In Brazil, carrot is among the five main garden crops grown and 80 % of the total production supplies Brazilian domestic market. The Southeast, Northeast, and South regions are the largest producers of this root crop. The State of Minas Gerais stands out in the production of this vegetable crop and the Alto Paranaiba region is one of the main producers. Yield potential of the crop is from 100 to 120 t ha^-1; however, the Brazilian average is much lower around 33 tons ha^-1 (Embrapa, 2010).

Free-living bacteria are found in the rhizosphere of plants  and  parts  of  them  are  known  as  Plant  Growth Promoting Rhizobacteria - PGPR (Alves, 2007). Pseudomonas, Bacillus, Azospirillum, Agrobacterium and Azotobacte have been reported as PGPR, being Pseudomonas and Bacillus the PGPR widely reported.

The mechanisms that PGPR, including Bacillus spp., increase the plant growth may vary from production or changes on phytohormones concentration, ethylene synthesis inhibition, siderophore and antibiotics production, nitrogen fixation, phosphate solubilization and systemic resistance induction to pathogens (Singh et al., 2011;  Vafadar   et   al.,   2014).  The  beneficial  effect  of rhizobacteria in commercially cultivated crops is world-wide known in onion (Harthmann et al., 2010), potato (Sottero et al., 2006) and tomato Mena-Violante and Olalde-Portugal, 2007). However, this potential is not yet known in several crops, such as carrot (Konusny-Andreani et al., 2014).

Carrot marketing depends on the roots pattern that considers the length, diameter and the defects absence (Ceasaminas, 2015). Thus, the application of PGPR may increase the production of commercial roots, resulting in lower waste and consequently higher profitability to farmers. Bacillus spp. in special is easily grown in liquid culture media with low cost, which make easier its mass production in industrial fermenters. Furthermore, they produce resistant endospores, which increase the shelf life of products and integrated use with chemical products (Lanna Filho, 2010). Therefore, the aim of this study was to evaluate the efficiency of five Bacillus spp. isolates in increasing the production of commercial roots of carrot under field conditions.

PGPR Isolates

The following isolates: SF 264 (Bacillus spp.), SF 268 (Bacillus spp.), SF 202 (Bacillus subtilis - commercial product Rizos^®), SF 266 (B. methylotrophicus - commercial product Quartz^®) and SF 267 (Bacillus methylotrophicus - commercial product Onix^®). They were tested in 4 fields with commercial carrot production in the municipality of Rio Paranaíba - MG (site 1 – S.: 19°19´09´´; W.: 46°14´04.2’’; site 2 – S.: 19°14´14.7´´; W.: 46°08´47.5.2’’; site 3 – S.: 19°25´45.8´´; W.: 46°14´10.2’’; site 4 – S.: 19°13´04´´; W.: 46°14´20’’). Isolates and commercial products belong to Farroupilha`s group from Patos de Minas, Minas Gerais. The experiments were carried out during February and March 2012.

Carrot cultivars

The cultivar Brasilia, developed by EMBRAPA, was used in two sites. In the two other sites the cultivars were Suprema Max^® (Isla Sementes Ltda, Porto Alegre, Brasil) and Juliana^® (Seminis Vegetable Seeds, St. Louis, EUA), respectively.

Soil and sites characterization

The chemical and physical characteristics of the soils are presented in the Table 1.

All isolates were prepared in liquid culture media and formulated as a protocol used in the company. The concentration of the isolates SF 264, SF 266, SF 267 and SF 268 was 1 x 10⁹ colony forming units (cfu) per mL, while the concentration of SF 202 was 5 x 109 cfu mL^-1.

At 15 days after planting, the isolate suspensions were applied on the plants using a CO₂ -pressurized backpack sprayer furnished with three 110-02 nozzles working in constant pressure of 2.2 bar, spaced at 0.5 m. The dose and volume of the spray solution were 4 L ha^-1 and 200 L ha^-1, respectively. Immediately after the application of the bacterial isolates, each experimental area was irrigated at 5 mm  water   depth   in   order  to  facilitate  the  bacteria  propagules percolation in the soil profile. The sprays were done after 3 pm.

Each plot was 6 m long and 1.75 m wide (10.5 m²), including four double rows. An additional treatment containing only water composed the control treatment. The experiment was designed as randomized blocks. Crop management was carried out following the pattern adopted by the growers, including irrigation, topdressing fertilizations, pest management and rough-hewing.

