Technological Quality of Sugarcane Inoculated with Plant-Growth-Promoting Bacteria and Residual Effect of Phosphorus Rates

Phosphate fertilization in highly weathered soils has been a major challenge for sugarcane production. The objective of this work was to evaluate the foliar levels of phosphorus (P) and nitrogen (N) and the technological quality and productivity of second ratoon cane as a function of inoculation with plant-growth-promoting bacteria (PGPBs) together with the residual effect of phosphate fertilization. The experiment was carried out at the research and extension farm of Ilha Solteira, state of São Paulo, Brazil. The experiment was designed in a randomized block with three replications in a 5 × 8 factorial scheme. The treatments consisted of five residual doses of phosphorus (0, 45, 90, 135 and 180 kg ha−1 of P2O5, 46% P) applied at planting from the source of triple superphosphate and eight inoculations from three species of PGPB (Azospirillum brasilense, Bacillus subtilis and Pseudomonas fluorescens), applied in single or co-inoculation at the base of stems of sugarcane variety RB92579. Inoculation with PGPBs influenced leaf N concentration, while inoculations with Pseudomonas fluorescens and combinations of bacteria together with the highest doses exerted a positive effect on leaf P concentration. Co-inoculation with A. brasilense + Pseudomonas fluorescens associated with a residual dose of 135 kg ha−1 of P2O5 increased stem productivity by 42%. Thus, it was concluded that inoculations with Pseudomonas fluorescens and their combinations are beneficial for the sugarcane crop, reducing phosphate fertilization and increasing productivity.


Introduction
Brazil is the world's largest producer of sugarcane (Saccharum officinarum), with a total production of 654.5 million tons in the 2020/2021 harvest. The state of São Paulo the production of organic acids and solubilization of insoluble inorganic phosphate compounds such as tricalcium phosphate, dicalcium phosphate, hydroxyapatite and rock phosphate [17,20], and their phosphatase activity to mineralize organic phosphorus [15], thus improving the residual effect of phosphorus fertilization.
It is believed that the interactions between the species of PGPBs of the genera Azospirillum, Bacillus and Pseudomonas have the capacity to increase the efficiency of fertilization with residual P in rat cane, with positive effects on the productivity and technological quality of the sugarcane crop [21]. In this context, the present objective was to evaluate the effect of inoculation with three species of plant-growth-promoting bacteria, applied alone or in combination with residual doses of P 2 O 5 (applied at the time of planting) on foliar nitrogen (N) and phosphorus (P) concentrations, yield of stalks, sugar and technological quality of cane second ratoon in tropical soil with low P content.

Leaf Concentrations of N and P as a Function of Inoculations and P 2 O 5 Doses
The effects of interactions between inoculations and phosphorus (P) doses, as well as the effect of P 2 O 5 doses, were not significant on leaf N concentrations of sugarcane; however, the effect of inoculations was significant on leaf N concentration ( Table 1). The treatments with inoculation of PGPBs provided higher leaf N concentration in sugarcane compared to those without inoculation treatments. Means followed by the same letter in the column were statistically not different according to the Scott-Knott test at 5% probability. TRS: total recoverable sugar, STY: stalk yield and SUY: sugar yield. CV: coefficient of variance. **, * and ns: significant at 1% and 5% at p < 0.01, p < 0.05 and non-significant, respectively. n = 4 replications.
There was a significant interaction (p < 0.05) between inoculations with PGPBs and P 2 O 5 doses for leaf P concentration of sugarcane (Table 1). Leaf P concentration of sugarcane plants was increased with inoculations of Azospirillum brasilense, Pseudomonas fluorescens, Azospirillum brasilense + Bacillus subtilis, Azospirillum brasilense + Pseudomonas fluorescens and Bacillus subtilis + Pseudomonas fluorescens in association with increasing residual P 2 O 5 doses (Figure 1). Inoculations with A. brasilense, P. fluorescens, A. brasilense + B. subtilis, A. brasilense + P. fluorescens, B. subtilis + P. fluorescens and A. brasilense + B. subtilis + P. fluorescens at a residual P 2 O 5 dose of 180 kg ha −1 were observed with higher leaf P concentration, which were statistically not different from the leaf P concentration obtained in the residual P 2 O 5 dose of 135 kg ha −1 in combination with inoculations of P. fluorescens, A. brasilense + B. subtilis, A. brasilense + P. fluorescens and B. subtilis + P. fluorescens (Figure 1).
