Introduction

Azospirillum is a Gram-negative, microaerophilic, non-fermentative, and nitrogen-fixing bacterial genus. It has been one of the most studied plant growth-promoting bacteria (PGPB) since its discovery by Martinus Beijerinck in the Netherlands in 1925. However, as a result of the research conducted by Johanna Döbereiner in Brazil in the 1970s, two main characteristics are used to define this bacterial genus: its ability to fix atmospheric nitrogen (N) (Day and Döbereiner 1976) and produce several phytohormones, including auxins, cytokinins, and gibberellins (Reynders and Vlassak 1979; Tien et al. 1979). Consequently, in subsequent studies, these two characteristics have been considered the cornerstone of the effect of this genus on plant growth and crops. Because Azospirillum is one the most studied PGPB worldwide, and it has been commercialized in several South American countries, including Argentina, Brazil, Uruguay, and Paraguay (Okon and Labandera-Gonzalez 1994; Cassán and Diaz-Zorita 2016), a significant amount of knowledge has been accumulated, demonstrating different aspects of the plant-bacteria interaction under in planta and in vitro conditions. It is difficult to identify and quantify the agronomical use of Azospirillum in countries other than those in South America. We are aware of products in Mexico, India, China, the United States (US), South Africa, Australia, and France, but no official information is available about the number of hectares (ha) treated, type of crops, type of products, and strains used. Therefore, this review focuses on its use in the South American countries, and several of the available references presenting the data are either in Spanish or Portuguese.

Major changes in the plant root architecture is the main outcome of inoculation with Azospirillum. It is generally accepted that these developmental responses are triggered by the production of bacterial phytohormones, and more specifically by the biosynthesis of indole-3-acetic acid (IAA) (Cassán et al. 2014). Despite exhaustive efforts to define a single mode of action to explain the plant growth facilitated by inoculation with Azospirillum, the mode is still undefined. However, some hypotheses have been proposed to better understand the benefits of the Azospirillum-plant interaction (Bashan et al. 2004; Bashan and de-Bashan 2010). This review aimed to understand the evolution of the research on the agronomical use of Azospirillum conducted over the last several decades, and to identify its novel use for environmental purposes and biotechnological applications beyond the agricultural industry. Based on the gathered information and new evidence brought to light in the past several years, a novel hypothesis is proposed to explain the plant growth promotion capability of these bacteria.

The genus Azospirillum

The Azospirillum species (Azospirillum spp.) are alpha-proteobacteria that are members of the Rhodospirillaceae family (Baldani et al. 2005). While most of the representatives of this family are found in aquatic environments, Azospirillum spp. have mainly been isolated from soil. Genomic analysis suggests that, throughout the evolutionary process, this genus transicioned from aquatic to terrestrial environments significatively later than the major Precambrian divergence of hydrobacteria and terrabacteria (nearly 2.5 billion years ago), coinciding with the major radiation of vascular plants on land 400 million years ago (Wisniewski-Dyé et al. 2011). However, for scientists, the history of this genus begins in 1925 when Beijerinck first observed a spirillum-like bacterium isolated from garden soil that was able to increase the N content in nitrogen-deficient malate-based media. Beijerinck initially named the organism Azotobacter largimobile; 3 years later, he renamed it Spirillum lipoferum (Beijerinck 1925). For 50 years, the importance of this bacterial genus as a research subject decreased until 1974 when its capacity to form a strong association with plant roots was discovered (Von Bülow and Döbereiner 1975). Another fact that awakened interest in these bacteria was their isolation from several types of soil and the roots of grasses and grain crops (Döbereiner et al. 1976). The genus Azospirillum was first proposed by Tarrand et al. (1978). Initially, A. lipoferum and A. brasilense were the only two species described (Tarrand et al. 1978). Since then, as summarized in Fig. 1 and Supplementary Fig. 1, a total of 22 species belonging to this bacterial genus have been identified, including A. halopraeferens (Reinhold et al. 1987), A. largimobile (Ben Dekhil et al. 1997), A. doebereinerae (Eckert et al. 2001), A. oryzae (Xie and Yokota 2005), A. melinis (Peng et al. 2006), A. canadense (Mehnaz et al. 2007a), A. zeae (Mehnaz et al. 2007b), and A. rugosum (Young et al. 2008). Subsequently, new species have been reported: A. picis (Lin et al. 2009), A. palatum (Zhou et al. 2009), A. thiophilum (Lavrinenko et al. 2010), A. formosense (Lin et al. 2012), A. humicireducens (Zhou et al. 2013), A. fermentarium (Lin et al. 2013), A. himalayense (Tyagi and Singh 2014), A. soli (Lin et al. 2015), and A. agricola (Lin et al. 2016). Three new species were identified in 2019: A. ramasamyi (Anandham et al. 2019), A. griseum (Zhang et al. 2019), and A. palustre (Tikhonova et al. 2019). A. amazonense (Falk et al. 1985) and A. irakense (Khammas et al. 1989) were relocated to separate genera, Nitrospirillum and Niveispirillum, respectively (Lin et al. 2014).

Fig. 1
figure 1

Azospirillum phylogenetic analysis using rpoD sequences obtained from NCBI database. Reference strains of each specie were used. Other members of the Rhodospirillaceae family and B. japonicum E109 were used as outgroups. Analysis was made by Maximum Likelihood method, Tamura-Nei substitution model, and a Bootstrap testing of 1000 iterations (Jones et al. 1992). Bootstrap values ≥ 50 are shown in the corresponding nodes. Evolutionary analyses were conducted in MEGA7 (Kumar et al. 2016)

Full size image

The genus Azospirillum is distributed worldwide and different strains and species have been isolated from several countries, including Argentina, Brasil, China, Taiwan, Korea, Russia, Pakistan, and Irak, among others (Table 1). This genus is considered to be versatile because it has been isolated from different environments (Reis et al. 2015). Although less common, Azospirillum spp. have also been found under extreme conditions, such as saline soil, oil-contaminated soil, fermented products, fermentation tanks, sulfide springs, and microbial fuel cells (Reis et al. 2015; Anandham et al. 2019; Tikhonova et al. 2019). Moreover, one member of the Azospirillum spp. was isolated from the Himalayan valley and others were found in Baiyang Lake (Reis et al. 2015; Zhang et al. 2019). Nearly 100 years have elapsed since the genus Azospirillum was first identified, and the taxonomy information about this type of bacteria continues to grow. Advances in molecular biology allow a better clasification of organisms, and the C. C. Young Group from National Chung Hsing University (Taichung, Taiwan) has made the greatest contribution to this research area (Young et al. 2008; Lin et al. 2009, 2012, 2013, 2014, 2015, 2016). Not only have they discovered a significant number of new species and redistributed other species, they have also made significant advances by developing methodologies for the identification of Azospirillum strains using polymerase chain reaction (PCR) (Lin et al. 2011).

