BACTERIAL STRAINS DECREASE SOLUBILITY OF POTASSIUM IN THE SOIL ESTIRPES BACTERIANAS REDUZEM POTENCIAL DE SOLUBILIZAÇÃO DE POTÁSSIO NO SOLO

This study aimed to verify whether inoculation with strains of diazotrophic bacteria, with proven ability to solubilize potassium (K) in vitro, contributes towards the release of K in the soil after fertilization with phonolite rock powder. The experiment was conducted in containers with 0.3 dm of soil containing low potassium content. Fifteen treatments were used, namely, 12 inoculated with the bacterial strains, a control treatment (without phonolite and without inoculation), one containing phonolite without inoculation and one containing KCl, soluble fertilizer, without inoculation. In treatments with phonolite and KCl, the doses of these materials were applied to provide the soil with 195 mg dm of K. A completely randomized design with four replications was used. The soil was incubated during 90 days at room temperature and humidity at about 70% retention capacity. After this period, the content of K (Mehlich and resin), pH value and potential acidity (H+Al) were evaluated. Phonolite, associated with inoculation with most bacterial strains, increased the availability of potassium in the soil, pH rate and reduced potential acidity. Among the strains tested, UNIFENAS 100-01, UNIFENAS 100-16, UNIFENAS 100-27, UNIFENAS 100-39 and UNIFENAS 100-93 were the most efficient for the solubilization of K of the phonolite. In spite of the observed results, K content released by the bacterial strains in the soil decreased when compared to in vitro conditions, thus justifying the need for studies on bio-solubilization of soil to select the most efficient strains in the process.


