Organomineral Fertilizer as Source of P and K for Sugarcane

Sugarcane (Saccharum spp) crop has high social, economic and environmental importance for several regions throughout the world. However, the increasing demand for efficiency and optimization of agricultural resources generates uncertainties regarding high mineral fertilizer consumption. Thereby, organomineral fertilizers are to reduce the conventional sources consumption. Thus, this study was carried out to evaluate the agronomic and economic sugarcane performancies and the residual effect of P and K under mineral and organomineral fertilization. Growth and technological parameters, leaf and soil nutrients concentration in surface and subsurface layers were analyzed from sugarcane planting (plant cane) until the first ratoon. Agronomic and economic sugarcane efficiency were evaluated. At the first ratoon, resin-extractable P provided by mineral and organomineral fertilizers were, respectively, 15 and 11 mg kg−1 in the 0.0–0.2 m, and 28 and 31 mg kg−1 in 0.2–0.4 m layer. However, exchangeable K in the 0.0–0.2 m layer was 1.88 and 1.58 mmolc kg−1 for mineral and organomineral fertilizers, respectively. The yield gains over the control reached with mineral and organomineral fertilizers were, respectively, 10.99 and 17 Mg ha−1 at the lowest fertilizer rate; and 29.25 and 61.3 Mg ha−1 at the highest fertilizer rate. Agronomic and economic organomineral fertilizer efficiencies are more pronounced in plant cane. Summing two harvests, the organomineral is 7% more profitable than mineral fertilizer.

The sugarcane (Saccharum spp) crop is recognized worldwide for its high biomass production capacity, sequestering thousands of tons of atmospheric CO 2 during its development, and it has a sustainable and very attractive balance related to greenhouse gas (GHG) emissions during its entire industrial process 1 . Thus, it is the main economically exploited crop in Brazil used for the production of clean and renewable energy, biofuel and sugar among other products 2 . In addition, the crop is responsible for creating millions of jobs and positively contributing to the environment 3,4 .
To meet the high nutritional demand of the crop, a great number of mineral fertilizers is commonly used to achieve satisfactory yields and sustainable revenue during all crop cycles. However, the increasing demand for efficiency and optimization of resources used during the agricultural production process generates uncertainties regarding the high mineral fertilizer consumption produced with imported raw material, substantially increasing the agricultural budget 5 .
In this sense, the recycling of agroindustrial organic waste appears as an alternative for substitution or supplementation of mineral fertilization for the sugarcane crop. Its reutilization in plant nutrition aiming for agricultural production is an excellent and sustainable form to supply the soil × plant × environment system demand 6 .
According to Raij 7 , the oldest fertilizers used in agriculture originated from organic sources, i.e. manure, sometimes referred to as natural fertilizers because of their origin. The agricultural use of organic waste constitutes an economically and environmentally viable practice mainly because it allows for the recovery of several chemical elements, such as nitrogen (N), phosphorus (P), potassium (K) and trace elements. In addition, it contributes through the addition of organic matter (OM) to the soil, improving the physical structure, water uptake capacity and nutrient supply to plants, thus increasing crop production 8 .

Material and Methods
The experiment was performed in a commercial area of the Agrodoce Agricultural Group at Boracéia-SP, Brazil, during 2016 to 2018. The location of the experimental area is 22°11' S and 48°48' W at 480 m altitude. According to the Köppen classification, the predominant climate in the region is Cwa, which is mainly tropical humid with a hot summer. The soil was classified as a sandy-textured Typic Hapludox 18 . Chemical and physical characterizations 19 were obtained from air-dried soil samples (0-0.20 and 0.20-0.40-m layers) passed through a 2-mm sieve (10 mesh) and showed the following results: Surface layer (sand, 705; clay, 48 and silt, 246 g kg −1 ); 5.6 pH (CaCl 2 ); 16 g dm −3 organic matter; 9 mg dm −3 P resin ; 1.1, 27, 10, and 15 mmol c dm −3 of exchangeable K, Ca, Mg, and H + Al, respectively; and a base saturation (BS) of 71%. For the subsurface layer (sand, 702; clay, 41 and silt, 256 g kg −1 ); 5.3 pH (CaCl 2 ); 10 g dm −3 organic matter; 14 mg dm −3 P resin ; 0.54, 21, 07, and 18 mmol c dm −3 of exchangeable K, Ca, Mg, and H + Al, respectively; and a base saturation (BS) of 61%.