Means followed by the same le er in the column were statistically not different according to the Sco -Kno test at 5% probability. TRS: total recoverable sugar, STY: stalk yield and SUY: sugar yield. CV: coefficient of variance. **, * and ns: significant at 1% and 5% at p < 0.01, p < 0.05 and nonsignificant, respectively. n = 4 replications.
There was a significant interaction (p < 0.05) between inoculations with PGPBs and P2O5 doses for leaf P concentration of sugarcane (Table 1).
Leaf P concentration of sugarcane plants was increased with inoculations of Azospirillum brasilense, Pseudomonas fluorescens, Azospirillum brasilense + Bacillus subtilis, Azospirillum brasilense + Pseudomonas fluorescens and Bacillus subtilis + Pseudomonas fluorescens in association with increasing residual P2O5 doses (Figure 1). Inoculations with A. brasilense, P. fluorescens, A. brasilense + B.subtilis, A. brasilense + P. fluorescens, B. subtilis + P. fluorescens and A. brasilense + B. subtilis + P. fluorescens at a residual P2O5 dose of 180 kg ha −1 were observed with higher leaf P concentration, which were statistically not different from the leaf P concentration obtained in the residual P2O5 dose of 135 kg ha −1 in combination with inoculations of P. fluorescens, A. brasilense + B. subtilis, A. brasilense + P. fluorescens and B. subtilis + P. fluorescens (Figure 1).

Technological Attributes of Sugarcane as a Function of Inoculations and P 2 O 5 Doses
Interactions between inoculations with PGPBs and phosphorus (P) doses were significant for % fiber, juice brix, pol, total recoverable sugar (ATR), stem yield (STY) and sugar yield (SUY) of cane sugar. The effect of the single treatment of inoculations and the residual dose of P 2 O 5 and their interactions were not significant for the purity of the sugarcane juice ( Table 2). Means followed by the same letter in the column were statistically not different according to the Scott-Knott test at 5% probability. TRS: total recoverable sugar, STY: stalk yield and SUY: sugar yield. CV: coefficient of variance. **, * and ns: significant at 1% and 5% at p < 0.01, p < 0.05 and non-significant, respectively. n = 4 replications.
The highest fiber % was observed in the treatments with inoculations of A. brasilense, P. fluorescens, A. brasilense + B. subtilis and A. brasilense + B. subtilis + P. fluorescens without residual P fertilization. In addition, inoculations with P. fluorescens, A. brasilense + P. fluorescens, B. subtilis + P. fluorescens and A. brasilense + B. subtilis + P. fluorescens at a residual P 2 O 5 dose of 45 kg ha −1 obtained the highest fiber % while the triple co-inoculation provided higher fiber % only at residual P 2 O 5 dose of 90 kg ha −1 . The treatments without inoculation and with co-inoculations of A. brasilense + P. fluorescens and A. brasilense + Bacillus subtilis + Pseudomonas fluorescens increased fiber % at a residual P 2 O 5 dose of 135 kg ha −1 , whereas inoculations with A. brasilense, B. subtilis, P. fluorescens, A. brasilense + P. fluorescens and A. brasilense + B. subtilis + P. fluorescens provided higher fiber % at a residual P 2 O 5 dose of 180 kg ha −1 (Figure 2A).
Inoculation with Azospirillum brasilense + Bacillus subtilis reduced sugarcane fiber % with increasing residual P 2 O 5 doses, indicating that increasing P fertilization doses collaborates effectively with the performance of these bacteria to reduce fiber %, which is undesirable for juice extraction. The inverse was observed with co-inoculation of A. brasilense + P. fluorescens, where fiber % of sugarcane was increased with increasing residual P doses, which is harmful to the quality of sugarcane ( Figure 2A).
