A Malian native Azospirillum sp. Az6-based biofertilizer improves growth and yield of both rice (Oryza sativa L.) and maize (Zea mays L.)

The objective of this study was to improve rice and maize yields using native Azospirillum-based biofertilizer. To reach this objective, samples of rhizosphere soil, non-rhizosphere soil and roots of maize plants were collected from the particular locations of Samanko and Bamako of the south Mali. Thirty-three different colonies of bacteria were isolated from the different samples. Based on their better growth in nitrogen free semi-solid medium, their morphological, biochemical and plant growth promotion characteristic, ten bacterial isolates were identified as Azospirillum isolates following the Bergey’s Manual of Determinative Bacteriology. Ten isolates were selected: Az1, Az2, Az3, Az4, Az5, Az6, Az7, Az8, Az9 and Az10. Strain Az6 showed great potential on both rice and maize production. Therefore, this strain is suggested for large scale rice and maize fields’ application. While the Azospirillum sp. Az5, Az6 and Az10 strains are suggested for large scale application in maize field, which may reduce production cost. Top dressing with 25% of the recommended nitrogen-fertilizer was found to decrease maize grain yield.


INTRODUCTION
In Mali, in recent years, a significant drop in cereal yields, mainly maize (Zea mays L.) which occupies an important place in agricultural production systems in agroecological zones has been registered. Maize is one of the main cereals used to feed people in Africa (Macauley and Ramadjita, 2015;Ranum et al., 2014). Although a food crop, maize represents also a cash crop for many small farmers. It has a high yield potential, and it is easy to prepare and digest. All parts of the plant (stem, leaves, tops and seeds) have economic value. They can all be used to produce a wide variety of food (for Human and animals) and non-food products (Ranum et al., 2014). Maize is one of the most crucial and strategic cereals for achieving food self-sufficiency in Mali. Several factors explain its reduced yields such as soil mineral deficiencies, soil salinity, pH, temperature, heavy metals, phytopathogens, poor farming practices and harmful insects. Among these factors, soil nitrogen deficiencies, soil salinity and phytopathogens constitute the main constraints in Mali.
In recent years, much work in the world has been devoted to the use of soil bacteria to improve plant production (by improving its growth and protection), while reducing the use of chemical compounds which are more are expensive and harmful to Human and the environment. These plant growth-promoting rhizobacteria (PGPR) belong to different genera such as Pseudomonas, Azospirillum, Azotobacter and Bacillus (Sharma et al., 2017). These plant growth-promoting rhizobacteria (PGPR) improve plant growth and nutrition through phyto-stimulatory activities, hormonal mechanisms and phytoprotective functions (antagonism, competition or induction of systemic resistance) (Kannojia et al., 2019). The use of phytosanitary products can impact the soil microbial biodiversity (Annapurna et al., 2013). To do this, a possible strategy is based on the use of microorganisms whose phytobenefic effects would limit the use of chemical inputs (Abiala et al., 2015;Biessy et al., 2019).
Although Mali is among the biggest rice and maize producing countries in Africa, and nitrogen fertilization is more than necessary to increase the production of these cereals, very few studies have been undertaken so far on the use of native free and endophyte Azospirillum as organic fertilizer to improve rice and maize yields. This study aims to isolate effective Azospirillum isolates and formulate efficient as well as easy to use Azospirillumbased biofertilizer to improve maize and rice growth and yields.

MATERIAL AND METHODS
Collection of samples: Maize fields of particular locations of two different sites of the districts of Bamako were selected for sample collection. The locations were Samanko agricultural fields and LaboREM experimental site. Rhizosphere soils were collected from the rhizosphere regions of the maize plant at the depth of 2-3 cm and non-rhizosphere soil sample was collected from 1.80 m away from each plant. The plants were uprooted for root sample and the soils attached to the roots were removed. All samples were taken in different polythene bags and brought to the laboratory. The samples were preserved in refrigerator.
Preparation of samples: By placing the roots under a gentle stream of water soils attached to the roots were removed. When the roots were free of adhering soil, these were thoroughly washed for several times with sterile distilled water. Using sterile scissors, the roots were cut into small pieces. One gram of each root samples was used for the purpose. The roots were rinsed successively in disinfectant 2% Chloramine T and subsequently sterile 0.5 M PO4 buffer (pH-7) five times. Afterwards, these have been washed with sterile distilled water three times; the root pieces were macerated in sterile mortars and serially diluted.