The organomineral fertilizer is a commercial grade fertilizer manufactured at Solvi group fertilizer industry located in Coroados, São Paulo, Brazil (21°23'07.1"S 50°15'15.0"W), commercialized under the name of Organosolvi ® and is openly available to customers on www.organosolvi.com. The organic matrix of the organomineral is originated from the agroindustries located near the fertilizer plant, and is composed by byproducts of meat industry (rumen, blood, bones), dairy products, Fuller's earth (high absorbent and high CEC clay minerals used in the tallow industry), pine bark and eucalyptus. Before its use for organomineral production, it is submitted to a composting process. After composting its chemical composition showed the following results, in g kg Then, the compost is sent to a rotating drum that consists of shaping and sphericity, and then to a rotary dryer with a hot air flow (120 °C). After drying, the granules were classified in sieves of 1 to 4 mm, acquiring a granulometry very similar to the standard mineral fertilizer. Following this process, it was mixed with mineral fertilizers (MAP, KCl and urea treated with a nitrification inhibitor + polymer-based additive that controls the N availability and minimizes N losses via volatilization). The nitrogen sources used in mineral fertilization (MAP for plant cane and ammonium nitrate for ratoon) are not likely to lose N by volatilization due to the low pH resulted of their dissolution [20][21][22][23] . Thus the N amount supplied by both organomineral and mineral fertilizers would be the same, allowing to evaluate the organic matter matrix effect regardless the N losses from nitrogen sources. Urea has a www.nature.com/scientificreports www.nature.com/scientificreports/ high N content (45%), thus this is the best N source for organomineral fertilizer because it allows increasing the N content in it with the smallest increment in its final mass.
Following 6 months of treatment application, a period of full vegetative growth of sugarcane, 10 + 1 leaves or TVD (Top Visible Dewlap leaf), were collected within each plot in the two central rows according to the numbering system suggested by Kuijper 24 . Disregarding the leaf midrib and considering only the middle third of the leaf blade, the material was dried in an oven with forced air circulation at 60 °C until a constant mass was obtained. It was then milled and leaf N, P and K contents were determined 25 . Prior to harvest, the stalk number m −1 was determined by counting the stalks in the two central rows within the useful area of each plot and then converting to the number of stalks m −1 . Stalk weight, diameter, plant height, internode number and length were calculated as the means of the ten stems collected from each plot, clipped at the apical bud height, defoliated and measured using a digital scale, caliper and a ruler marked in meters from the soil surface up to the auricle region of the +1 or TDV leaf. After the growth evaluations, the cleaned stalks were sent to the Middle Tietê Sugarcane Planters Association (ASCANA) Laboratory, in Lençóis Paulista, SP, Brazil for processing according to the methodology defined in the Sucrose Content-Based Sugarcane Payment System, in accordance with Consecana's semiannual updates for the technological evaluations as described by Fernandes 26 . At harvesting, the four central rows of each plot were mechanically harvested and stalks were weighed using an electronic load cell. Then, the stalk yield was estimated extrapolating the values to stalk yield ha −1 , disregarding planting holes (gaps greater than 0.5 m). Sugar yield ha −1 was estimated as the product of the multiplication of the sucrose concentration (%) and stalk yield (Mg ha −1 ) at harvest. Following harvest, soil samples were collected at depths of 0.00-0.20 and 0.20-0.40 m in all plots using a Dutch auger. Five subsamples were randomly collected in planting row of each plot and combined into a composite sample. Soil chemical attributes were determined following the method of Raij et al. 19 .