Plants 2023, 11, x FOR PEER REVIEW 6 fluorescens and A. brasilense + B. subtilis + P. fluorescens provided higher fiber % at a res P2O5 dose of 180 kg ha −1 (Figure 2A).  Inoculation with Azospirillum brasilense + Bacillus subtilis reduced sugarcane fib with increasing residual P2O5 doses, indicating that increasing P fertilization d The percentage of apparent soluble solids (% brix) increased with increasing doses of residual P 2 O 5 under co-inoculation of B. subtilis + P. fluorescens ( Figure 2B). Inoculation with B. subtilis at the residual P 2 O 5 dose of 45 kg ha −1 was observed with highest % of brix. The lowest % brix was determined in the treatments with inoculation of P. fluorescens and with co-inoculation of A. brasilense + P. fluorescens and A. brasilense + B. subtilis + Pseudomonas fluorescens at a residual P 2 O 5 dose of 90 kg ha −1 . Inoculations of A. brasilense, B. subtilis, P. fluorescens and A. brasilense + P. fluorescens associated with a residual P 2 O 5 dose of 180 kg ha −1 provided lower % brix ( Figure 2B).
There was an increase in pol percentage of sugarcane (% of pol) with increasing residual P 2 O 5 doses up to 75 kg ha −1 in combination with inoculation of Bacillus subtilis, while a further increase in residual P 2 O 5 doses led to the reduction in pol % in sugarcane. In addition, pol% of sugarcane was also reduced with increasing residual P 2 O 5 doses under inoculations with Pseudomonas fluorescens and A. brasilense + B. subtilis ( Figure 2C).

Total Recoverable Sugar, Stalks and Sugar Yield as a Function of Inoculations and P Doses
The concentration of total recoverable sugar (TRS) was reduced in the treatments with inoculation of P. fluorescens and A. brasilense + B. subtilis + P. fluorescens as compared to other inoculations and co-inoculations at a residual P 2 O 5 dose of 90 kg ha −1 . The same effect was observed with inoculations of A. brasilense, B. subtilis, P. fluorescens and A. brasilense + P. fluorescens at a residual P 2 O 5 dose of 180 kg ha −1 . In addition, TRS concentration was increased with inoculation of B. subtilis under the increasing residual dose of P 2 O 5 up to 72 kg ha −1 ( Figure 3A).
The treatments inoculated with A. brasilense and B. subtilis were observed with greater stalk productivity or stalk yield (STY) ha −1 and sugar yield (SUY) ha− 1 in the absence of P 2 O 5 fertilization in relation to other inoculations. Stalk yield and SUY were increased with inoculations of A. brasilense, B. subtilis, P. fluorescens and A. brasilense + B. subtilis under the residual P 2 O 5 dose of 45 kg ha −1 , while the same effect was observed with inoculation of P. fluorescens at the residual P 2 O 5 dose of 90 kg ha −1 , and with co-inoculation of A. brasilense + P. fluorescens at a P 2 O 5 dose of 135 kg ha −1 (Figure 3A, C). The residual P 2 O 5 dose of 180 kg ha −1 under inoculations of A. brasilense, P. fluorescens, A. brasilense + B. subtilis and B. subtilis + P. fluorescens were observed with greater STY as compared with other treatments at the same P 2 O 5 dose. Sugarcane stalk yield was also increased in the treatments with inoculation of P. fluorescens and without inoculation in association with increasing residual P 2 O 5 doses up to 102 and 114 kg ha −1 , respectively. The treatments without inoculation and inoculation with P. fluorescens could produce a maximum STY of 110.5 and 142.8 t stalks ha −1 at calculated residual P 2 O 5 dose of 102 and 114 kg ha −1 . In addition, there was a reduction in STY in the treatments with inoculation of B. subtilis and increasing residual P 2 O 5 doses ( Figure 3B).
Sugar yield (SUY) of sugarcane was increased without inoculation and with inoculation of P. fluorescens and A. brasilense + P. fluorescens under maximum residual P 2 O 5 doses of 116, 98 and 110 kg ha −1 , giving a maximum estimated SUY of 15, 16.7 and 16 t ha −1 , respectively. Sugar yield was increased with increasing residual P 2 O 5 doses under co-inoculation of B. subtilis + P. fluorescens and A. brasilense + B. subtilis + P. fluorescens. In addition, there was a reduction in SUY with increasing P doses and inoculations with A. brasilense and B. subtilis ( Figure 3C).

