RESEARCH WORK

Effect of the inoculation with rhizobia from Alberta, Canada, in sorghum (Sorghum bicolor L. Moench), under field conditions

C. J. Bécquer, Beatriz Salas, U. Ávila, L. A. Palmero, J. A. Nápoles, Yamilka Ramos, Ivis Pasarón y Lisbet Ulloa

Instituto de Investigaciones de Pastos y Forrajes, Estación Experimental Sancti Spíritus, Apdo 2228, Sancti Spíritus, Cuba
E-mail: pastossp@yayabo.inf.cu

ABSTRACT

A field trial was conducted with the objective of measuring the effect of rhizobium strains on the agronomic variables of sorghum under the environmental conditions of Sancti Spiritus, Cuba. Ten Sinorhizobium meliloti strains, from livestock production ecosystems of Alberta, Canada, were used; as well as four reference strains belonging to different rhizobium genera and strains, which were from the collection of Agriculture and AgriFood Canada. The inoculi confection and seed inoculation were made by standard methods. The experimental design was randomized blocks, with 16 treatments and four replications. The dry aerial weight, stem length and ear length were evaluated; in addition, the increase of aerial dry weight was calculated in the inoculated treatments as compared to the absolute control. The results proved the capacity of the studied strains to influence the agronomic variables, because the selected treatments equaled their values to those of the fertilized control and showed an increase higher than 100% of the aerial dry weight, with regards to the absolute control.

Key words: Ecosystem, inoculation, Sinorhizobium, Sorghum.

INTRODUCTION

Although there are studies about the possibility of inducing root nodules with rhizobia in plants which do not belong to the legume family (Kalia and Gupta, 2002), the success of the biofertilization of cereals and other plants is based mainly on the stimulation of plant growth through the secretion of hormonal substances by rhizosphere bacteria (Chabot et al., 1996; Antoun et al., 1998; Prévost et al., 2000).

Chi et al. (2005) stated that rhizobia can infect cereal roots and positively influence the physiological indicators of the plant. Hafeez et al. (2004) determined that like Azotobacter, rhizobia also increase the aerial biomass, nitrogen content and other variables in cotton plants.

Sorghum (Sorghum vulgare or Sorghum bicolor) is a grass (from the family Poaceae) which is used as forage, and its seeds are used for meal. It is an important food crop in Africa, Central America and southern Asia (Nápoles, 2006). In Cuba it is indistinctly used, as a source of grain for monogastric animals and of forage for cattle.

In the country there are salinity (one million hectares), erosion (moderate to strong) and high compaction problems, as well as an increase of non fertile soils and deforestation. Sorghum can be a favorable option for agriculture, because it is capable of standing drought conditions due to its highly developed and fibrous root system. The antecedents existing in the country about the crop fertilization are limited, although according to Chaviano (2005) the cereal responds well to organic fertilization.

Before this work, there were no antecedents in Cuba of the utilization of rhizobia in sorghum. Nevertheless, the need to develop this crop on low fertility soils necessarily demands a correct agronomic management in nutrition, which leads to a maximum precision in the ways, sources, doses and moment of fertilizer application. The use of bacterial inoculants can be one of the best alternatives, from the economic and environmental points of view.

The rhizobium strains used in this experiment came from stressing soil and climate environments in Alberta, Canada, which makes them promising for their application in extreme ecosystems. Although the strains from local ecosystems are adapted to the environment and can be more competitive than the imported ones (Bhattarai and Hess, 1993; Neves and Rumjanek, 1997), Alberta Innovates-Technology Futures, Bioresource Technologies of Canada and the Cuban Research Institute of Pastures and Forages were interested in testing the effectiveness of these microorganisms in different geographical zones. Thus, the range of possibilities for their practical and commercial application, as well as the enrichment of biodiversity, would be widened, for which the objective of the trial was to measure the effect of Canadian rhizobium strains on agronomic variables of sorghum, under the environmental conditions of Sancti Spiritus, Cuba.

MATERIALS AND METHODS

For the trial four reference strains were used, belonging to different rhizobium genera and species and 10 Canadian strains of the genus Sinorhizobium, which were isolated from naturalized legume roots (Melilotus and Medicago) in livestock production ecosystems of the Alberta prairies, Canada (table 1); they were identified and characterized by the methods of classical microbiology and molecular biology (Bécquer, C.J. et al., unpublished data). Such ecosystems have a loam, sandy, eroded soil and little rainfall. One of the strains (CAS2) was isolated from legumes adapted to hydrocarbon-contaminated soils (Slaski, J., personal communication). The isolation and identification works were conducted at the Unit of Environmental Technologies, Laboratory of Microbiology of the Alberta Research Council, Vegreville, Canada.