Table 1 Country of origin and source of isolation of Azospirillum species
Full size table

Functional analysis of plant growth promotion

Azospirillum spp. have been associated to several mechanisms to promote plant growth and a wide range of studies have detailed the beneficial effects of inoculation with these rhizobacteria. The improvement of plant growth by Azospirillum spp. has been mostly attributed to their capacity to fix atmospheric N and to produce phytohormones; it is less attributed to the bio-disposition of nutrients, expression of enzymes, synthesis of compounds related to plant stress mitigation, and competition against phytopathogens, among other mechanisms. However, taken individually, none of these mechanisms has been found to be fully responsible for the changes observed in inoculated plants (Bashan and de-Bashan 2010). Azospirillum spp. modes of action were initially explained by the Additive Hypothesis where the effects of small mechanisms operating either at the same time or consecutively create a larger final effect on plants (Bashan and Levanony 1990). In 2010, this hypothesis was replaced by the Multiple Mechanisms Hypothesis, which posits that no single mechanism is involved in the promotion of plant growth; rather, in each case of inoculation a combination of a few or many mechanisms is responsible for the beneficial effect (Bashan and de-Bashan 2010). In the following sections, evidence related to the mechanisms most often studied is summarized to explain the plant growth resulting from inoculation by Azospirillum spp.

N fixation

N fixation was the first mechanism to be identified that demonstrated the way in which Azospirillum positively affects plant growth (Döbereiner et al. 1976; Okon et al. 1983); therefore, many studies have investigated it and a substantial about of information about it has been published (Kennedy et al. 2004; Baldani and Baldani 2005; Bashan and de-Bashan 2010). The emphasis on this mechanism is due to the significant increase in the total amount of N in shoots and grains observed after Azospirillum inoculation in wheat, sorghum, and panicum, among other cereal and grass species (Kapulnik et al. 1981). However, the evidence collected during subsequent decades is controversial. Numerous greenhouse and field experiments demonstrated the contribution of fixed N by bacteria on crops by a reduction in the doses of N fertilizers used under field conditions (for a review, see Bashan and de-Bashan et al. 2010). Incorporation of atmospheric N into the host plant by inoculating with Azospirillum was initially evaluated using the acetylene reduction assay (ARA) and later using isotopic 15N2 and 15N-dilution techniques. ARA has contributed to the understanding of Azospirillum-gramineae associations, but in its use for definitive quantification of biological nitrogen fixation (BNF), it has many disadvantages, mainly due to the fact that it is a short-term assay of enzyme activity and such activity is drastically reduced when plants are disturbed. While isotope techniques (15N) have been more popular, they are not easily adaptable under field conditions due to the uniform labeling of soils and the selection of suitable non-N2-fixing control plants (Boddey and Knowles 1987). Solid evidence that N fixation contributes to the N balance of plants has been mainly based on the observation of an increase in the nitrogenase activity within inoculated roots with sufficient magnitude to increase the total N yield of the inoculated plants (Bashan and Holguin 1997; Kennedy et al. 1997). However, many studies have shown that the contribution of N fixation by Azospirillum to plants is minimal (an increase of 5–18% in the total N of inoculated plants); consequently, plant growth promotion was induced by other mechanisms. These findings almost resulted in the abandonment of the N fixation aspects of Azospirillum, except in pure genetic and molecular studies.

In the last years, several studies have focused on N metabolism within bacterial cells, and many details of molecular mechanisms have been studied in Azospirillum, which is considered a bacterial model for investigating non-symbiotic N fixation. In this sense, during the genomic era, the Sp245 strain of A. brasilense has been used as a model to understand the N metabolism pathways since its genome had been completely sequenced and this strain has been physiologically characterized. The nif gene cluster was identified in two specific positions of the genome; in one case, it was probably codified for an alternative iron or vanadium nitrogenase. The ammonia assimilation in Azospirillum occurs via two pathways, one involving glutamate dehydrogenase (gdhA) under a high NH4+ concentration and the other involving glutamine synthetase (glnA) and glutamate synthase (gltBgltD) under limiting NH4+. The genes involved in both pathways are present in all the Azospirillum species that have been analyzed to date (de Souza and Pedrosa 2015).

Two innovative approaches regarding N fixation research have been developed in the last decades: (a) obtaining the spontaneous ammonium excreting mutants of A. brasilense (see Bashan and de-Bashan 2010) and (b) induction of a specialized sites for N fixation on the roots of legume plants known as paranodules. Externally, paranodules resemble a legume nodule and they can be induced in grasses by exogenous application of auxins (Tchan et al. 1991). Under the premise that Azospirillum does not secrete significant amounts of ammonium obtained from BNF on plant tissues, A. brasilense cells were inoculated into rice and evaluated for their capacity to colonize root paranodules previously induced by treating the roots with auxins. The bacteria colonization of paranodules in the treated plants was correlated with significant increases in plant biomass in comparison to the non-inoculated plants (Christiansen-Weniger and van Veen 1991). Additionally, the nitrogenase activity was significantly higher in the Azospirillum-inoculated paranodules of the roots of the rice plants in comparison to the control plants (Christiansen-Weniger 1997). According to Christiansen-Weniger (1997), this was likely because nitrogenase was less sensitive to the oxygen tension in the paranodules than in the rest of the root. Similar increases in nitrogenase activity were reported by Tchan et al. (1991), Zeman et al. (1992), and Yu et al. (1993) in wheat roots containing paranodules colonized by Azospirillum. In addition to rice and wheat (Katupitiya et al. 1995), paranodules were also obtained on the roots of maize seedlings (Saikia et al. 2004, 2007).