Discussion
Nitrogen and phosphorus are considered the most important nutrients for the growth and yield of crop plants. Leaf concentrations of N and P in the range of 18-26 and 1.5-3.0 g kg −1 , respectively, are considered adequate for sugarcane ratoon leaves [22,23]. Leaf N concentrations in the present study are within the established standard range, except for treatments without inoculation (17.35 g kg −1 ) and with inoculation of A. brasilense (3.20 g ha −1 ). Leaf P concentration in treatments with inoculations of A. brasilense and A. brasilense + B. subtilis under residual P 2 O 5 dose of 180 kg ha −1 was increased by 31 and 28% as compared to without inoculation. This may be due to the ability of PGPBs to improve plant growth directly and indirectly through various mechanisms [11], including nitrogen fixation and phosphate solubilization. This may be because of greater root growth (increasing soil exploitation by roots) and biological N fixation by A. brasilense [24] and increasing P availability and accumulation with inoculations of Bacillus subtilis and P. fluorescens [21]. The mechanism of P solubilization has been associated with the release of organic acids, which chelate cations of phosphate through hydroxyl and carboxyl groups and then convert them into soluble forms [25,26] with greater availability to the roots, thus increasing its uptake by plant roots and transport to the leaves, resulting in higher leaf P concentrations. It has also been reported that by-products enriched with phosphate rock and inoculated with P-solubilizing bacteria (P. aeruginosa, Bacillus sp. and Rhizobium sp.) in soils cultivated with sugarcane [27]. Rosa et al. [28] reported that inoculation and coinoculation with A. brasilense, B. subtilis and P. fluorescens increased foliar P concentrations in the first sugarcane harvest.
The % of sugarcane fiber was increased by 11.65% (Table 2; Figure 2A); this is within the agro-industrial quality of sugarcane within the standard range of 11-13% found in a previous study [29]. It is interesting and worth mentioning that % fiber values could not exceed the maximum value of 13%, as the higher % fiber content would negatively interfere at the time of juice extraction. Renan et al. [10] showed that the percentage of sugarcane fiber reflects on the juice extraction efficiency, where higher values of % fiber reduce the extraction efficiency. Furthermore, it should be considered that sugarcane varieties with low fiber content are more susceptible to mechanical damage during cutting and transport, thus impairing % brix, Pol and SUY.
Sugarcane requires low % brix, and minimum % brix values produced 18% higher sugar yield regardless of residual P doses and inoculations ( Table 2). The % brix is directly related to sugar yield and the ideal % varies from 18 to 25% [30]. Inoculation with B. subtilis at a residual P 2 O 5 dose of 45 kg ha −1 increased brix by 0.8% as compared to without inoculation with the highest value for the present work. This demonstrated the ability of these PGPBs in associate with residual P doses to improve quality of sugarcane juice ( Figure 2B). It has previously reported that inoculation and co-inoculation with A. brasilense, P. fluorescens and B. subtilis under reduced P 2 O 5 dose improved brix% in the first crop of sugarcane, proved that P solubilization favors the production of sugar [28].
The sugarcane pol% indicates all the apparent sucrose of the absolute juice of the harvested cane. Thus, increasing pol values could improve industrial yield of sugarcane [31]. The present study indicated that pol % of sugarcane was reduced in the treatments with increasing residual P 2 O 5 doses under inoculation of P. fluorescens and co-inoculation of A. brasilense + B. subtilis, indicating that increase in P fertilization can reduce efficiency of these bacteria in the industrial sugarcane production. The co-inoculation with A. brasilense + B. subtilis + P. fluorescens provided the highest pol% of sugarcane in comparison to other inoculations, being at least 1.4% higher than the other inoculations ( Figure 2C).
Inoculation with B. subtilis and A. brasilense + B. subtilis at the residual P 2 O 5 dose of 45 kg ha −1 provided the highest total recoverable sugar (TRS) in relation to the other P doses within the same inoculation ( Figure 3A), which allows us to infer that these bacteria benefited from the low P 2 O 5 dose. Phosphate fertilizer allowing the release of compounds that could help to improve industrial quality of sugarcane juice of this variety. Rosa et al. [28] describes that inoculation and co-inoculation with A. brasilense, B. subtilis and P. fluorescens provided higher TRS of sugarcane juice under reduced P doses.