The strains grew in yeast-manitol-agar medium (Vincent, 1970) and were suspended in yeast-manitol-broth (Vincent, 1970) until achieving a cell concentration of 10⁶-10⁸ CFU/mL. The inoculation was made by immersing the seeds for 24 hours in the inoculum, at room temperature, and they were later extracted from the broth to be dried under shade and planted immediately (Sabry et al., 1997). For the absolute control and the non inoculated fertilized control only the yeast-manitol broth was used. The reinoculation of the treatments took place 18 days after planting to ensure an effective presence of the bacteria in the rhizosphere for root colonization, with a bacterial inoculum (10 mL/plant) containing 10⁶-10⁸ CFU/mL, for which a backpack sprayer was used. The backpack spray was aimed at the plant stem base. This activity was carried out in fresh morning hours, to prevent the excessive desiccation of the product.

The climatic conditions of the experimental period are shown in table 2. The planting dose was 12 kg/ha; seeding was done by spaced drilling, with a 50-cm frame between rows. Each plot measured 3 m x 15 m. Four irrigations were applied. Ninety days after planting the harvest was manually performed.

The fertilized treatment consisted in an application of 150 kg N/ha (NH₄NO₃). On the other hand, due to the poor mineral content of the experimental soil (table 3) full fertilizer was applied (NPK: 9-13-17) to all treatments, including the absolute control and the fertilized control, 21 days after seeding, at a rate of 80 kg N/ha.

An experimental design in completely randomized blocks was used (Somasegaran and Hoben, 1994), with 16 treatments and four replications. The data were analyzed through a variance analysis (ANOVA) (StatGraphics Plus, Version 2.0, 1994-1996, Statistical Graphics Corporation). The differences among means were found through the LSD (Least Significant Difference) test of Fisher (p<0,05) (Lerch, 1977) and the correlation coefficient among variables was determined (Ostle, 1984). The agronomic variables aerial dry weight (g/plot), stem length (cm) and ear length (cm) were evaluated. The increase in the aerial dry weight of the inoculated treatments was calculated with regards to the absolute control (%).

RESULTS AND DISCUSSION

The soil of the experimental area (table 3) corresponds to the Alluvial type (Anon, 1979), it has deficit of P₂O₅ and organic matter, which is in correspondence with the report by Hernández et al. (1999) for this soil type. These agrochemical characteristics were taken into consideration to fertilize the experiment in all treatments, where the absolute and the fertilized control were included, and thus stimulate plant growth in their first phenological stages.

This remarkable nutrient deficit in the soil coincides, in general, with some of the stressing conditions that should be faced in sorghum cultivation in Sancti Spiritus, and also propitiate higher reliability on the experimental results, because there is no evidence of a significant interference due to a high availability of macroelements, which could mask the positive effect of the strains, based on their properties as plant growth promoting organisms.

The aerial dry weight (table 4) in the treatments inoculated with the strains ATCC 10004, CAC17, CAC8 and CAS2 showed statistically higher values (p<0,05) than the absolute control. At the same time, such treatments showed common letters with the fertilized control (959,9 g/plot). In this sense, Mia and Shamsuddin (2010) stated that the inoculation of certain rice cultivars with rhizobia increased the dry weight of the plant, as well as the stem length, among other agronomic variables.

Regarding stem length (table 4), the treatments inoculated with the strains USDA 191 (101,3 cm), 25B6 (100,5 cm) and CAS2 (101,6 cm) showed statistically significant values (p<0,05) as compared to the absolute control (92 cm), and in the case of USDA 191 and CAS2, they turned out to be statistically higher than the fertilized control (93,5 cm). The treatments inoculated with the strains ATCC 10004, CAC8, CAC9, CAC14, CAC16, CAC17, CAS2, 25B6, CAC7, ATCC 10317 and CAC4, CAC2 and CAC5 did not show significant differences with regards to the fertilized control and shared common letters with the absolute control, with the exception of the strain 25B6. These treatments constituted most of the ones used in the experiment, and although the effect achieved by these strains in stem elongation was not remarkable, it was proven to be positive, in general, and the strains USDA 191, 25B6 and CAS2 stood out. There are evidences about the positive effect of rhizobia on sorghum and millet growth and yield. Matiru and Dakora (2004) stated that a correct selection of the rhizobial strains and plant varieties can lead to a better effect of biofertilization.