Machado et al. (1991) characterized a spontaneous mutant, HM053, derived from A. brasilense FP2 (Sp7 ATCC 29145, SmR, NalR), which was resistant to ethylenediamine (EDAR). This mutant was able to excrete ammonium and fix N in the presence of high concentrations of NH4+; hence, it is an interesting candidate for use as a biofertilizer to supply N to gramineaceous plants. Machado et al. (1991) suggested that the mutant HM053’s ability to excrete ammonium is related to low glutamine synthetase activity, resulting in a deficiency of NH4+ assimilation; this explains the excretion of excess ammonium produced during N fixation. Pankievicz et al. (2015) showed that Setaria viridis inoculated with the HM053 strain incorporates a significant N level via BNF, and this level may be enough to provide the plant’s daily N demand. Moreover, HM053 was able to promote wheat and barley growth (Santos et al. 2017) and nif expression in planta during wheat root colonization, which was shown to be about 300-fold higher growth than with the wild type strain. The same strain outperformed the parental strain in field experiments, leading to a maize yield increase of up to 28% (Pedrosa et al. 2019). Similar ammonium excreting mutants of A. brasilense have been reported to enhance plant growth (Van Dommelen et al. 2009). Moreover, some of the mutants have been evaluated using the paranodules colonization system (Christiansen-Weniger and Van Veen 1991).

Phytohormone production

Due to evidence reported in studies published over the past 90 years, it is known that the Azospirillum genus is associated with the production of several phytohormones. Simultaneously, Reynders and Vlassak (1979) and Tien et al. (1979) reported the capacity of Azospirillum to produce indole-3-acetic acid (IAA) under in vitro and in vivo conditions, respectively. Additional investigations revealed the capacity to produce cytokinins (Tien et al. 1979), gibberellins (Bottini et al. 1989), ethylene (Strzelczyk et al. 1994), and other plant growth regulators, such as abscisic acid (ABA) (Kolb and Martin 1985), nitric oxide (Creus et al. 2005), and polyamines, such as spermidine, spermine, and the diamine cadaverine (Thuler et al. 2003; Cassán et al. 2009). The plant growth regulators and phytohormones produced by Azospirillum have been summarized and ranked according their effects on plants in previous reports (see Table 1, Cassán and Diaz-Zorita 2016). In a culture medium, the concentrations of the most important groups of plant hormones produced by this bacterium, such as auxins, cytokinins, and gibberellins, increase with bacterial growth because these compounds are continuously accumulated in the medium according to a batch fermentation model (Ona et al. 2003; Cassán et al. 2009; Molina et al. 2018). Based on the active principles of inoculants, both the bacteria (cell number) and the metabolites (mainly phytohormones) are biosinthetized, released, and accumulated in the culture medium. Then, inoculants with a different metabolite profile should have a different capacity to promote the growth of inoculated plants, even if the number of cells is equal. In the case of seed inoculation, the use of inoculants containing Azospirillum and phytohormones in the culture medium will produce a “seed priming” effect. In this sense, Okon (1982) reported that, after seed inoculation, the number of viable Azospirillum cells decreases very rapidly. Then, the short-term benefits of seeds inoculation should not be strictly related to the presence of the bacterial cells in the inoculant; instead, they are, at least partly, related to the presence and concentration of several phytohormones and plant growth regulators. This has been defined as the hormonal effect of inoculation (Cassán et al. 2014).

Auxin metabolism

Auxins are a group of plant gowth regulators that are involved in numerous aspects of plant growth and development (Teale et al. 2006). IAA is the predominant plant gowth regulator found in plants. It is acknowledged that 80% of rhizobacteria, including Azospirillum, are able to produce IAA and the synthesis pathways are similar to those found in plants (Spaepen et al. 2007). At present, members of the genus Azospirillum have provided an excellent experimental model for investigating the physiological role of auxins in PGPB-plant interactions, and several naturally occurring auxin-like molecules have been described as products of bacterial metabolism. The genome sequence of A. brasilense Az39 revealed the existence of all the genes involved in the indole-3 pyruvate (IPyA) pathway: hisC1 coding for an aromatic amino transferase, ipdC coding for an indole-3-pyruvate decarboxylase, which is considered to be the key enzyme of this pathway (Broek et al. 1999), and an aldehyde dehydrogenase gene (see Table 1, Cassán et al. 2014). For the Sp245 and CBG497 strains, only the hisC1 and ipdC genes were identified; no evidence of aldehyde dehydrogenase was observed in these genomes. Considering that the genome sequences of A. brasilense Sp245 and Az39 are very similar, it is not surprising that all the genes encoding for the IPyA pathway are very similar in both strains. No evidence has been found for the existence of ipdC or aldehyde dehydrogenase in the genome sequence of A. lipoferum 4B. Only a putative aromatic amino transferase sequence with homology to AAT1 from A. brasilense Sp7 has been identified (Wisniewski-Dyé et al. 2011). Azospirillum sp. B510 genome sequence analysis revealed a putative aromatic amino transferase with homology to AAT1 from A. brasilense Sp7 (Wisniewski-Dyé et al. 2011). Kaneko et al. (2010) proposed that two candidate genes are involved in the indole acetamide (IAM) pathway, but we question their role in IAA biosynthesis due to the low similarity (especially for the putative iaaM gene) between them and the known iaaM and iaaH genes. Finally, gene encoding nitrilases have also been identified in the Azospirillum sp. B510 genome (Wisniewski-Dyé et al. 2012).