At a residual P 2 O 5 dose of 45 kg ha −1 , the stalk yield (STY) (t ha −1 ) was increased with inoculation of A. brasilense, B. subtilis, P. fluorescens and A. brasilense + B. subtilis + P. fluorescens, being 28, 21, 18 and 16% higher in relation to without inoculation, respectively ( Figure 3B). Inoculation with Pseudomonas fluorescens increased the STY of sugarcane by 17% (17.96 t ha −1 ) as compared to without inoculation at a residual P 2 O 5 dose of 90 kg ha −1 . In addition, co-inoculation with A. brasilense + P. fluorescens at residual P 2 O 5 dose of 135 kg ha −1 provided 42% (36.69 t ha −1 ) higher STY than without inoculation, being the highest STY observed in the entire study, also allowing a 25% reduction in phosphate fertilization. It was reported in a previous study that ratoon cane productivity was significantly improved with joint application of PGPBs, green manure and adequate fertilizers [27].
The P use efficiency for the treatments with inoculation of A. brasilense, B. subtilis and P. fluorescens at a residual P 2 O 5 dose of 45 kg ha −1 , were 0.76, 0.62 and 0.58 t of sugarcane kg −1 P 2 O 5 , respectively. For inoculation with Pseudomonas fluorescens at a residual P 2 O 5 dose of 90 kg ha −1 , the P use efficiency was 0.54 t of sugarcane kg −1 P 2 O 5 . The P use efficiency for the co-inoculation with A. brasilense + P. fluorescens at residual P 2 O 5 dose of 135 kg ha −1 was 0.38 t of sugarcane kg −1 P 2 O 5 .
Sugar yield (SUY) (t ha −1 ) was increased with inoculation of A. brasilense, B. subtilis, P. fluorescens and A. brasilense + B. subtilis + P. fluorescens at a residual P 2 O 5 dose of 45 kg ha −1 , being 26, 23, 14 and 17% higher as compared to without inoculation treatments ( Figure 3C). Inoculation with P. fluorescens and co-inoculation with A. brasilense + P. fluorescens at a residual P 2 O 5 dose of 90 kg ha −1 provided 8 and 43% higher SUY, respectively, in comparison of without inoculated treatments (Table 2; Figure 3C). The previous study with the first sugarcane crop indicated that co-inoculation of A. brasilense + B. subtilis reduced P 2 O 5 fertilization by 75% during first sugarcane crop while inoculation with A. brasilense improved agro-industrial quality of sugarcane under P 2 O 5 dose of 45 kg ha −1 [28].
The exact mechanisms of inoculation with A. brasilense and B. subtilis on the productivity and industrial quality of sugarcane is still needed to be disclosed. However, use of these microorganisms increased availability and absorption of P, which is mainly responsible for high leaf P concentration, growth, industrial quality and yield of sugarcane crops [32][33][34]. The gene sequencing of A. brasilense (Ab-V5 and Ab-V6 strains) highlighted its role in auxin synthesis [34], nutrient availability and cycling [14,35] and biological N fixation [36,37]. Furthermore, B. subtilis has been described as having the potential ability to promote plant growth, solubilize P, and inhibit infestation of phyto-pathogenic attack and heavy metal accumulation [38,39]. Bacillus sp. is a genus that is characterized by being an excellent phosphate solubilizer, including B. subtilis, and P is considered one of the most critical nutrients to increase crop yield [40,41]. Studies have described that plants adapt different mechanisms under inoculation with A. brasilense and B. subtilis to potentially increase soil P availability and food security [39,42]. Bacteria of the genera Arthrobacter, Bacillus, Pseudomonas and Rhizobium, among others, have demonstrated the ability to solubilize phosphates [17]. Therefore, these PGPBs have the potential to make better use of phosphorus present in the solid phase of the soil, constituting viable alternatives for the inoculation of crops. However, it is noteworthy that the genera Bacillus, Pseudomonas and Rhizobium are considered the most efficient solubilizers of inorganic phosphates [15].
Single inoculation with P. fluorescens and/or co-inoculation with A. brasilense and B. subtilis increased leaf P concentration, industrial quality and productivity of sugarcane in the current study. It may be possible due to the efficiency of P. fluorescens bacterium to use metabolites as a bio-control agent with the synthesis of antibiotics and volatile organic compounds to combat soil pathogens [43,44]. This is promising in terms of phosphate solubilization [45] and N-fixing activities [46]. The genus Pseudomonas sp. produces gluconic acid, which is among the different types of organic acids involved in the solubilization of phosphate, which is considered as an important technique to improve management of P fertilization in modern agriculture [47]. Therefore, co-inoculation with A. brasilense + P. fluorescens can be an important tool in the sustainable management of phosphate fertilization in sugarcane cultivated in tropical soils, optimizing the use of residual P 2 O 5 doses and allowing to reduce P 2 O 5 fertilization by 25% in the ratoon cane field.