The agronomic efficiency index (AEI) 27 and the economic efficiency index (EEI) were calculated as the percentage ratio between the stalk and sugar yields (AEI) and net profits (EEI) resulting from the mineral and organomineral fertilizers applied at the same rate in the plant cane and the residual effect in first ratoon. For AEI, the crop yield obtained in the treatment control was subtracted from both yields as follows (Eq. 1): where Y1 = crop yield in the control treatment; Y2 = crop yield using organomineral fertilizer at the corresponding rate; and Y3 = crop yield with mineral fertilizer at the corresponding rate. For EEI, the percentage ratio of the net profits obtained for both fertilizers was obtained by Eq. 2: where X2 = net profit using organomineral fertilizer at the corresponding rate and X1 = net profit using mineral fertilizer at the corresponding rate. The net profit was calculated using the Eq. 3: The revenue and fertilizers cost were converted using dollar quotation at each respective time (harvests, September 2017 and 2018; planting fertilization, February 2016 and first ratoon fertilization, October 2017). The revenue was obtained using the Eq. 4: The total economic efficiency index (EEI total) was estimated similarly to the EEI, but using the sum of net profit of plant cane and first ratoon.
Data for each crop season were separately subjected to analysis of variance. The fertilizer source means were compared using the t test (LSD) at 5% probability. The rate effects were evaluated via regression analysis using the statistical software package SISVAR 28 .

Results plant cane.
There was significant interaction between the sources and rates for the resin-extractable P content in the surface layer (Table 1; Fig. 1A). The mathematical adjustment for the mineral fertilizer was quadratic, while in the organomineral fertilizer it was linear, with a significant difference only when applying 120 kg ha −1 of P 2 O 5. Related to the subsurface layer, soluble P content significantly varied only as a function of the rates of P 2 O 5 ( Table 1; Fig. 1B).
The exchangeable K content in the surface layer linearly increased as a function of P and K fertilizer rates, without differences between sources (Table 1; Fig. 1E). For the subsurface layer, a linear interaction effect was observed between the sources and rates, being different at 120 and 180 kg ha −1 of K 2 O (Table 1; Fig. 1F).
There was not a significant change in the N, P and K leaf concentration (Table 2). On average, the values for N (18-25 g kg −1 ) and P (1.5-3.0 g kg −1 ) were within the reference limit for sugarcane 29,30 . K (10-16 g kg −1 ), although showing a similar value, is below the lower boundary range 29,30 .
Related to growth parameters, on average, plant height was significantly higher where organomineral fertilizer was applied (Table 2). For internodes per plant, the interaction effect between sources and rates was significant only for the organomineral fertilizer, linearly increasing as a function of fertilizer rates ( Fig. 2A).
There was a significant interaction effect between sources and rates for stalks (Fig. 3A) and sugar yield (Fig. 3C). For both parameters, the adjustments were linear, showing a response until the highest applied rate for organomineral fertilizer. For the mineral fertilizer, the adjustments were linear until the estimated rate at 132 and 120 kg ha −1 P 2 O 5 /K 2 O for stalks and sugar yield, respectively, stabilizing for the two last rates. On average, the organomineral fertilizer produced 9% more yield than the mineral fertilizer. www.nature.com/scientificreports www.nature.com/scientificreports/ The AEI followed the same trend of stalk and sugar yield results (Table 3). Organomineral fertilizer provided greater increases in stalk and sugar yields than those of the mineral fertilizer for all P and K rates, especially at the highest rate, being, on average, 98 and 113% more efficient, respectively. In spite of the higher fertilization cost, organomineral fertilizer was, on average, 12% more profitable when compared to the standard mineral fertilizer. The higher P 2 O 5 /K 2 O rate (180 kg ha −1 ) was 27% more profitable. first ratoon. Related to the soil residual effect of P, there was significant interaction between the sources and rates for soluble P content in the surface layer (Table 1). Following the same trend as that of the plant cane, the adjustment in the soil soluble P content was quadratic for the mineral fertilizer application and linear for the organomineral fertilizer (Fig. 1C). The mineral fertilizer was higher up to 150 kg ha −1 P 2 O 5 with no significant difference at the highest rate (Fig. 1C). Related to the subsurface layer, the soil soluble P content positively ranged as a function of P 2 O 5 rates (Fig. 1D). For P 2 O 5 sources, on average, soil soluble P content was statistically higher using organomineral fertilizer. Because P 2 O 5 was not applied in the first ratoon, the P content was smaller than the plant cane evaluation, as expected because of the sugarcane uptake.
