriaceae or Methylobacteriaceae, whereas isolate IIB was assigned to the family Nitrobacteraceae. These isolates were subsequently selected for inoculation studies with C. ensiformis under semi-controlled and field conditions between coffee rows.

3.3. Effects of Rhizobia on Nodulation and Growth of C. ensiformis

The inoculation of the previously selected putative rhizobial isolates in C. ensiformis plants grown under semi-controlled conditions showed that isolate 4R induced the highest number of effective nodules (36 ± 9.76). Plants inoculated with isolates 5R and 6R also exhibited a greater number of nodules (16.2 ± 3.8 and 20.2 ± 4.4, respectively) compared with the uninoculated control. No statistically significant differences were observed between plants inoculated with isolates 7R and IIB and the uninoculated control (Figure 1A). Inoculation with isolate 5R resulted in the highest nodule dry weight, exceeding the control treatment by 31.6%. No significant differences were detected between plants inoculated with isolates IIB and 6R and the uninoculated control (Figure 1A). Additionally, inoculation with isolates 4R, 5R, and 7R increased the number of trifoliate leaves at 90 days post-inoculation, with isolate 5R producing the highest values. No significant differences were detected among plants inoculated with isolates 4R, 6R, and 7R, nor between the control and plants inoculated with isolates IIB and 6R (Figure 1B). Moreover, inoculation with isolates 5R and 7R enhanced plant dry weight by 68.9% and 23.5%, respectively, compared with the uninoculated control, whereas no significant differences were detected for isolates IIB, 4R, and 6R (Figure 1B).

Field inoculation trials of C. ensiformis plants grown between coffee rows demonstrated that all isolates significantly increased the number of trifoliate leaves at 90 days post-inoculation. Plants inoculated with isolates 5R and 6R exhibited the highest values, with no significant differences compared to those inoculated with isolate 4R (Figure 1C). Analysis of shoot dry weight showed that inoculation with isolates 4R, 5R, 6R, and 7R significantly increased this variable. The highest values were again observed in plants inoculated with isolates 5R and 6R, which increased shoot dry weight by 24.2% and 30.8%, respectively, relative to the control. No significant differences were found between uninoculated plants and those inoculated with isolate IIB (Figure 1C). These results indicate that certain native rhizobial isolates, particularly 5R and 6R, exhibit strong symbiotic performance and growth-promoting potential in C. ensiformis under both semi-controlled and field conditions. Based on these findings, isolate 5R was selected as the most promising for stimulating nodulation and plant growth in C. ensiformis and was therefore chosen for molecular identification.

3.4. Identification of Promising Bacterium

The 16S rRNA gene sequencing results indicated that isolate 5R belongs to the genus Phyllobacterium within the family Phyllobacteriaceae (Figure 2). Phylogenetic analysis showed that the isolate clustered closely with the reference strains Phyllobacterium brassicacearum and Phyllobacterium zundukense.

4. Discussion

The results of our analyses confirmed the presence of diverse bacterial taxa in the nodules of C. ensiformis plants grown in a Chromic Eutric Cambisol soil from the eastern region of Cuba. This diversity was reflected in the distinct cultural, morphological, and biochemical characteristics of the isolates obtained on YM medium, which is traditionally used for rhizobial studies. However, this medium can also support the growth of bacteria that are not taxonomically classified within the rhizobial group. Similar observations have been reported for chickpea (Cicer arietinum), broad bean (Vicia faba), and cowpea (Vigna unguiculata), where more than 90% of isolates recovered from surface-sterilized nodules and cultured on YM medium were later identified, through partial 16S rRNA gene sequencing, as non-rhizobial bacteria [42].

The growth of isolates displaying cultural characteristics that deviate from those typically described for rhizobia on YM medium may help explain the results of the in vitro nodulation tests, particularly for isolates 1R, 9R, and IVB, which did not induce nodule formation on siratro plants. It is well established that legume nodules can harbor not only rhizobia but also a diverse range of non-rhizobial bacterial taxa, commonly referred to as nodule endophytes. More than 80 genera have been reported within legume nodules; these bacteria can penetrate root tissues and occupy the nodule through infection threads formed during the early stages of the rhizobium-legume interaction [43,44]. Although these microorganisms are known to promote legume growth and enhance symbiotic performance with rhizobia, their functional roles remain an active area of investigation. Another possible explanation is the low symbiotic efficiency between siratro plants and isolates 1R, 9R, and IVB. Siratro is widely used as a model legume because of its versatility in establishing symbioses with rhizobia belonging to different genera. However, variability in symbiotic compatibility has been documented. For instance, Bradyrhizobium elkanii and Bradyrhizobium japonicum strains exhibit differential symbiotic efficiencies on siratro. In a previous study, the symbiotic effectiveness of 34 isolates, initially selected for their ability to nodulate soybean (Glycine max [L.] Merrill), was evaluated 35% of them failed to produce nodules in siratro plants [45]. Together, these findings indicate a moderate level of specificity between certain rhizobial isolates and siratro during symbiosis establishment.