In addition to IAA, other molecules, such as indole-butyric acid (IBA) (Martínez-Morales et al. 2003), phenyl acetic acid (PAA) (Somers et al. 2005), indole-3-lactic acid (ILA), indole-3-ethanol and indole-3-methanol (Crozier et al. 1988), indole-3-acetamide (IAM) (Hartmann et al. 1983), indole-3-acetaldehyde (Costacurta et al. 1994), tryptamine (TAM), and anthranilate (Hartmann et al. 1983), have been identified in an Azospirillum spp. culture medium. At least four different IAA biosynthesis pathways have been proposed in Azospirillum spp.: the tryptophan-dependent pathways IPyA, IAM, and TAM, and a putative tryptophan-independent pathway (Prinsen et al. 1993). Despite this diversity, IPyA is considered to be the most important pathway for IAA biosynthesis in this genus. The question about why some bacteria are able to produce phytohormones remains unanswered; however, in the case of auxins, a co-evolutionary mechanism could be hypothetized. Plants release several compounds, such as amino acids and organic acids, into the rhizosphere through root exudates. In the case of amino acids, and particularly for L-trp, this precursor could be used by auxin-producing bacteria to biosynthetize IAA. This molecule increases the amount of this hormone in the rhizosphere, which induces changes in the plant, increasing its root morphology and growth. Thus, a higher amount of root exudate in the rhizoshere will increase the availability of nutrients for the bacteria living in the rhizosphere, enhancing their population. Higher levels of IAA in the rhizosphere will induce a higher ipdC gene expression by Azospirillum, thereby enhancing the IAA concentration in the rhizosphere and stimulating root growth. In other words, some bacteria are able to increase their own population within the rhizosphere by producing IAA using the L-trp produced by plants as a co-evolutionary mechanism. How do plants regulate the IAA levels in the rhizosphere? This should be the most important question for this model; the answer is related to the ability of the plant to regulate the release of L-trp and other amino acids in the rhizosphere by the exudates. In this sense, the full IAA metabolism of A. brasilense has been recently revealed (Rivera et al. 2018). Rivera et al. (2018) found that some amino acids, such as L-met, L-val, L-cys, and L-ser, inhibit bacterial growth and reduce IAA biosynthesis, while the expression of ipdC and IAA biosynthesis, but not bacterial growth, are affected by L-leu, L-phe, L-ala, L-ile, and L-pro. Furthermore, L-arg, L-glu, L-his, L-lyis, L-asp, and L-thr do not affect bacterial growth, IAA biosynthesis, or ipdC gene expression; this fact should have some impact on the rhizosphere during plant-microbe interactions (see Fig. 2, Rivera et al. 2018). It was also confirmed that the A. brasilense strains Sp245, Az39, and Cd can only produce IAA in the presence of L-trp (biosynthesis); these strains are unable to degrade auxins (catabolism), conjugate IAA with sugars and/or L-amino acids (conjugation), or hydrolize conjugates to release free IAA (hydrolysis). IAA biosynthesis was also evaluated under abiotic and biotic stress conditions; it was found to increase with daylight or in the presence of PEG6000, ABA, salicylic acid (SA), chitosan, and a filtered supernatant of Fusarium oxysporum. In contrast, exposure to 45 °C or treatment with H2O2, NaCl, Na2SO4, 1-aminocyclopropane 1-carboxylic acid, methyl jasmonate, and a filtered supernatant of Pseudomonas savastanoi decreases IAA biosynthesis (Molina et al. 2018).

Fig. 2
figure 2

Changes in the architecture of the 12-day-old seedling root system of Arabidopsis. Seedlings were grown for 7 days on MS medium, then were inoculated for 5 days with A. brasilense Az39 or Az39 ipdC-mutant deficient in IAA production. Images show different zones of the root: basal, middle, and apical root. Primary root (PR), lateral roots (LR), hair root (HR), root cap (RC). Photography credits: Mora V

Full size image

Root growth phytostimulation

Roots are the plant organs that are preferentially modified by Azospirillum (see Bashan and de-Bashan 2010). In the 1990s, enhanced water and mineral uptake by roots was frequently used to explain the beneficial effects of Azospirillum inoculation (see Bashan and Levanony 1990; Bashan and Holguin 1997). Increased mineral uptake and water absorption have been related to changes in root growth, architecture, and volume instead of any specific metabolic enhancement process (Murty and Ladha 1988). This fact has been strictly related to the bacterial capacity to produce phytohormones. However, the descriptive data presented thus far have not shown whether these improvements are the cause or the result of other mechanisms (Bashan and de-Bashan 2010). The first evidence of phytostimulation by A. brasilense was observed in pearl millet and sorghum seedlings, and it was similar to that observed by exogenous application of IAA (Tien et al. 1979). Later, it was shown that inoculation of Beta vulgaris increased the number of lateral roots in the inoculated plants in comparison to the uninoculated plants. This effect was correlated with the high levels of IAA produced by bacteria in a pure liquid culture medium and it was mimicked by the exogenous addition of similar concentrations of the phyhormone (Kolb and Martin 1985). The current model of root growth phytostimulation by Azospirillum includes a number of morphological changes that can be summarized as follows: (1) decrease in the elongation of the main root (Dobbelaere et al. 1999; Spaepen et al. 2007); (2) increase in the lateral and adventitious roots (Fallik et al. 1994; Molina-Favero et al. 2008); (3) increase in the number of root hairs (Okon and Kapulnik 1986; Hadas and Okon 1987); (4) branching of the root hairs (Jain and Patriquin 1985); and (5) significant increase in the root surface and volume, probably related to the improvement in water and nutrient acquisition (Spaepen et al. 2014). Modifications in the root architecture mediated by Azospirillum have shown that there is an IAA-dependent response to inoculation. However, recent evidence suggests that other molecules or cell components would be able to induce an IAA-like response to inoculation (IAA independent response). In this sense, and as shown in Fig. 2, A. brasilense Az39 is able to induce the typical root phytostimulation effect in Arabidopsis thaliana under in vitro conditions due to IAA production. However, inoculation with A. brasilense Az39 ipdC- (a non-IAA producer mutant) still induced a stimulatory effect similar to the one induced by IAA on Arabidopsis roots (V. Mora, personal communication). This result increases the complexity of the current model and forces us to work with alternative hypotheses to establish the definitive model, which, in spite of many published papers and a significant amount of effort, has not yet been finalized.

The stimulation of plant root growth by Azospirillum induces an increase in the water absorption and nutrient adquisition rates (including N), which clearly improves the assimilation of N in the biomass and, more generally, plant growth. This capacity would be mediated by the bacterial colonization of the roots and/or their ability to produce different phytohormones, mostly during early stages of plant development. Consequently, the increase in the root biomass would increase the supply of root exudates into the rhizosphere, which would increase the bacterial population associated with the roots and improve their ability to colonize this organ and the rest of the plant. Once the plant is colonized with a high number of bacteria, e.g., > 105 cfu g−1 according to Okon (1982), these cells would be able to provide the plant with significant amounts of NH4+ via BNF. During the advanced stages of plant development, this would have a greater impact on the N economy for the plant. In summary, the Eficient Nutrients Acquisition Hypothesis by inoculated plants would depend on both biological N fixation and phytohormone biosynthesis by the effectively colonized bacteria.

The impact of Azospirillum inoculation on agriculture

Worldwide, the market of inoculants containing Azospirillum spp. is flourishing in South America. Here, the inoculation was initially focused on cereal production, but nowadays, and mostly in Brazil, inoculation is additionally focused on legumes, such as soybeans, combining it with rhizobia inoculants (co-inoculation). The changes in plant growth observed by Azospirillum inoculation and the bacterial capacity to improve the negative effects of abiotic stress on crops has attracted the attention of researchers interested in developing field applied studies (Okon and Labandera-Gonzalez 1994). Okon et al. (2015) suggested that because the diverse modes of action of Azospirillum mostly stimulate plant root growth, inoculation with this microbe could contribute to the increase and stabilization of crop production. However, evaluations of the efficacy of Azospirillum under current crop management practices and at regular environmental conditions are scarce and have been conducted on different crops and in different regions.