For the exchangeable K content, in both layers, a significant interaction between sources and rates were observed with a quadratic adjustment (Table 1). In both cases, the exchangeable K content was higher when the mineral K fertilizer was applied and, in the surface layer, the interaction was not significant only at the highest K 2 O rate (Fig. 1G). Regarding to subsurface layer, there was not a significant difference in exchangeable K content only up to 60 kg ha −1 of K 2 O above which it was higher with the mineral fertilizer application for the other three higher K 2 O rates (Fig. 1H).
There was significant polynomial interaction effect between the sources and rates for the N and P leaf concentration ( Table 2). The mathematical adjustment was quadratic for both (Fig. 2B,C). Related to N, there was no significant difference for the first rate applied; however, for the other rates, the N uptake was higher using the mineral fertilizer. In spite of this, the N leaf concentration was less than the lower boundary range (18-25 g kg −1 ). The P leaf concentration was also higher using mineral fertilizer at rates equivalent to 120 and 150 kg ha −1 P 2 O 5. Only the P leaf concentration in the treatment control was below the range considered optimal for sugarcane (1.5-3.0 g kg −1 ). The K leaf concentration linearly increased as a function of the applied rates, without differences between the sources (Fig. 2D). Similar to P, only in the control was the K leaf concentration less than the range considered optimal for sugarcane (10-16 g kg −1 ).
A significant difference was not observed related to growth parameters; however, stalks and sugar yield linearly increased as a function of fertilizer rates ( Table 2; Fig. 3B,D). For stalk yield, as well, there was a significant difference related to the sources, being, on average, 3% higher with organomineral fertilizer application ( Table 2). The AEI (Table 3) for both parameters was lower than the plant cane. On average, the increase in stalk and sugar yields was 44 and 12% respectively higher than mineral source. Nevertheless, the EEI was practically the same, not showing economic gains at the ratoon. Considering the sum of the two evaluated harvests, the use of organomineral fertilizer proved to be more profitable (7%, on average), mainly in the higher rate of P and K (14% at rate of 180 kg ha −1 of P 2 O 5 /K 2 O) ( Table 3).

Discussion
Soil soluble P in the surface layer, was higher in the treatments with mineral fertilizer application with quadratic distribution of the points (Fig. 1A,C). However, soil P content with organomineral fertilization linearly increases, without a difference in the mineral P at the highest P 2 O 5 rate. This difference in P content may be related to its solubility. The P from the mineral fertilizer is readily soluble, detectable in the resin-extractable P analysis; otherwise, organomineral P depends on soil mineralization for availability, demanding an OM-P linkage breaking from the organic structure, which can be characterized as slow P release 31,32 .
Because the soil P application occurred in the planting furrow, the highest soil P content was in the subsurface layer, linearly increasing as a function of the P 2 O 5 rates. Although there was no P application in the first ratoon, the P content remained high because of its soil residual effect. It is likely the increase in the stalks and sugar yields was related to the increase in soil P content in the subsurface layer, because P plays an important role in sugarcane rooting and tillering, positively affecting stalk and sugar yields 33,34 . Similarly, in this study, other authors have shown that mineral fertilizer, when associated with any organic compound, can be more efficient mainly by the reduction in P adsorption onto Al and Fe minerals 11,35,36 . Despite soil resin-extractable P content in organomineral fertilizer treatments have been higher, it should be emphasized that this study was conducted in a largely sandy-textured soil, in which P adsorption is naturally reduced 36,37 allowing high resin-extractable P content regardless of the P 2 O 5 source. Thus, stalks and sugar yield in first ratoon were not affected by these variables. Besides the effects of the organomineral organic matrix on reducing P adsorption; the addition of organic matter to the soil may help to maintain its moisture near fertilizer placement for more time which favor the soil P diffusion 38,39 , and increases the P amount reaching the roots.