Azospirillum isolation
The isolation of Azospirillum was done according to the technique of Hossain et al. (2015). Briefly, Nfb semi-solid medium in screwcapped tubes was inoculated with 0.1 ml of each sample suspension, using a sterile pipette and was incubated at 37°C for 72 h. After incubation, Azospirillum appeared in the tubes forming characteristic thin dense, white pellicle few mm below the surface at the medium (Dobereiner, 1980). The pellicles were examined microscopically for the presence of gram negative, fibroid and actively motile cells. According to Krieg (1981), a loopful of the pellicle developed in tubes was transferred to fresh semi-solid Nfbmedium in screw-capped tubes and the tubes were incubated at 37°C. The white sub-surface pellicle formed after 72 h in the fresh medium was checked by microscopic examination for the presence of gram negative, curved, motile cells and transferred into the fresh semi-solid Nfb-medium thrice, each transfer being made at 72 h intervals. Then, a loopful of the pellicle was streaked on the plates of Nfb-medium, containing 20 mg yeast extract per liter and solidified with 1.5% agar. The plates were incubated at 37°C for one week. Small, dry, slightly convex and rugose colonies were transferred to the slants of solid malate medium containing 0.1% ammonium chloride. Cultures in the slants were streaked on the Congo red medium to get pure colony. The pure colonies were transferred to the slants of same medium and preserved.

Morphological characteristics of the isolates
From the morphological view point, the isolates have been observed macroscopically and their specific morphology and cultural characteristics have been appreciated by culturing them on potato dextrose agar (PDA) plate media (Rosemary et al., 2013).

Biochemical characteristics of the isolates
The catalase (Venkateshwarran et al., 1999), oxydase (Caridis et al., 1991), cellulase (Gupta et al., 2012), and urease (Varenyam et al., 2010) tests were done according to methods described by the cited authors. Indole, hydrogen sulfide, nitrate reduction, carbohydrate fermentation, Voges and methyl red tests were done using Hossain et al. (2015). In each case, growth of the isolates was recorded by visual observation (Hossain et al., 2015). Bergey's Manual of Determinative Bacteriology (1994) was used to identify the tested isolates.

Experiment 1: Determination of the effect of Azospirillum isolates on the growth of rice
In this experiment, a complete randomized design (CRD) with 11 treatments was used. The Azospirillum isolates and a noninoculated control represented the treatments. Rice cultivar (Adny11), the most appreciated by farmers and consumers in Mali, was used as plant test in this experiment. Each treatment was replicated 3 times. Five seeds of the tested rice variety were seeded directly in a plastic pot (100 mm w x 75mm d x 85mm h) filled with 2.50 kg of air-dried soil from LaboREM-Biotech test site. Rice seeds treated with sterile distilled water alone were considered as control. The pots were held on racks and grown under greenhouse conditions and watered regularly. After one week, the seedling was thinned to two plants per pot. Growth parameters such as shoot length and root length were recorded 45 days after planting. Collected data were subjected to analysis of variance and comparison of means using protected LSD test (P≤0.05). The statistical package SAS (Version 9 -SAS Institute) was used for all analysis (SAS, 1990).

Experiment 2: Determination of the effect of Azospirillum isolates on the growth of maize
Efficiency of Azospirillum isolates on plant growth and nutrient uptake in maize was evaluated under greenhouse conditions by seed bacterization. Bacterization of surface sterilized seeds was performed by imbibing the seeds in each Azospirillum isolate cell suspension (A600=0.5) for 6 h at 60 rpm. Maize and rice seeds treated with sterile distilled water alone were considered as control. Seeds either inoculated with bacteria or untreated were sown in plastic pots (100 mm w × 75 mm d × 85 mm h) filled with approximately 2.50 kg of air-dried soil from LaboREM-Biotech test site. The pots were held on racks in a complete randomized block design (CRD) with the dose of N applied as blocks and the Azospirillum isolates as treatments. Each treatment was replicated two times. The pots were maintained under greenhouse conditions and watered regularly. Growth parameters such as shoot length and root length were recorded 45 days after planting. Grain production (number and weight of the grains harvested per treatment) was recorded after harvesting.