On the other hand, the in vitro inoculation assays conducted to confirm the nodulating ability of the isolates showed that nearly 70% of them induced the formation of nodules with a reddish internal coloration in siratro plants (Table 2). A relevant study conducted in Cuba reported the isolation, characterization on YM medium, and partial 16S rRNA gene sequencing of rhizobia belonging to the genera Rhizobium, Sinorhizobium, Bradyrhizobium, and Mesorhizobium from nodules of forage legumes such as stylo [Stylosanthes guianensis (Aubl.) Sw.], spurred butterfly pea (Centrosema molle Benth.), kudzú [Pueraria phaseoloides (Roxb.) Benth.], and siratro [19].

In this study, we initially aimed to establish a taxonomic framework for the isolates based on their cultural, morphological, biochemical/physiological, and symbiotic characteristics. Based on this evidence, it is possible to propose at least a taxonomic family for those isolates that (i) originated from surface-disinfected nodules, (ii) exhibited typical colony morphology, a defined growth range, and the ability to produce acid or base on YM medium, and (iii) formed effective nodules with a reddish interior in in vitro assays, in accordance with Koch’s postulates [29]. Genera such as Rhizobium, Ensifer, Allorhizobium, and Shinella, although sharing certain phenotypic characteristics, may differ substantially in genetic composition, physiology, and symbiotic compatibility with different plant species. This observation is consistent with previous studies on isolates obtained from forage legumes [19,46]. Considering these aspects, a higher-order taxonomic classification, specifically at the family level, was assigned to the isolates that fulfilled the three criteria listed above.

Isolates 4R, 5R, 6R, and 7R were designated as potential members of the families Phyllobacteriaceae or Methylobacteriaceae. Differentiation between these two families was not possible based on the cultural, morphological, and physiological traits evaluated in this study. Both families are known to include isolates that form small colonies (1–2 mm in diameter), exhibit moderate growth within 3–5 days, and generate alkaline reactions on YM medium [47,48]. Additionally, isolate IIB was assigned to the family Nitrobacteraceae. Members of this family produce colonies that resemble those of Phyllobacteriaceae and Methylobacteriaceae, although they tend to be smaller (<1 mm in diameter) and grow more slowly on YM medium, typically requiring 7–10 days for development [29,49]. Previous studies have also reported rhizobial isolates belonging to the families Phyllobacteriaceae and Methylobacteriaceae in nodules of forage legumes such as Crotalaria spp. and Chinese sophora (Sophora flavescens Aiton) [50,51].

Inoculation assays conducted under semi-controlled conditions confirmed the greater effectiveness of the selected isolates compared with the rhizobial populations naturally present in the experimental soil (control treatment). This was evidenced by significant increases in both the number and dry weight of effective nodules across all inoculated treatments (Figure 1A). Previous studies in Cuba reported increases of more than 35% in nodule number and 45% in nodule dry weight in C. ensiformis plants inoculated with rhizobia in a Carbonated Mull Pardo Sialitic soil [6]. However, the present study provides the first evidence demonstrating the positive effects of native rhizobial strains on C. ensiformis grown in a Chromic Eutric Cambisol. Earlier research has emphasized the advantages of using pre-selected isolates adapted to local soil conditions to enhance nodulation and symbiotic efficiency in C. ensiformis, outperforming non-native commercial strains and uninoculated controls [52]. Similar trends have been observed in other green manure legumes such as sunn hemp (Crotalaria juncea L.) and Chinese milk vetch (Astragalus sinicus L.), where native strains of Rhizobium tropici and Mesorhizobium huakuii exhibited superior performance [53]. In Cuba, rhizobial inoculants have also improved nodulation in economically important crops such as soybean and chickpea, achieving increases of over 80% and 75% in nodule number and nodule dry weight, respectively, compared with uninoculated plants [54,55].