Based on 347 trials obtained from 12 countries, including Brazil, Argentina, and several countries in Southeast Asia, and 47 published articles, mainly focusing on maize and other cereals, the impact of Azospirillum inoculation has been analyzed (Díaz-Zorita et al. 2015). From this analysis, the greatest contribution of Azospirillum inoculation to grain yield was observed in winter cereals followed by summer cereals and other crops (Fig. 3). The reviewed studies on inoculation with Azospirillum showed variable results and a multiplicity of interactions related not only to crop management practices but also to environmental conditions. Most field assays have been performed in single geographical locations during one or two consecutive seasons. Thus, of the ability to analyze the performance of bacterial inoculation under random temporal and spatial conditions is limited.

Fig. 3
figure 3

Mean contribution of the inoculation with Azospirillum sp. on crop grain production reviewed from 47 worldwide published field trials under regular production practices. (Adapted from Díaz-Zorita et al. 2015)

Full size image

Based on a total of 316 field experiments performed in the pampas region (Argentina), the relative yield increase in maize due to inoculation with A. brasilense showed positive results, ranging between 66 and 80% of positive responses in comparison with untreated control (Díaz-Zorita et al. 2015). Among the growing seasons, the relative contribution of Azospirillum to maize yield increases under conditions with less rainfall during the early growth stages (Supplementary Fig. 2). In wheat, the early season effects of inoculation with Azospirillum decrease if favorable growing conditions occur during the seed filling stage (Kazi et al. 2016). Okon and Labandera-Gonzalez (1994) and Díaz-Zorita and Fernández-Canigia (2009) found that the grain production responses to inoculation with Azospirillum spp. in wheat and other crops were successful in 70–80% of the cases, regardless of the production conditions. In part, this behavior is caused by the complexity of the impact of Azospirillum on plants interacting with the impact of several abiotic stressful conditions. Kazi et al. (2016) reported that azospirilla inoculation increased the bacteria population in the rhizosphere during the early stages of growth. Most of the benefits have been observed during the early growth stages of plants with greater and more consistent responses seen in the root and shoot dry matter production and a minimal contribution to the grain yield components during the seed filling period (Díaz-Zorita and Fernández-Canigia 2009; Veresoglou and Menexes 2010). Based on the analysis of 480 greenhouse and field experiments, Veresoglou and Menexes (2010) validated the benefits of wheat inoculation with Azospirillum, but they considered variable responses based on differences in the management practices, such as N fertilization, wheat genotype, or Azospirillum strain, that were used to inoculate the crops. Although N fertilization benefits wheat production (Saubidet et al. 2002), the relative contribution of Azospirillum decreased as the dose of the N fertilizer increased. Under high N availability, the bacterial response was not observed (Ozturk et al. 2003).

The combined inoculation of legumes with rhizobia and azospirilla, defined as co-inoculation, could improve plant performance due to the complementary nature of the mechanisms of both bacteria. In soybean crops, co-inoculation resulted in both early initiation of nodule ontogenesis and an increase in the number of nodules, thereby increasing the concentration of N in the shoots and improving the plant growth, particularly under drought conditions (Chibeba et al. 2015; Cerezini et al. 2016). Although the contribution of co-inoculation to the productivity of diverse legume crops is promising, the available information about its use under large production conditions is limited. The results from 21 field trials with alfalfa performed in the pampas region (Argentina) showed that the seed treatment combining Ensifer meliloti and A. brasilense resulted in a response that was nearly two times better than the response obtained from a single inoculation with rhizobia (Díaz-Zorita 2012). Hungria et al. (2013) also reported an increase in grain yields in soybeans and common beans (Phaseolus vulgaris) when combining rhizobia seed inoculation with in-furrow application of A. brasilense at four sites in Brazil. The single inoculation of Bradyrhizobium in soybeans resulted in mean grain yield increases of 8.4% in comparison to the uninoculated control, whereas co-inoculation with Bradyrhizobium and A. brasilense resulted in an increase of 16.1%. For common beans, the single inoculation with R. tropici increased the yield by 8.3%, and co-inoculation of R. tropici and A. brasilense improved the yield by 19.6% (Hungria et al. 2013). The mean soybean yields from 37 field trials under regular management when Bradyrhizobium was co-inoculated with A. brasilense were 227 kg per ha greater than when the soybeans were inoculated with Bradyrhizobium alone and 335 kg per ha greater than the uninoculated control (Nogueira et al. 2018). The mean effects of co-inoculation on soybean nodulation were evaluated under 22 regular crop production conditions; the results showed differences in the effects between tropical-subtropical and temperate environments. On average, the percentage of soybean nodulation increased by around 5% at the Brazilian sites (Hungria et al. 2015; Fipke et al. 2016; Galindo et al. 2018) and around 12% at the Argentinian sites (Benintende et al. 2010; Ferraris and Couretot 2011, 2013; Morla et al. 2019). However, opposite results were found for grain yields. This limited dataset was insufficient to show a consistent and direct relationship between the use of co-inoculation and changes in nodulation and grain yield. Currently, the use of azospirilla inoculants in Brazil is increasing due to co-inoculation. In the state of Parana (Brazil), the use of co-inoculation between 2016 and 2018 increased by almost 30% (Prando et al. 2016, 2018).

Alternative methods of inoculation that are as effective as the standard seed inoculation technique may represent an important strategy to avoid the incompatibility that can occur between the inoculants and pesticides used during seed treatment. However, these technologies need to be thoroughly evaluated before promoting their extensive use. Fukami et al. (2016) described the beneficial effects of spraying leaves with Azospirillum at the beginning of the vegetative phase. Morais et al. (2016) observed that seed furrow inoculation also increased the maize grain yield under current Brazilian production practices. The benefits of foliar inoculation with A. brasilense were evaluated and explained using an auxin signaling model (Puente et al. 2017). The results confirmed soybean growth promotion after seed treatment with B. japonicum and foliar co-inoculation with the IAA producer A. brasilense Az39. Both auxin production and A. brasilense colonization were responsible, via plant signaling, for the positive effects on plant growth and the symbiosis establishment (see Fig. 5 in Puente et al. 2017). An improvement in the nutritional quality of soybean grain due to foliar inoculation with A. brasilense Az39 under greenhouse and field conditions was reported 1 year later (Puente et al. 2018). These findings provide new insights into soybean agricultural technology.