Nitrogen-fixing bacteria isolated
Sixty-six nitrogen-fixing bacteria were isolated from soils and different parts (Rhizoplane, xylem and endosphere) of maize plants, sampled in the two sites (Samanko and LaboREM, Bamako, Mali) investigated in this study. Based on their ability to grow better and faster in Nfb semi-solid medium in screw capped test tubes but not in Nfb agar medium in plates, 10 isolates were selected out for further studies. No isolate was obtained from nonrhizosphere soil, contrary to the rhizosphere soil and the rhizoplane where the maximum number of isolates was obtained. All the bacterial isolates from LaboREM-Biotech location were from the rhizoplane.

Azospirillum bacteria identified
Colony aspects of isolate Az10 on Nitrogen free medium ( Figure 1A) and Congo red medium ( Figure 1B) are presented in Figure 1, while Gram and vegetative cells of Azospirillum sp. Az6 are presented in Figure 2.
Data collected on morphological and biochemical characteristics of the ten nitrogen-fixing bacterial isolates showed that: all the ten selected isolates presented brownish and flat colonies on Azospirillum medium. They were curved, Gram negative, mobile, catalase and oxidase positive. They produced hydrogen sulfide, reduced nitrate and utilized lactose, mannitol and galactose as carbon source. The ten selected bacteria isolate can utilize biotin, but cannot hydrolyze the gelatin nor utilize xylose, glucose and sucrose, as carbon source. Considering all the identified characteristics, Az1, Az2, Az5, Az6 and Az10 were identified as Azospirillum brasinense.
The analysis of data in Table 1, shows that only the Azospirillum isolate Az5 produced all the compounds assessed, followed by the isolates Az3 and AZ8 who were only cellulase negative. Contrary to Az3 and Az8, the isolate AZ9 produced all the compounds but did not solubilize phosphates. Siderophores and cyanhydric acid have been produced by all tested isolates, but the isolates Az2, Az4, Az7 and Az10 were not able to produce the indole acetic acid. All selected Azospirillum isolates can grow normally at a temperature between 22 to 45°C, at pH range of 4.5 to 9.8, and in media containing 40 to 500 mM of NaCl.

Effect of the Azospirillum isolates on rice growth
Inoculation with the isolated Azospirillum sp. strains significantly increased seed germination, plant height, root and shoot fresh and dry weights of rice (Oryza sativa L. cv. Adny11) ( Table 2). No significant effect was observed between the different repetitions. This indicates,      for example, that the percentage germination of seed treated or not by Azospirillum sp. will not be affected differently by the different repetitions. The effects of Azospirillum sp. isolates on germination rate, plant height, root length, fresh shoot and root biomass and stem diameter of rice (O. sativa L. cv. Adny11); showed that seed inoculation significantly enhanced rice seed germination. However, the rate of enhancement varied with bacterial strains. All the Azospirillum sp. strains tested, significantly increased seed germination over the non-treated control. The highest percentage of seed germination was recorded with Az4 (85%) and highest plant height, root and shoot biomass were recorded with Az6, followed by Az3, Az8 and Az5 (Figure 3).

Effect of the Azospirillum isolates on maize growth and production
Inoculation with Azospirillum strains significantly affected leaves number, stem length, grain number and grain weight (Table 3). Grain number and grain weight were also significantly affected by nitrogen level, while no significant effect of nitrogen level on the efficacy of the isolates was observed.
Maize seed inoculation with the Azospirillum isolates significantly enhanced all the analyzed parameters. In average, compared to non-inoculated maize plants, an increase in the number of leaves by 9.83%, the stem length by 32.33%, the number of grains by 90.75% and total grain weight by 70.83% were obtained (Table 4). No significant difference was observed between the Azospirillum isolates tested (Table 4). However, they showed high quality and well-filled ears compared to the control (Figure 4). The analysis of data in Table 5 showed that the application of 25% of the recommended dose of nitrogen, after inoculation with the Azospirillum isolates, decreased the number of maize grains by 45.34% and the total grains weight by 31.40%.