Studies have shown that legumes often exhibit greater symbiotic efficiency under field conditions than in semi-controlled environments, as natural settings enable the full expression of key physiological processes involved in effective nodulation. Wu and Guo [56] demonstrated that field-grown plants develop root architectures that more strongly favor rhizobial infection compared with those grown under semi-controlled conditions. In natural environments, heterogeneous nutrient distributions (particularly nitrogen and phosphorus), temperature fluctuations, moderate water stress, and interactions with the native soil microbiota enhance root sensitivity and receptivity to rhizobial infection during flavonoid-Nod factor signaling, thereby regulating nodulation more effectively than the homogeneous substrates typically used in semi-controlled trials [57]. Additionally, the greater photosynthetic capacity achieved under field conditions, driven by higher radiation levels and dynamic moisture and thermal regimes, directly increases nitrogenase activity and symbiotic performance [58]. Field soils also support richer and more complex microbial communities, including plant growth-promoting rhizobacteria (PGPR) that stimulate root growth and nodulation, thereby improving inoculant competitiveness, interactions that are largely absent in semi-controlled environments [59]. Considering these factors, the IIB strain was included in the field trials despite its lack of significant nodulation in C. ensiformis under semi-controlled conditions.

Our results also showed that inoculation with some isolates exerted a positive effect on the number of trifoliate leaves and plant biomass of C. ensiformis under both semi-controlled conditions and in field trials conducted between coffee rows (Figure 1B,C). In terms of biomass accumulation, the average increase observed among inoculated treatments was higher under semi-controlled conditions (≈70%) than under field conditions (≈30%) relative to the uninoculated control. Similar trends have been reported previously, where biomass of inoculated C. ensiformis plants increased by more than 100% in semi-controlled environments and by 30–60% under field conditions [52,60,61]. This pattern is influenced by several factors known to modulate legume-rhizobium symbiosis performance, including water deficit, extreme temperatures, nutrient imbalances, soil type, plant genotype, rhizobial strain, and the degree of rhizobium-legume compatibility during symbiosis establishment [52]. Additionally, differences in light availability between semi-controlled conditions (full sunlight) and field conditions (partial shade within coffee plantations) may further explain the biomass variations observed in this study.

Among the isolates, 5R exhibited the most promising performance under both experimental conditions. This strain induced the formation of nodules in C. ensiformis with the highest dry weight. Similar findings have been previously reported for broad bean and soybean, where selected rhizobial strains significantly increased nodule biomass [54,62]. Nodule dry weight is directly related to the total nitrogen fixation capacity of the symbiotic bacteria, which in turn enhances nitrogen availability for the host legume and promotes overall plant growth [63,64]. In C. ensiformis, a positive correlation between nodule biomass and total plant biomass has also been documented [52,65]. Nitrogen derived from biological nitrogen fixation contributes to increased leaf nitrogen content, enhanced synthesis of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), a key enzyme in photosynthesis, and higher chlorophyll levels. Collectively, these processes support higher photosynthetic rates and carbohydrate accumulation, ultimately improving plant biomass production [66,67].

The reddish coloration observed inside C. ensiformis nodules is indicative of the presence of leghemoglobin. This protein is synthesized by leguminous plants during their interaction with rhizobia. Although its visualization provides only indirect evidence of symbiotic efficiency, it is consistently associated with nitrogen fixation activity inside the nodule. Mathematical analyses in soybean and cowpea have demonstrated a strong functional correlation between leghemoglobin content and nitrogenase activity, confirming that both proteins are tightly linked to N₂-fixing efficiency in legume nodules [68]. Additional studies have shown that Lotus japonicus plants unable to synthesize leghemoglobin exhibit markedly reduced bacterial nitrogenase production and a complete absence of symbiotic nitrogen fixation [69].

Beyond BNF, additional plant growth-promoting mechanisms exhibited by rhizobia (PGPR traits) may also have contributed to the biomass increases observed in inoculated C. ensiformis plants, particularly in those treated with isolate 5R. Phytostimulation through the synthesis and release of secondary metabolites such as phytohormones, as well as improved nutrient availability via phosphate solubilization and siderophore production, are among the principal mechanisms involved [55]. Several of these traits have been previously reported in rhizobial isolates obtained from green manure legumes such as velvet bean (Mucuna pruriens) and sunn hemp (Crotalaria juncea) [70,71].

