The potential for bradyrhizobia and phosphate solubilising microorganisms to improve soybean (Glycine max (L.) Merr.) production in acid soils in Ethiopia

Abstract

Nitrogen (N) and phosphorus (P) are two significant plant growth-limiting elements that are required in relatively large amounts and often are deficient. Both N and P are typically deficient in soils of Sub Saharan Africa. Amelioration of these deficiencies is difficult in Sub Saharan Africa since chemical fertilizers are prohibitively expensive for small holder farmers. Microbial inoculants that enhance the access to N and P could potentially be used to alleviate N and P deficiencies and provide an inexpensive means of providing these nutrients in this region. In Ethiopia, rhizobial inoculants are currently being promoted to legume growers. Site-specific field experiments have demonstrated yield improvements due to rhizobial inoculants for different legume crops. However, for some of the grain legumes, such as soybean, multi-location demonstrations have shown inconsistent responses to inoculation. The reasons for the variability in the responses to inoculants is, however, not clear. Around forty percent of the arable land in Ethiopia is of low pH (less than 5.5), which is one of the constraints for successful cultivation of crops. Soil acidity limits N fixation through its detrimental effect on legume growth, the survival of rhizobia, and its influence on the symbiotic interactions. In addition, soil acidity reduces the availability of P to plants, further limiting both plant growth and N fixation. Therefore, this study aimed to identify acid tolerant rhizobial inoculants and phosphate dissolving bacterial inoculants as a means to improve soybean production in acid soils of Ethiopia. In doing so, the presence, effectiveness, acid tolerance and diversity of soybean rhizobial populations resident in Ethiopian soils were investigated. Phosphate solubilising microorganisms were also isolated from soils that grew soybean and their effect on soybean yield was investigated. Rhizobial strains isolated from nodules of soybean grown on Ethiopian soils were screened in vitro for their acid tolerance. The acid tolerant strains were then evaluated for symbiotic effectiveness in a controlled environment. Following this, the most effective acid tolerant strains were evaluated in six field experiments in major soybean growing areas of Ethiopia. Inoculation with a commercial rhizobial strain, or two locally-sourced isolates of rhizobia were used as inoculants resulted in improved soybean yield. The yield increase due to inoculation with the commercial strain was consistent and greater than that of other treatments, while the increase due to these two most effective locally-sourced strains was comparable to, or greater than, application of 46 kg N ha⁻¹ in soils, where the resident rhizobial population was ≤ 1.4 x 10³ cfu g⁻¹ soil. For soils with high background rhizobial populations, there was no nodulation response to inoculation. At one of the experimental sites (Bako), the percentage of N derived from the atmosphere (%Ndfa) [dfa subscript] was 55% for the commercial strain and 35% for a local isolate, Bradyrhizobium japonicum strain ES3. Field validation was observed to be a necessary step in the selection of acid tolerant strains of rhizobia, to increase soybean production in Ethiopia. Genetic diversity of twenty of the 55 acid tolerant isolates was determined by comparing their 16S-23S internal transcribed spacer sequences. The acid tolerant strains were found to have high symbiotic and phylogenetic diversity, relative to the type strains. The acid tolerant strains were also shown to be phylogenetically distinct from most of the type strains used, as well as from most of the previously isolated Ethiopian soybean strains. However, multilocus sequence analysis of the core and symbiotic genes are required to determine their exact taxonomic position relative to Bradyrhizobium and other related genera. During acid tolerance screening of rhizobial isolates, P solubilising Bacillus spp. and a Pseudomonas sp. were isolated from soils that grew soybean. Selected phosphate solubilising strains significantly increased the soluble P relative to controls in liquid cultures containing Al, Fe and Ca bound P sources. The increase in available P in the culture solution was accompanied by a decrease of up to 1.7 pH units in the culture media. Strains that dissolved the highest amount of P in liquid cultures were selected and tested in six field experiments in Ethiopia, in separate plots adjacent to the N fixation experiments. Field experiments where soybean was inoculated with phosphate solubilising organisms appeared to show trends of yield increases over the controls, but the increases were not statistically significant. As an example, among the inoculant strains, strain EPS1 resulted in an average yield increase of 13.8% over the control that was not supplied with P. Further investigation of this strain in a new inoculant formulation, such as seed co-inoculation with an effective rhizobial inoculant would be worthwhile. Finally, the presence and abundance of soybean nodulating rhizobia and the N fixing effectiveness of soil rhizobial populations were tested in 55 soils collected from major soybean growing areas of Ethiopia, using the most probable number and whole soil inoculation techniques in a controlled environment. Rhizobial population estimates of the soils ranged from non-detectable to >1.5×10⁴ cfu g⁻¹ soil and 49% of the soils had rhizobial populations of <300 cfu g⁻¹ soil. Soybean plants that received soil suspensions had shoot dry weights ranging between 45 to 142 mg plant⁻¹, compared with 60 mg plant⁻¹ when un-inoculated and 136 mg plant-1 when inoculated with a reference strain (Bradyrhizobium japonicum strain CB1809). Comparison of shoot dry weights of plants inoculated with a reference bradyrhizobial strain (CB1809), soil suspensions or control treatments (un-inoculated and not fertilized) showed that among soils with rhizobial populations of >300 cfu g⁻¹ soil, 13% contained ineffective rhizobial populations, while 15% contained moderately effective and 72% of soils contained effective rhizobial populations. Soils from southwestern Ethiopia had larger and more effective rhizobial populations while soils from South Ethiopia, West Ethiopia, and Assossa areas mostly contained few and/or ineffective populations Therefore, widespread inoculation responses are unlikely in Southwestern Ethiopia, while extension efforts related to inoculation of soybean should be targeted to South Ethiopia, West Ethiopia, and Assossa areas to provide the greatest likely benefit. The results of this study are relevant to the soybean industry, inoculant companies, and the rhizobial inoculation programs of various governmental and non-governmental institutions in Ethiopia and potentially in other African countries with low soil pH where soybean is grown. Low soil pH and the population density of resident soil rhizobia are shown here to be important factors that have contributed to inconsistent responses to soybean inoculation in Ethiopia. The results showed soybean yield in acid soils can be improved as a result of selection and testing of new isolates from a combination of controlled environment screening and field evaluation. Genetic analysis has indicated effective rhizobia are not specific to a particular taxonomic group, or to a particular location, which indicates that an effective strain may be selected randomly from any region.