Inoculants formulated with Azospirillum in South America

Currently, the use of azospirilla inoculants for crop production is a consolidated practice in South America (i.e., Brazil, Argentina, Uruguay, and Paraguay), where the extensive agriculture is frequent (Cassán and Diaz-Zorita 2016). In Argentina, Uruguay, and Brazil, there many biological products contain Azospirillum as an active principle. However, the first inoculant in the region was registered 23 years ago (1996) in Argentina with the Servicio Nacional de Sanidad y Calidad Agroalimentaria (SENASA) using the name of Nodumax-L by Laboratorios Lopez SRL (Jesús Maria, Córdoba). It was formulated with A. brasilense Az39, one decade after the isolation and selection of this strain by Enrique Rodriguez Caceres from the Instituto Nacional de Tecnología Agropecuaria (INTA). The inoculant was initially recommended for the treatment of wheat and maize seeds, but it is now recommended for several crops. In Brazil, paradoxically, the first inoculant was registered by Stoller do Brasil SA (Campinas, São Paulo), 14 years after the first one was registered in Argentina. It was named Masterfix L gramineas, and it was formulated with a combination of the A. brasilense Abv5 and Abv6 strains. This product was initially recommended for the treatment of maize and rice seeds, but in the last several years, it has also been recommended in combination with B. japonicum for soybean co-inoculation. Finally, in Uruguay, the first inoculant product was registered in 2015 by Lage y CIA SA (Montevideo, Montevideo) under the name Graminosoil. It contains a combination of A. brasilense Az39 and CFN535. The product was initially recommended for the treatment of maize and sorghum. Currently in South America, there are 106 products (inoculants) produced by 74 companies representing 79 commercial brands. Most of them are produced in Argentina (90 products); 14 products are produced in Brazil and two products are produced in Uruguay. All the available products for commercialization in the Argentinian market are produced in Argentina, but in Brazil and Uruguay, the inoculants are either locally produced or imported from Argentina. All of the products (100%) are formulated with A. brasilense, and the Az39 strain is the active principle in 75% of these inoculants (79 products). In 13 products, Az39 is combined with other A. brasilense strains (one product containing CFN535), Pseudomonas fluorescens (one product), or B. japonicum (11 products). In the last case, this is due to the increase in the number of products registered as a premium technology (co-inoculation) for soybeans. The combination of the A. brasilense Abv5 and Abv6 strains is used to formulate 18 products and the combination of the A. brasilense Az78 and Az70 strains is used to formulate three products. The rest of the azospirilla inoculants are formulated with single strains (Abv5, AzM3, AzT5, 1003, Tuc 27/85, Tuc 10/1, and 11005). Liquid carriers are most often used to formulate these biological products (94%); 6% of the products are formulated on solid carriers, such as peat or bentonite. In 2014, 82% of the formulations in the market were liquid carriers and 18% were solid carriers. This clearly shows the formulation preferences of the companies that are manufacturing these products. The most frequent shelf life of the registered products is approximately 6 months from production with a minimal concentration of 1 × 107 cfu ml−1 in Argentina or 1 × 108 cfu ml−1 in Brazil and Uruguay. Although the use of these biological products has been recommended for 16 types of crops, the registration is mainly for wheat (67), maize (65), sunflowers (16), and soybeans (12). The other plant species recommend for the treatment with A. brasilense are sorghum (Sorghum bicolor) (9), grasses, and winter cereals for grazing (4), rice (5), barley (3), cotton (Gossypium hirsutum) (3), oats (Avena sativa) (2), sugar cane (Saccharum officinarum) (1), tobacco (Nicotiana tabacum) (1), and lettuce (Lactuca sativa) (1). In Brazil, most of the commercialized products are allocated in the maize and soybean grain production market. Based on 2018 data, approximately 7.0 million doses of azospirilla inoculants were commercialized, covering almost 5.0 million ha in South America. In 2014, 3 million ha of plants were inoculated with A. brasilense corresponding to 3.5 million doses of these products. This shows a clear trend in the region of increased use of products formulated with these bacteria.

Extending the use of Azospirillum beyond the agricultural industry

In addition to its proven usefulness in agriculture, Azospirillum possesses the potential to solve environmental problems, such as preventing soil erosion by improving the growth of plants on barren and degraded lands that have lost their capacity to support regeneration, and participating in phytoremediation strategies to decontaminate soils, all leading to healthier environments (de-Bashan et al. 2012). Although these uses are not yet widespread, some examples are presented in this section.

Puente and Bashan (1993) demonstrated that A. brasilense inoculated on the cardon cactus, Pachycereus pringlei, the world’s largest cactus that stabilizes topsoil in its usual scrub habitat in the Sonoran Desert (Mexico), improves the growth characteristics of the plant. In a field trial, three species of cacti inoculated with A. brasilense had a significantly higher survival rate in comparison to the non-inoculated controls. The most important outcome from this trial was the significant reduction in soil erosion and the reclamation of topsoil (Bashan et al. 1999). Growth chamber experiments have demonstrated that A. brasilense enhances enzymes in the phosphogluconate pathway and facilitates the growth of mesquite seedlings (Prosopis articulata) that are cultivated in poor soils (Leyva and Bashan 2008). In a greenhouse environment, the effect of A. brasilense combined with Bacillus pumilus, unidentified arbuscular mycorrhizal (AM) fungi (mainly Glomus spp.) and compost, were measured on the growth of leguminous trees, such as mesquite, yellow palo verde (Parkinsonia microphylla), and blue palo verde (Parkinsonia florida), used in desert reforestation and urban gardening in arid northwestern Mexico and the southwestern region of the US (de-Bashan et al. 2012). The mesquite and yellow palo verde had different, positive responses to several parameters, while blue palo verde did not respond (de-Bashan et al. 2012). Later, seven field trials were undertaken with cardon cacti and the same species of leguminous trees (Bashan et al. 2012). The trial showed that, a decade later, a combination of a legume tree with a cardon cactus, while detrimental to the legume, significantly increased the chances of the cactus surviving and growing in degraded soil. (Moreno et al. 2017). Recently, inoculation of Brachiaria spp. with A. brasilense demonstrated the potential for successful reclamation of degraded pastures in Brazil (Hungria et al. 2016).