Phylogenetic analysis based on 16S rRNA gene sequencing revealed that isolate 5R belonged to the genus Phyllobacterium, within the family Phyllobacteriaceae, one of the two families previously proposed during the cultural, morphological, and biochemical/physiological characterization of the isolates (Table 1). Isolate 5R clustered closely with Phyllobacterium brassicacearum strain STM 196 and Phyllobacterium zundukense strain Tri-48, originally isolated from Brassica napus cv. Euro (canola) and from the legume Oxytropis triphylla (three-leaf locoweed), respectively [72,73]. Notably, the latter species has demonstrated symbiotic and nitrogen-fixing capabilities in its host plant. The second family initially suggested was Methylobacteriaceae, which includes symbiotic species within the genus Microvirga, such as M. lupini Lut6, a known symbiont of Lupinus texensis [74]. More distantly related to Phyllobacteriaceae is the family Nitrobacteraceae (Figure 2), to which isolate IIB was tentatively assigned; however, molecular analyses would be required to confirm this. Within Nitrobacteraceae, the genus Bradyrhizobium comprises major nitrogen-fixing symbionts of legumes such as soybean, Vigna subterranea (bambara groundnut), and siratro [75,76,77].

The family Phyllobacteriaceae, to which isolate 5R belongs, is part of the order Hyphomicrobiales within the class Alphaproteobacteria. Members of this family are Gram-negative, rod-shaped or pleomorphic, generally non-spore-forming, and motile by means of polar or subpolar flagella. They are chemoorganotrophic and typically aerobic or facultatively microaerophilic. Phyllobacteriaceae includes several genera commonly associated with plants, colonizing roots, leaves, seeds, soil, and the rhizosphere, and many representatives are recognized as plant-growth-promoting rhizobacteria (PGPR) [78]. Species within this family may function as plant-associated bacteria, including endophytes, rhizosphere colonizers, or nodule-forming symbionts. Notably, some genera such as Phyllobacterium and Mesorhizobium include species capable of forming nitrogen-fixing nodules with legumes [79].

In summary, this study examines a portion of the bacterial microbiota associated with nodules of C. ensiformis, one of the most widely used green manure species due to the benefits it confers on the physical, chemical, and biological properties of soils, as well as on economically important crops such as coffee. Unlike previous studies conducted in Cuba, this research focused on the selection of rhizobial strains symbiotic with C. ensiformis grown in Chromic Eutric Cambisol soils from one of the country’s principal coffee-producing regions. For the first time, a collection of promising rhizobial strains adapted to the edaphoclimatic conditions of this region and capable of establishing effective symbiosis under coffee production scenarios has been obtained.

Although the establishment of C. ensiformis in coffee-growing soils may indirectly increase coffee yield, the incorporation of this legume inoculated with rhizobia would constitute a particularly relevant agroecological strategy for coffee production systems in Cuba. The symbiotic association between this legume and rhizobia selected for their symbiotic efficiency and adaptation to local edaphoclimatic conditions enables highly effective biological nitrogen fixation, increasing the availability of this nutrient in the soil and reducing the need for industrial nitrogen fertilizers, with corresponding economic and environmental benefits. In parallel, the substantial above- and belowground biomass produced by C. ensiformis contributes to the enrichment of soil organic matter, the improvement of soil physical structure, and the enhancement of beneficial microbial activity—factors especially important in coffee plantations located in mountainous regions where erosion, compaction, and reduced fertility are common constraints. Furthermore, its use as a cover crop helps suppress weed emergence, minimize erosion, and promote greater functional biodiversity, thereby strengthening overall system resilience. Taken together, these attributes support the notion that the introduction of C. ensiformis inoculated with rhizobia represents a sustainable, efficient, and agroecologically coherent alternative for improving soil health and productivity in Cuban coffee production.

5. Conclusions

Isolate 5R, belonging to the genus Phyllobacterium, was identified as the most promising candidate for the inoculation of C. ensiformis in Chromic Eutric Cambisol soils from coffee-growing regions of Cuba. This isolate, a member of the native microbiota inhabiting legume nodules, exhibited greater symbiotic efficiency than the resident rhizobia naturally present in the soil, demonstrated by its enhanced rhizosphere colonization and its ability to form effective nitrogen-fixing nodules. Moreover, its capacity to promote the growth of C. ensiformis under coffee cultivation conditions underscores its potential as a bioinoculant for the sustainable and integrated management of coffee agroecosystems in Cuba. Overall, this research supports the selection of promising bacterial strains for the development of inoculants that improve the growth of green manure crops such as C. ensiformis, thereby contributing to more resilient coffee production systems in Chromic Eutric Cambisol soils across key coffee-producing regions of the country.