DISCUSSION
In this study, no bacterial isolate was obtained from bulk soil with few nutrients to support high quality growth of Figure 1. Effect of inoculation with of Azospirillum isolates (Az3, Az5, Az6 and Az8) on maize growth compared to non-inoculated maize plants. Table 3. Analyze of variance for the leave number, the stem length, the number of ears, the number of grains/pots and the weight of grains produced (g)/pots of maize (Zea mays. cv. Dembagnuman).  Figure 4. Effect of Azospirillum sp. isolates Az6, Az5 and Az10 on the ears, the filling of the ears and the quality of maize grains produced. Table 5. Effect of application of 25% of the recommended dose of nitrogen after inoculation with the Azospirillum isolates on the number of grains/pots and the weight of grains produced(g)/pots. These results are in conformity with those previously reported by Kabir et al. (1996) in Mali. According to New and Kennedy (1989), wheat rhizosphere provides an ecological niche protected against soil acidity. These results suggest that plant rhizosphere offer a high quantity and diversity of nutrient for Azospirillum and a protective environment for a high-quality growth and activities. In fact, Wang et al. (2017) working on the effects of plant root exudates on the composition of the belowground microbiome, demonstrated that plant species and plant genotype were key factors driving the changes in the belowground bacterial community composition in agro ecosystems. Likewise, Brusamarello-Santos et al. (2017) suggesting to exploit the highly diverse maize genetic resources in terms of beneficial plant-bacterial interactions for optimizing maize growth, with reduced N fertilization inputs. In fact, Rilling et al. (2018) showed a compartmentalization between rhizosphere and root endosphere for both the abundance and diversity of total (16S rRNA) and putative N 2 -fixing bacterial communities on wheat plants, and Johnston-Monje et al. (2016) studying bacterial populations in juvenile maize rhizospheres originate from both seed and soil, concluded that the most common bacterial cells in juvenile maize rhizospheres are seed transmitted. All the Azospirillum sp. strains isolated in this study produce siderophores and some of them produce indole acetic acid. Inoculation of these isolates significantly enhanced rice and maize seed germination as well as allowed healthy plants. In fact, Araújo et al. (2010) reported that faster germination reduces the period of heterotrophism and reduces the chances of attack by soil pathogens. Besides the production of phytohormones, all the isolated Azospirillum sp. strains produce siderophores, and some of them can produce cyanhydric acid and solubilize phosphorus as well as increase the plant height, the number of leaves, the stem diameter of rice and maize plants; and the number of grains and the total weight of maize grains. The results were also supported by the results of Isawa et al. (2010) and Bao et al. (2013), who reported a significant enhancement of rice growth in terms of tiller numbers, shoot length and shoot biomass consecutive to rice seed inoculation with Azospirillum B510, under greenhouse and field conditions. Pandiaranja and Govindara (2012) reported that the strains of Azospirillum help the plants for better growing by means of utilization of various parameters from the soil to the plants.

Nitrogen doses Number of grains/pots
In this study, the application of 25% of the recommended dose of nitrogen after inoculation with the selected Azospirillum sp. strains tested decreased the number of maize grains by 45.34% and the total grains weight by 31.40%. In fact, Zeffa et al. (2018) reported no significant increase in grain yield when inoculation with bacteria from the Azospirillum genus was realized together with nitrogen fertilization, but did not observe a significant decrease in production. According to research results of Steenhoudt and Vanderleyden (2006), Nif gene is inactive in excess N 2 . These results indicate that addition of nitrogen after inoculation with Azospirillum strains is non-additive.

Conclusion
The present studies specify that the isolated strain Azospirillum sp. Az6 was suitable for inoculating on paddy (O. sativa L. cv. Adny 11), while Azospirillum sp. Az6, Az5 and Az10 strains were suitable for inoculating on maize (Zea mays cv. Dembagnuman). Azospirillum sp. Az5, Az6, and Az10 strains produced the highest number of leaves, stem length and yield (number of grains and total grains weight) when compared to the non-inoculated control other isolates. However, only Azospirillum sp. Az6 strain showed great potential on both rice and maize production. Therefore, this strain is suggested for large scale rice and maize fields' application, while the Azospirillum sp. Az5 Az6 and Az10 strains are suggested for large scale application in maize field, which may reduce production cost.