The Location of the Experimental Area
This experiment was conducted during the 2019/20 cropping season at Limoeiro Farm, a part of the ethanol and sugar industry in Ilha Solteira, in the northwest of the state of São Paulo, Brazil. The experimental area is located at the geographical coordinates of 20 • 21 14 S latitude and 51 • 04 51 W longitude, with an altitude of 371 m. The experiment was conducted during the third year of sugar-cane cultivation (second ratoon). The climate in the region is Aw (Köppen scale, defined as tropical with dry winters). Climatic data referring to the period of the experiment were recorded (Figure 4). gluconic acid, which is among the different types of organic acids involved in the solubilization of phosphate, which is considered as an important technique to improve management of P fertilization in modern agriculture [47]. Therefore, co-inoculation with A. brasilense + P. fluorescens can be an important tool in the sustainable management of phosphate fertilization in sugarcane cultivated in tropical soils, optimizing the use of residual P2O5 doses and allowing to reduce P2O5 fertilization by 25% in the ratoon cane field.

The Location of the Experimental Area
This experiment was conducted during the 2019/20 cropping season at Limoeiro Farm, a part of the ethanol and sugar industry in Ilha Solteira, in the northwest of the state of São Paulo, Brazil. The experimental area is located at the geographical coordinates of 20° 21′ 14″ S latitude and 51° 04′ 51″ W longitude, with an altitude of 371 m. The experiment was conducted during the third year of sugar-cane cultivation (second ratoon). The climate in the region is Aw (Köppen scale, defined as tropical with dry winters). Climatic data referring to the period of the experiment were recorded (Figure 4). The cultivation history of the area in the last ten years and prior to the implantation of sugarcane (2017) showed that this area was cultivated with a pasture (Urochloa brizantha) with some degree of degradation. The soil was classified as Rhodic Haplustox according to USDA [48], and as medium-to-sandy-textured Red Dystrophic soil according to SiBCS [49], with a granulometry of 777, 98, 125 g kg −1 and 747, 88, 165 g kg −1 of sand, silt and clay at depths of 0.00-0.25 and 0.25-0.50 m, respectively. The chemical a ributes of the soil were determined before implantation of sugarcane and regrowth of second ratoon cane crop (Table 3).  The cultivation history of the area in the last ten years and prior to the implantation of sugarcane (2017) showed that this area was cultivated with a pasture (Urochloa brizantha) with some degree of degradation. The soil was classified as Rhodic Haplustox according to USDA [48], and as medium-to-sandy-textured Red Dystrophic soil according to SiBCS [49], with a granulometry of 777, 98, 125 g kg −1 and 747, 88, 165 g kg −1 of sand, silt and clay at depths of 0.00-0.25 and 0.25-0.50 m, respectively. The chemical attributes of the soil were determined before implantation of sugarcane and regrowth of second ratoon cane crop (Table 3). Table 3. Chemical characterization of the soil in the experimental area before planting sugarcane and regrowth of the second ratoon cane in the 0.00-0. 25

Installation and Conduct of the Experiment
The experimental field was applied with soil 1.0 t ha −1 of dolomitic limestone (PRNT 85%) to obtain 60% base saturation as well as 1.0 t ha −1 of gypsum to raise sulfur content (S) after initial soil chemical characterization and 15 days before planting, as recommended for sugarcane cultivation. Three harrowing and one sub-soiling operations were carried out, followed by furrowing at a depth of 0.40 m and applying the insecticide fipronil (180 g ha −1 of ai) + fungicide pyraclostrobin (125 g ha −1 of ai) in planting furrow.
Manual plantation of a sugarcane variety RB92579 was carried out in planting furrows on 11 July 2017, containing about 22 buds m −1 . All the treatments were equally fertilized with 30 kg ha −1 of N (ammonium nitrate, 33% N) and 120 kg ha −1 of K 2 O (potassium chloride, 60% K 2 O) based on soil analysis. In addition, respective doses of P 2 O 5 were applied in a single dose for each treatment in the planting furrow at the time of sugarcane implantation in 2017.