The treatments inoculated with the strains CAC8, CAC9, USDA 191 and ATCC 10004 showed statistically higher values (p<0,05) than the absolute control in ear length (table 4). On the other hand, the ones inoculated with the strains CAC2, CAC5, CAC14, ATCC 10004, CAC16, CAC17, CAS2, 25B6, CAC7, ATCC 10317 and CAC4, in addition to the above-mentioned, showed values with equal letters as the fertilized control. The positive influence of the inoculated strains on the first treatments is obvious. This coincides with the reports by different authors, such as Saubidet et al. (2001) regarding the fact that hormone production by rhizobacteria favors a higher nutrient extraction in the soil, which is shown in the different agronomic variables of the plant. Likewise, if the agrochemical characteristics of the experimental area (table 3) are taken into consideration, which showed the low phosphorus and organic matter availability, it could be inferred that the higher influence of these bacteria could have occurred in the solubilization of phosphates, property which, according to Richardson (2001), is the most common form of action implied in plant growth promotion by rhizobacteria, which increases the utilization of nutrients by the host plant.

It must be stated that the treatment inoculated with the strain CAS2, which did not differ from the fertilized control in the variables aerial dry weight and stem length (table 4), was from root nodules of Medicago sativa plants that grew on hydrocarbon-contaminated soils (Slaski, J. personal communication), which could indicate, in addition to its survival capacity in extreme environments, its high ability to produce plant growth promoting substances. This characteristic turns such strain into promising, for its use in ecosystems which are under stressing physical-chemical factors.

The increase in the aerial dry weight of the inoculated treatments, as compared to the absolute control, is shown in table 4. All the treatments showed values that exceeded the absolute control, but the ones inoculated with the native strains CAC2, CAC8, CAC9, CAC17, CAS2 and CAC7 stood out (with 100% increase or more); while only the treatment inoculated with the commercial strain ATCC 10004 surpassed this value; it expresses a higher effect of the native strains on this agronomic variable of the plant.

When relating ear length to stem length in the treatments inoculated with the native strains and in the ones inoculated with the reference strains, correlation coefficients of 0,19 were found for the ear length-native strains, which indicates that the variable stem length was not directly interrelated to the former, perhaps due to the intervention of external factors, such as the edaphoclimatic conditions, or internal factors, such as the insufficient production of certain hormones for these strains. However, for this variable, but in the treatments inoculated with the reference strains, a correlation coefficient of 0,62 was observed, which indicated a moderately strong interrelation of stem length to ear length in such treatments.

The aerial dry weight as compared to stem length, in the treatments inoculated with the native strains, was observed to show a correlation coefficient of 0,79 (p<0,01), which showed a strong interrelation of stem length with aerial dry weight; while in the treatments inoculated with the reference strains the correlation coefficient was -0,51 (p<0,01), for which a negative interrelation was found between these two variables. On the contrary, the variables ear length and stem length, in the treatments inoculated with the native strains, showed a weak correlation (r = 0,19), while for the treatments inoculated with the reference strains, the correlation was moderately strong (r=0,62).

The results referred to ear length with regards to stem length, like those of aerial dry weight as compared to stem length, seem to have a close relation to the specific secretion of plant growth stimulating substances by the rhizobia of different genera and species. This mechanism of the rhizobacteria-plant interaction and its effect on the different plant parts has not been exactly determined (Saubidet et al., 2002), although according to Sahin et al. (2004), some authors have evidence in this regard. On the other hand, Vessey (2003) stated that the indoleacetic acid produced by rhizobacteria can cause root initiation and cell elongation; the production of cytokinins can favor cell division and tissue expansion, as well as gibberellins influence stem elongation. The authors of this work do not know the exact way in which these substances could have influenced the above-mentioned results, but evidently these processes must depend, in one way or the other, on the type and quantity of metabolites segregated by the different rhizobium genera and species which are used in biofertilization. Further research is recommended in this field of study.

In general, the capacity of the studied strains to positively influence growth and other agronomic variables of the plants was proven. In case the effect of these strains on grain production as feedstuff or seed source is to be tested, other evaluations aimed at determining the valuables in that variable type are recommended.