INTRODUCTION
The bio-solubilization of nutrients, such as potassium, in rock minerals, performed by several soil microorganism groups has been studied over and over again due to their capacity for partial or total replacement of imported soluble potassium fertilizers by those found in Brazil. Potassium (K) ranks second as the most absorbed macronutrient by plants (MALAVOLTA, 2006). Its functions within the vegetal metabolism are related to photosynthetic capacity, translocation of carbohydrates, cell osmosis regulation, enzyme co-factor and a greater efficiency in nitrogen fertilization (ANDRADE et al., 2000;MALAVOLTA, 2006;FILGUEIRA, 2008). Lack or deficiency of K in the soil limits the development of plants and their productivity.
Since soils in Brazil normally have low K + rates, high doses of the nutrient are frequently applied through mineral fertilizers. More than 90% of potassium fertilizers used in agriculture in Brazil are imported from Canada, Germany and Byelorussia (DNPM, 2014). Needless to say, high dependence on foreign market increases costs and triggers studies on alternative sources of the nutrient (MARTINS et al., 2015b).
Powder from silicate rocks found in Brazil, such as verdetes, gneisses, granites and phonolites, may be an alternative source for potassium fertilizers (TEIXEIRA et al., 2012;MANCUSO et al., 2014;SANTOS et al., 2015). The solubility of the powder of the above minerals is low and the release of potassium is slow and gradual, with decrease in losses by leaching and soil salinity. It actually provides long term activity (MANCUSO et al., 2014;MARTINS et al., 2015b). On the other hand, slow K release is not a positive characteristic from the nutritional point of view since nutrient absorption by plants is impaired. Deficiency may occur and the development and production of plants may be affected (BRANDÃO; LOPES-ASSAD; CECCATO-ANTONINI, 2014).
The use of physical-chemical treatments, such as high temperature and the addition of NH 4 OH (MARTINS et al., 2015a), triggers the release of nutrients in the rocks. Costs are high and PAREDES FILHO et al. Biosci consumption may become unfeasible. A low-cost and sustainable form, without the generation of possible environmental damages, is the use of K solubilizing microorganisms which increase the availability of K of silicate rocks by the release of organic acids (GIRGIS; KHALIL; SHARAF, 2008;  BRANDÃO; LOPES-ASSAD; CECCATO-ANTONINI, 2014;MEENA;MAURYA;VERMA, 2014;FLORENTINO et al., 2017).
However, most research works above have been developed in in vitro conditions and require evaluations with the microorganisms behavior in the soil. The above is highly relevant due to the fact that other groups of microorganisms and different edaphoclimatic conditions may interfere in the survival and efficiency of K solubilizing microorganisms in the soil (MEENA; MAURYA; VERMA, 2014).
This study aimed to verify whether inoculation by diazotrophic bacteria strains of proven capacity for in vitro solubilization of K contribute towards the release of potassium in the soil after fertilization by phonolite powder.
Assay comprised fifteen treatments: 12 treatments with phonolite and inoculation with bacterial strains; three control treatments: a control treatment (without phonolite and without inoculation); one treatment with phonolite without inoculation; one treatment with KCl alone. Design was totally randomized, with four replications.
Further, 0.25 dm 3 soil lots were weighed and transferred to 0.3 dm 3 plastic containers. Phonolite and KCl dose was applied to provide the soil with 195 mg dm -3 K. Phonolite was mixed with a volume of soil from each recipient. KCl was applied by the solution. Further, 3 mL of bacterial suspension was used on the soil surface of each recipient in inoculated treatments. Assay was performed during 90 days, with humidity at about 70% by periodic weighing of flasks. Distilled water replaced water lost through evaporation. Flasks were maintained in lab conditions, at room temperature.
At the end of the incubation, soil lots were removed from the flasks, air-dried and a sample was retrieved from each flask to determine K + rates, extracted by Mehlich and resin methods. They were quantified by flame photometry; potential acidity (H+Al); pH in CaCl 2 (RAIJ et al., 2001;SILVA, 2009).
Solubility efficiency was calculated according to results obtained from resin extracting solution, according to the formula: where: ES = efficiency of solubility; K soluble treatment = rate of K extracted by resin; K KCl = rate of K in treatment soil with KCl Data underwent analysis of variance and means of the four replications were compared by Scott-Knott test at 5% probability, with Sisvar (FERREIRA, 2011).
K + rates were quantified by extraction methods, Melich (Km) and Resin (Kr). Table 2 provides final pH and potential acidity (H+Al) rates in the soil. Table 2. Rates of residual K + , evaluated by Mehlich method (Km) and by resin (Kr); efficiency of solubility (ES), pH and potential acidity (H+Al) of soil samples.
Means followed by same letter did not differ by Scott-Knott test at 5% probability.
Highest K + rates in the soil, extracted by Mehlich extractor and by resin, occurred in treatment with KCl. K was 5.5 times greater than that in treatment with phonolite without inoculation, and 10.8 times than that in treatment without K. This result is due to the fact that KCl is highly soluble in water. Similar results with regard to KCl were obtained by Duarte, Pereira and Korndörfer (2013). It is actually an undesirable characteristic from the environmental point of view since losses through leaching increase and may cause eutrophication of water courses and underwater waters (LEONARDOS; THEODORO; ASSAD, 2000). There was a 2.4 times increase in K + rates in the treatment with phonolite when compared to control with no potassium fertilizer. Increase in K + rates in the soil through the application of phonolite was also obtained by Teixeira et al (2012) and by Martins et al. (2015b). It has also been observed that phonolite associated with bacterial strain inoculation provided a three-fold increase in K + rates in the soil. The latter were similar to treatment with phonolite without inoculation only in treatments inoculated by strains 100-26 and 100-85.
In their in vitro study, Florentino et al. (2017) used the same bacterial strains and reported that bio-solubility provided K rates ranging between 4.0 and 12.5 times more when compared to treatment without inoculation. This fact showed the great potential of strains to increase K release from phonolite rocks and thus their feasibility in agriculture. K rates triggered by bacterial strains in the soil revealed a significantly reduced effect when compared to in vitro conditions (FLORENTINO et al., 2017). Consequently, there is a need for studies on soil bio-solubility to select the most efficient strains in the process.
Solubility efficiency (ES) may be underscored by strains UNIFENAS 100-01, UNIFENAS 100-16, UNIFENAS 100-27, UNIFENAS 100-39 and UNIFENAS 100-93, with an approximately increase of 28% of K + rate in the soil. According to Meena, Maurya and Verma (2014), soil microorganisms, including bacteria and fungi, have an important role in the solubility of non-available K + inorganic forms, with an increase in availability to plants.
Further, pH rates in CaCl 2 ranged between 4.1 and 5.1. There was an increase in pH of the soil in treatments inoculated with all bacterial strains, except strains UNIFENAS 100-21 and UNIFENAS 100-94, when compared to control. Results differ from in vitro bio-solubility assays which correlate pH decrease of culture medium with the production of organic acids and with greater efficiency in the process of bio-solubility (SHENG et al., 2008).
However, in spite of the great capacity of bacterial strains in in vitro conditions shown above (FLORENTINO et al., 2017), it may be verified that different behaviors occur in the soil with regard to the efficiency in K solubility and in pH rates. Therefore, in-depth knowledge is required on microbial ecology and population dynamics in the soil (RICHARDSON, 2001).
Increase in pH and decrease in the soil´s potential acidity (H+Al) through the application of phonolite may be due to calcium carbonate and magnesium in the phonolite (MARTINS et al., 2015b). These authors showed increase in pH in the soil when rock powder was applied. According to Lapido-Loureiro, Melamed and Figueiredo Neto (2009), increase of pH in the soil is one of the assets of rock powder, besides low costs and slow release in the soil, avoiding losses by leaching. Increase in the soil´s pH occurred mainly in treatments with inoculation, which may be associated to a greater release of carbonates from the rock triggered by bacterial strains.
Lapido-Loureiro and Nascimento (2009) underscore that, due to the mineral characteristics of the rocks, the K + in the rock powder has low solubility, or rather, slower than by conventional fertilizers even though they provide other macro and micro nutrients in their composition. Consequently, they have favorable traits for pH rise in the soil.
The use of phonolite plus inoculation of diazotrophic strains contributed towards a greater availability of potassium in the soil. Among the tested strains, strains UNIFENAS 100-01, UNIFENAS 100-16, UNIFENAS 100-27, UNIFENAS 100-39 and UNIFENAS 100-93 may be underscored. They were more efficient in the solubilization of the phonolites K + .
In spite of these results, K rates released by bacterial strains in the soil were low when compared to those in in vitro conditions. Further studies on bio-solubility in the soil are required to select the most efficient strains for the process.

ACKNOWLEDGMENTS
The author thank the Coordination for the Improvement of Higher Education Personnel (CAPES) for the PROSUP scholarship.