Yield evaluation

The harvest was carried out at 105 days after planting. The fresh weight of commercial roots of carrot was evaluated in 1 m², located at the center of each plot (useful plot). All roots in the useful plots were classified according to the carrot standard classification of a Brazilian program for commercial standards and horticultural packaging improving (Ceasaminas, 2015). It was considered only the roots with a length between 14 and 26 cm without defects. Later the selected roots were weighed using a digital scale.

Statistical analysis

The data were submitted to variance analysis (F test, P = 0.05) and the means compared by Scott-Knott test (P = 0.05) using the software SISVAR 5.1 Build 72 (Ferreira, 2007).

The bacteria isolates increased the production of commercial roots of the cultivar Juliana in both sites (Table 2). For the cultivar Suprema Max, the SF 267, SF 266, SF 264 and SF 202, even in different groups, increased the carrot production. For the cultivar Juliana, yields with SF 202, SF 266 and SF 267 application were 10.40; 12.48 and 23.32 higher than the control.

Isolates of B. subtilis also has the ability to induce hormonal regulation in plants, as reported by Tsavkelova et al. (2006), Persello-Cartieaux et al. (2003) and Lanna Filho  (2010)  controlling  the  root  growth by auxin, gibberellin and cytokinin synthesis. It may also explain the responses of B. subtilis isolates in all experiments. Besides the genetic differences among isolates and their possible influence on root colonization, the interaction between isolates and carrots cultivar may also influenced the plant growth (Romeiro et al., 2005; Choudhary et al., 2007).

For the cultivar Brasilia, all isolates increased the carrot production. The SF 264 isolate in the cultivar Juliana and the SF 268 in the cultivar Suprema Max and Juliana did not have the same beneficial effect. It is possible that the production and root exudates composition among cultivars are different; therefore, the SF 268 and SF 264 isolates have not been able to colonize efficiently the rhizosphere of Suprema Max^®and Juliana^®cultivars, and consequently act as growth promoter.

When a microorganism is introduced into a new environment, it must overcome the competition with other microorganisms (microbiostase), in order to develop and perform its ecological functions (Tsavkelova et al., 2006).There are reports that show the high specificity of growth - promoting bacteria to colonize specific hosts. This specificity was not observed in the present study. Raasch et al. (2013) stated that eucalyptus production with Rizolyptus® (B. subtilis) varied according to the tested clones, suggesting a high specificity of root colonization regarding the host.

Mello et al. (2002) reported the lack of specificity of the C210, ENF10, ENF16 and RAB9 isolates, obtained from cabbage, beans and radish, respectively, regarding growth promotion in seedlings of pineapple micro-propagation. In another study, isolates of C116 (Bacillus pumilus) and C25 (Bacillus thuringiensis subvar. kenyae), from cabbage were effective in promoting the growth of lettuce seedlings (Gomes et al., 2003).

Besides the specificity of bacterial isolates to the host, stands out the differences in the soil of rooting for rhizobacteria colonization, once abiotic factors (pH, nutrient availability, moisture retention, aeration etc.) and biotic (qualitative and quantitative composition of the microbiota and others) may favor or not the colonization, survival and beneficial activity of rhizobacteria (Mafia et al., 2009).

Such variables may explain the difference in the production of commercial carrot when using a same bacterial isolate and in different experiments, i.e. under different environmental conditions. This was observed for the cultivar Brasilia - site 1, where the SF 267 isolate (B. methylotrophicus, Onix^®) was used, the fresh weight was 6.41 kg while in the site 2 it was 8.51 kg.

In the cultivation fields of the cultivar Suprema Max and Juliana, the potassium content in the soil was around 40% lower than the other fields as highlighted in the Table 1. There may be some relationship between potassium and phosphorus contents in the soil and the capacity of the SF 268 isolate colonize the rhizosphere of carrot plants and promote the plant growth. Moreover, the SF 267, SF  266  and  SF  202  isolates do  not  seem  to depend on the presence of high amounts of these nutrients in the soil, considering that some PGPR solubilize phosphates and/or provide potassium to the plants (Gomes et al., 2003; Viruel et al., 2014). Further studies are recommended to investigate whether the beneficial effect of the SF 267, SF 266 and SF 202 isolates in increasing the production of commercial carrot roots is due to their action as nutrients solubilizers in the soil.

The application of Bacillus spp., present in the products Rizos®, Quartz® and Onix®, enhanced the yield of commercial carrot roots.

The authors have not declared any conflict of interest.

À Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), à Coordenação de Aperfeiçoamento de Pessoal do Nível Superior (CAPES) e ao Laboratório de Bio Controle Farroupilha S.A. pelo apoio financeiro; e ao Grupo VAgropesg, Grupo Leópolis, Sekita Agronegócios e Fazenda Shimada pela concessão das áreas experimentais.