In terms of phytoremediation, A. brasilense improved the growth of the shrub quailbush, Atriplex lentiformis, and affected the rhizosphere microbial community in acidic, metalliferous tailings in Arizona (de-Bashan et al. 2010). Tugarova et al. (2013) proved the capacity of A. brasilense strains to reduce selenium (IV) to selenium (0), indicating the possibility of applying Azospirillum as a microsymbiont for the phytoremediation of selenium-contaminated soils; moreover, the bioremediation potential of Panicum virgatum (switchgrass), along with AM fungi and Azospirillum, was tested against lead and cadmium in pot trials (Arora et al. 2016).

In 2000, the Yoav Bashan research group began an interesting study to investigate extending the use of Azospirillum from agricultural plants to aquatic green microalgae (Gonzalez and Bashan 2000). Specifically, they created a synthetic mutualism between the microalga Chlorella spp. and A. brasilense, and they proposed it as a simple, quantitative experimental model to study the beneficial interactions between the plant and the bacteria (Fig. 4). To facilitate the interaction and maintain the mutualistic associations, the two microorganisms were initially immobilized in small alginate beads (de-Bashan and Bashan 2008). The hypothesis behind proposing such an interaction was that, as an unspecified PGPB, A. brasilense would affect green microalgae in ways that were similar to how it impacted higher plants. They found that the effects occurred at all levels, presenting a new avenue for the application of A. brasilense.

Fig. 4
figure 4

Chlorella sorokiniana and Azospirillum brasilense in co-culture. a Auto-fluorescence of microalgae appears in orange while bacteria appears in green, as result of fluorescent in situ hybridization (FISH) using three specific probes targeting Eubacteria (FAM dye) and one specific probe for A. brasilense (CY3 dye). b Scanning electron microscopy (SEM). c C: C. sorokiniana. Az: A. brasilense. Arrows show cells of A. brasilense attached to the microalgae

Full size image

Thus far, physiological studies have shown the effects of A. brasilense on microalgae pigments (de-Bashan et al. 2002), carbohydrates (Choix et al. 2012a, b, 2018), total lipids (Leyva et al. 2015), and vitamins (Palacios et al. 2016). Similar to higher plants, the production of IAA is a key mechanism affecting microalgae (de-Bashan et al. 2008a). Azospirillum enhances the growth of Chlorella spp., Scenedesmus obliquus, and Chlamydomonas reinhardtii (de-Bashan and Bashan 2008; Choix et al. 2018), but it also affects the activities of enzymes, including glutamine synthetase and glutamate dehydrogenase in C. vulgaris. A higher uptake of N from the culture medium and a higher accumulation of intracellular N were observed in the plants inoculated with Azospirillum than those that were not inoculated (de-Bashan et al. 2008b; Meza et al. 2015). Similarly, it was found that Azospirillum had an effect on ADP-glucose pyrophosphorylase, leading to increased accumulation of starch (Choix et al. 2014) and on acetyl-CoA carboxylase, resulting in higher synthesis of fatty acids (Leyva et al. 2014) in microalgae. A direct exchange of N and C between A. brasilense Cd and C. sorokiniana was demonstrated by nanoSIMS (de-Bashan et al. 2016), and the positive effect of the volatile compounds produced by A. brasilense in C. sorokiniana was also reported (Amavizca et al. 2017). Lopez et al. (2019) showed that riboflavin and lumichrome produced by A. brasilense had a significant effect on photosynthetic and auxiliary pigments in C. sorokiniana. The combination has been successfully used for wastewater treatment (de-Bashan et al. 2002; Bashan et al. 2004; Perez-Garcia et al. 2010) and recovery of desert degraded soil after amendment of wastewater debris (Trejo et al. 2012; Lopez et al. 2013).

Overall, these results have extended the use of Azospirillum beyond agriculture to tackling environmental issues, such as revegetation, reforestation, phytoremediation, and wastewater treatment programs.

An overview of the research timeline

Over the last 90 years, studies on Azospirillum-plant interaction have suggested a wide range of mechanisms through which the bacterium enhances plant growth, as summarized in Supplementary Fig. 1.

Despite this body of evidence, two main mechanisms have defined this genus as a model of PGPB: BNF and phytohormone production. The history of the effects of Azospirillum as a bacterium capable of fixing atmospheric N dates to 1976 in Brazil. It was revealed for the first time that A. lipoferum was able to efficiently fix N in the roots of Digitaria decumbens (Day and Döbereiner 1976). This mechanism lost its research importance because the results obtained in greenhouse and field experiments were controversial; however, new mechanisms were proposed to explain the positive effects of inoculation. That was how, at the Katholieke Universiteit Leuven (Belgium), it was demonstrated for the first time that tryptophan was involved in IAA production since A. brasilense was able to convert tryptophan into IAA (Reynders and Vlassak 1979). Meanwhile, a study conducted in the US reported that A. brasilense was able to produce plant growth substances, such as auxins, cytokinins, and gibberellins (Tien et al. 1979). These two reports were the first to show that Azospirillum had the ability to improve plant growth due to the production of phytohormones. Two years later, Oliveira and Drozdowicz (1981) demonstrated for the first time the ability of the genus Azospirillum to produce bacteriocins (molecules able to inhibit bacterial growth) in a pure culture medium. One year later, Reynders and Vlassak (1982) investigated the use of A. brasilense as a biofertilizer in intensive wheat cropping. Simultaneously, a selective culture medium, Congo Red medium (CR), was developed to isolate Azospirillum spp. from soil or seeds (Rodriguez Caceres 1982).

As the interest in the phytohormonal effects of Azospirillum in plants intensified, numerous field studies were conducted in the 1980s to analyze the growth and yield of inoculated crops. Thus, Azospirillum began to emerge as a powerful crop inoculant in co-inoculation systems. The first co-inoculation studies were conducted using A. brasilense and the mycorrhizal fungus, Glomus mosseae, to study their effects on the growth and nutritional quality of maize and ryegrass (Barea et al. 1983). Furthermore, co-inoculation of Azospirillum and Rhizobium was found to have a positive effect on winged beans and soybeans (Iruthayathas et al. 1983). Sarig et al. (1984) reported that the best effects on plants inoculated with Azospirillum were obtained when the culture conditions were sub-optimal. It was first found that the grain yield of non-irrigated Sorghum bicolor increased under abiotic stress when inoculated with Azospirillum (Sarig et al. 1984). As interest in the phytohormonal effects of these bacteria continued, it was observed that A. brasilense was able to produce ABA in a chemically defined culture medium (Kolb and Martin 1985). The production of plant growth substances (phytohormones), classified as cytokinins, by Azospirillum and other related bacteria continued to be analyzed (Horemans et al. 1986). Four years later, gibberellins A1, A3, and iso-A3 were identified in cultures of A. lipoferum (Bottini et al. 1989). Similar results were obtained using A. brasilense (Janzen et al. 1992). Later, ethylene production by Azospirillum was evaluated in chemically defined media modified with the amino acid L-methionine (Strzelczyk et al. 1994).