In the plant cane, although there were linear increases in soil K content as a function of the K 2 O rates for both sources, there was no difference in the K leaf concentration. In relation to the first ratoon, the maximum estimated soil K content for the first layer was obtained at the rate of 100 kg ha −1 of K 2 O (mineral fertilizer) and 99 kg ha −1 of K 2 O (organomineral fertilizer). Although this dose is practically the same, a higher solubility for the mineral fertilizer was noted, because the quantified soil level was, respectively, 2.4 and 1.9 mmol c kg −1 . The high CEC value, quantified to the OM portion of the organomineral fertilizer, may adsorb part of the K applied, controlling its solubility in the soil. Rosolem et al. 14 observed a higher efficiency with humic substances coating KCl when the fertilizer was applied in a single dose. They attributed this to the slow release because of the high CEC of the humic and fulvic acids. For perennial crops, such as sugarcane, this process can aid in the gradual nutrient release and decrease potential losses via leaching and runoff 37 . The higher P concentration in soil subsurface is due to the fertilizers were applied in the planting furrow (0.20-0.30 m depth). The higher K content on the surface is due to the fact that sugarcane straw releases more than 50 kg K 2 O ha −1 40 . Even in the sugarcane plant, where the amount of straw on the soil is much smaller, there may be a contribution of K present in the leaves that fall on the soil as the plant grows, since the release of K by the straw is relatively fast. Oliveira et al. 41 observed that the sugarcane straw released 85% of its K content during the first year following sugarcane harvest.
Although the soil K content showed a quadratic adjustment, in the leaf, the increase was linear. The linear increase, also observed in stalk and sugar yields, may indicate there was higher K uptake by the plant, justifying the decrease in soil K content at the highest K 2 O rates. As well reported by Almeida et al. 42 , the increase in sugar yield by K 2 O rates may be related to the increase in stalk yield and K leaf concentration, because K acts in the transport via the phloem and carbohydrate storage 43 .
There was no variation in the N, P and K leaf concentration in the plant cane. Notably, there was source efficiency allowing the culture to maintain its nutritional status within the proper range. In addition, as the sugarcane was   www.nature.com/scientificreports www.nature.com/scientificreports/ planted in a conventional soil tillage system, with rotation of the soil arable layer, there is natural organic matter mineralization, releasing nutrients to the plants and contributed to their nutrition 44 .
In the first ratoon, the range of N, P and K leaf concentrations among the treatments did not alter the stalks and sugar yield. For N, even considering that the organomineral source uses a urease enzyme polymer inhibitor, the N leaf concentration was less than that of the treatments using mineral fertilization and both were below the range considered optimal for sugarcane. The quadrac fit find for N and P concentration in leaf may be due to antagonism between K and N as well as between Cl and P. Considering that the source of K in both fertilizers was mainly muriate of potash (KCl). However, while the effect of K on N absorption is well known and accepted; 45,46 the antagonism between Cl and P remains uncertain 47,48 . Also, quadratic behavior of both nutrients may be caused by the dilution effect, that is biomass accumulation is constant while nutrient absorption rate is reduced 49,50 .
The AEI (Table 3) showed a sugarcane positive response in both harvests, but more pronounced in plant cane. Similarly, De Souza 15 concluded that in 90% of evaluated areas, organomineral fertilizer promoted better sugarcane performance, mainly in plant cane. Teixeira et al. 16 also reported higher efficiency in stalk and sugar yield in plant cane using organomineral fertilizer at the highest rate, obtaining the same stalk yield with approximately 30 kg ha −1 less P 2 O 5 fertilizer. Ramos et al. 17 also observed a higher stalk yield applying organomineral fertilizer compared to poultry litter and mineral fertilizer in plant cane, recommending it for sugarcane. Because of intense soil tillage for sugarcane establishment, the soil organic fraction is stimulated to mineralize and organomineral fertilizer or other organic sources, e.g., filter cake, can contribute to higher agronomic stability 31 as reflected in higher yields.