Fertilization of topdressing in the second ratoon was performed 30 days after sprouting (DAS) on 12 September 2019 from ammonium nitrate as a source of N and potassium chloride as a source of K 2 O based on the recommendations of [23], as well as land slope and precipitation. The inoculation with PGPBs was performed at 30 DAS of the second ratoon sugarcane crop, using the following liquid inoculants doses: 1.0 L ha −1 of A. brasilense (strains Ab-V5 and Ab-V6 (guarantee of 2 × 10 8 colony forming units (CFU) mL −1 )), 0.5 L ha −1 of B. subtilis (strains CCTB04 (guarantee of 1 × 10 8 CFU mL −1 )) and 0.5 L ha −1 of P. fluorescens (strains CCTB03 (guarantee of 2 × 10 8 CFU mL −1 )) on the basis of manufacturer's recommendations. The spray volume was 200 L ha −1 (for all applied inoculants), applied with a backpack sprayer at the base of the second ratoon cane tillers during early morning due to milder temperature.
The correction of boron and zinc deficiency was carried out by applying 5 kg ha −1 of zinc (zinc sulfate, 20% Zn and 10% S) and 2 kg ha −1 of boron (boric acid, 17% B) in coverage at about 15 cm from the base of the plants at 60 DAS.
Weeds, pests and diseases were controlled via chemicals and biological ways according to the crop needs throughout the growth cycle. The Insecticides Chlorantraniliprole (10 g ha −1 ai) + Lambda cyhalothrin (5 g ha −1 ai) were applied. Four releases of Trichogramma galloi were performed throughout the crop cycle at 142, 148, 213 and 220 DAS to biologically control sugarcane borer (Diatraea saccharalis). The ratoon sugarcane was manually harvested on the 5th August 2020, at 357 DAS.

Assessments
The middle third of 12 flag leaves per plot were collected at the full vegetative growth phase (157 DAS) in the morning, and the midrib was removed as described by [50] for the determination of leaf N and P concentrations [51]. The weight was quantified at the time of harvest in order to estimate stalk yield per hectare (STY) (t ha −1 ).
Ten commercial-quality sugarcane stalks were collected from each plot at the time of harvest and determined for technological quality at Technological Analysis Laboratory by following the methodology defined in the System Payment of Sugarcane Based on Sucrose Content (SPSBSC), according to [52]. Fiber (%), juice purity (%), soluble solids or juice brix content (brix %), apparent sucrose content of cane (Pol-%) and total recoverable sugar (TRS) (kg sugar per t cane −1 ) were also evaluated at the technological laboratory of ethanol and sugar industry at Suzanápolis, Brazil. Sugar yield SUY (t sugar ha −1 ) was quantified from the product of STY and pol %, divided by 100 in each individual plot.

Statistical Analysis
The results were submitted to analysis of variance (F test) and Scott-Knott tests (p ≤ 0.05) to compare inoculation means, and regression equations were adjusted for the residual effect of P 2 O 5 doses. Statistical analyses were processed using the computer program SISVAR [53], and graphs were plotted with the aid of SigmaPlot 12.5 software.

Conclusions
Inoculation with Pseudomonas fluorescens alone, and the co-inoculations Azospirillum brasilense + B. subtilis, Azospirillum brasilense + Pseudomonas fluorescens and Bacillus subtilis + Pseudomonas fluorescens under residual P 2 O 5 doses of 135 and 180 kg ha −1 increased leaf P concentration. In addition, the isolated or combined inoculation with Azospirullum brasilense, Pseudomonas fluorescens and Bacillus subtilis promoted leaf N concentration in relation to control.
The co-inoculation of Azospirillum brasilense + Pseudomonas fluorescens in association with a residual P 2 O 5 dose of 45 kg ha −1 improved agro-industrial quality of the second ratoon cane. Co-inoculation with Azospirillum brasilense + Pseudomonas fluorescens at a residual P 2 O 5 dose of 135 kg ha −1 increased stalk yield by 42% and sugar by 43%, providing a 25% reduction in phosphate fertilization in the second ratoon cane.