The arrival of the molecular biology and genomics era shifted the focus to investigating the functional effects of Azospirillum on plants at the molecular level. The first studies to emerge focused on Arabidopsis plants as a model to investigate the A. brasilense-Arabidopsis root interaction system; they demonstrated that this bacterium more than doubled the root hair growth in comparison to the non-inoculated control in a consistent and reproducible way (Dubrovsky et al. 1994). Subsequently, it was established that this effect had a strong phytohormonal component mediated by IAA (Spaepen et al. 2014).

The study of the Azospirillum genome began in 2000 with the analysis of five Azospirillum spp. genomes using pulsed-field gel electrophoresis (Martin-Didonet et al. 2000). This biochemical characterization continued, and new plant growth-promoting mechanisms were proposed. It was found that Azospirillum was able to solubilize insoluble phosphates through the production of gluconic acid (Rodriguez et al. 2004). The same year, the sequence of the pRhico plasmid in A. brasilense Sp7 was analyzed and it was found to have an important role in plant-root interactions and bacterial viability (Vanbleu et al. 2004). In 2005, it was demonstrated that the nitric oxide produced in vitro by A. brasilense Sp245 was a promoter of lateral root initiation in tomato seedlings (Creus et al. 2005). Another interesting mechanism emerged in 2006. Four strains belonging to A. lipoferum isolated from the rice rhizosphere were able to synthesize N-acyl-homoserine lactones (AHLs), which regulate crucial functions for plant-bacteria interactions (Vial et al. 2006). A similar paper reported the production of cadaverine by A. brasilense Sp245 and Az39 (Perrig et al. 2007). Another study confirmed that A. brasilense had the capacity to produce several polyamines, such as putrescine, spermine, and spermidine, under similar culture medium conditions (Thuler et al. 2003). Supporting evidence was reported in later studies. It was reported that A. brasilense Az39 promoted root growth and helped mitigate osmotic stress in rice seedlings, in part due to cadaverine production (Cassán et al. 2009).

In 2010, the complete genomic structure of Azospirillum sp. B510 isolated from stems of rice plants was obtained (Kaneko et al. 2010). That study was the first to report on the genome structure of a member of the genus Azospirillum. In the same year, a new hypothesis about the action of Azospirillum on plants was proposed (Bashan and de-Bashan 2010). A year later, Wisniewski-Dyé et al. (2011) obtained the genome sequences of the model strains A. brasilense Sp245 and A. lipoferum 4B and analyzed the taxonomic origin of this bacterial genus. Through genome sequencing and analysis, they showed that Azospirillum spp. transitioned from aquatic to terrestrial environments. Most of the Azospirillum genes were acquired horizontally, and they encode functions that are critical for rhizosphere-plants adaptation and interaction. In 2014, the complete genome sequence of A. brasilense Az39 was presented (Rivera Botia et al. 2014); it is one of the strains that is most often used for agriculture in South America. One year later, the complete genome sequences of A. brasilense Sp7 (Kwak and Shin 2015) and A. thiophilum isolated from a sulfide spring (Fomenkov et al. 2016) were analyzed and annotated. More recent studies have identified the draft genome sequences of Azospirillum sp. B2, isolated from a raised Sphagnum bog (Grouzdev et al. 2018), A. brasilense strains Ab-V5 and Ab-V6 (Hungria et al. 2018), extensively used as biofertilizers in Brazil, and A. brasilense REC3 (Fontana et al. 2018), isolated from strawberry plants in Argentina. Recently, the quorum-sensing and quorum-quenching mechanisms based on N-acyl-L-homoserine lactones in A. brasilense Az39 were analyzed in silico and in vitro (Gualpa et al. 2019). That study reported that although A. brasilense Az39 this strain is a silent bacterium unable to produce AHL signals, it can interrupt the communication between other bacteria and/or plants via its quorum-quenching activity.

Concluding remarks and perspectives

Since its re-discovery in the 1970s, Azospirillum has become a cornerstone in the study of PGPB. Its potential as an effective inoculant for a wide variety of crops has been recognized. Yet, the exact mode of action is still not completely understood. Azospirillum modes of action were initially explained by the Additive Hypothesis; 20 years later, that was replaced by the multiple mechanisms hypothesis. In this review, we proposed the Eficient Nutrients Acquisition Hypothesis, which posits that plant growth promotion occurs via two major mechanisms, biological N fixation and phytohormone production, which are effectively induced by the colonized bacteria. Thus, some of these molecules have the capacity to alter the root morphology, thereby improving mineral uptake and inducing higher yields, even if using lower doses of chemical fertilizers. The contribution of N fixation is more controversial, and its effect may be less potent than previously believed. Although mixed results have been reported for inoculation, this has not prevented numerous companies around the world from offering inoculants containing Azospirillum. More specifically, in South America, 10 million doses of inoculants containing Azospirillum were used in 2018.

The use of Azospirillum under field conditions has been widely shown to improve plant growth and crop productivity. Thus, the use of azospirilla inoculants for crop production should be understood as a consolidated practice, in terms of grain yield production in summer and winter cereals, as well as legume production (co-inoculation). As an improvement in the use of Azospirillum, co-inoculation with rhizobia has proven to be a novel technology to enhance legume performance. Part of the current challenges of azospirilla inoculants has been the need for inoculant companies to develop effective formulations that can be used for diverse applications and under different storage handling and environmental conditions. In summary, the development of alternative application systems, such as the delivery of azospirilla by foliar inoculation, is seen as a solution to overcoming the limitations of on-seed treatment. There is an urgent need to promote a regional coordinated communication program about the already measured benefits of inoculation with Azospirillum as a complement to current extensive and intensive crop practices. These networks should include direct users of these products as well as other actors from rural and urban environments and local regulatory agencies.

Additionally, Azospirillum inoculation may serve as a valuable method for the remediation of contaminated soil and water and the revegetation and reforestation of degraded lands. Furthermore, the interaction of Azospirillum with green microalgae was proven to be an independent sub-field of Azospirillum research, presenting a new and interesting avenue to produce metabolites, such as lipids and pigments. However, this biotechnological application is yet to be tested under scale-up conditions to evaluate its real-life potential.