Regarding the first ratoon, perhaps sugarcane straw can influence mineral fertilizer efficiency. Satiro et al. 51 evaluated sugarcane straw removal's effects on soil degradation in the first and second ratoons and concluded that the impacts on soil chemical attributes were significant, mainly in sandy-textured soil. Because in this study sugarcane straw was not removed from the area, OM additions could contribute to a more balanced soil environment. However, further long-term field-scale research of differently textured soils is needed to explore in depth organomineral fertilizer performance for all planned sugarcane ratoons, providing more information regarding the soil residual effect of the nutrients and their influence on the stalk and sugar yield.
The higher efficiency of organomineral fertilizer, added to indirect parameters related to organic compounds reported in the literature, such as the slow release effect of nutrients, adequate soil biological activity, improvements in physical and chemical soil quality, better water retention capacity and soil porosity 32 , perhaps can explain the better sugarcane performance in this study.
Economically, in plant cane, in spite of the higher cost of production, due to the higher amount applied with organomineral, the increase in stalks and sugar yield guaranteed higher net profit (12%) compared to the mineral fertilizer. However, in the first ratoon, the profitability of both sources was equivalent. Considering the sum of the net profit in the two harvests, a greater profitability was obtained with the use of the organomineral fertilizer (7%), which justifies its use.
Considering these results, organomineral fertilizer is a promising agronomic and economic alternative for sugarcane producers. Cherubin et al. 31 highlighted the importance of fertilizer management strategies including a balance between organic and mineral P sources to improve the yield, soil quality, and environmental sustainability of Brazilian sugarcane production. However, some concerns such as the concentration of potentially toxic elementssuch as plumbum (Pb) and arsenic (AS) 52 , availability of the required amount and logistics for large areas should be considered.

conclusions
Organomineral fertilizer is suitable to supply sugarcane requirements and can completely replace mineral fertilizer. However, its influence on sugar yield is lower than on stalks yield. In addition, organomineral fertilizer efficiency in stalks and sugar yield is more pronounced in plant cane, being, on average, 96 and 113% more efficient than mineral fertilizer, respectively. Consequently, organomineral fertilizer is more economically efficient in plant cane, being, on average, 12%. In the two harvests summed, the organomineral is 7% more profitable than mineral fertilizer.    Table 3. Increased yield of stalks (IStY) and sugar (ISgY), agronomic efficiency index of stalks (AEIst) and sugar (AEIsg), economic efficiency index (EEI) and total economic efficiency index (EEI total) of of stalks yield as a function of P 2 O 5 /K 2 O sources and rates. (a) Increase in stalks yield relative to the mean yield in control; (b) Agronomic efficiency index of the organomineral relative to the mineral fertilizer in stalks yield; (c) Increase in sugar yield relative to the mean yield in control; (d) Agronomic efficiency index of the organomineral relative to the mineral fertilizer in sugar yield; (e) Revenue (US$ ha −1 ) = ((Value of TRS (US$ kg −1 ) × sugarcane TRS (kg Mg −1 )) × stalk yield (Mg ha −1 )); (f) Costs of fertilizers (US$ ha −1 ) = Fertilizers price × respective applied rate; (g) Net profit (US$ ha −1 ) = Revenue -Costs; (h) Economic efficiency index (EEI) of the organomineral in relation to the mineral fertilizer in profitability in the stalk yield; (i) Total Economic efficiency index (EEI total) is the sum of EEI in 2017 + 2018. *Average dollar value (february/2016 − US$ 1.00 is equivalent to R$ 3.99; september/2017 -US$ 1.00 is equivalent to R$ 3.16; november/2017 − US$ 1.00 is equivalent to R$ 3.