COFFEE SILVERSKIN AND EXPIRED COFFEE POWDER USED AS ORGANIC FERTILIZERS

The coffee industry produces a wide range of organic wastes, some in large amounts, and most of them do not have a well-defined final disposal. The agricultural use of these wastes can be based on the recycling of nitrogen (N), but their chemical characterization and evaluation with soil under controlled conditions are required. The aim in this work was to evaluate the use of coffee silverskin and expired coffee powder as organic fertilizers. The wastes were chemically characterized according to CONAMA’S resolution No. 375 and passed through Brazil’s Agriculture Ministry normative for organic fertilizers. A greenhouse experiment with Rhodic Ferralsol soil and maize as a model plant was carried out to obtain the agronomic efficiency index (AEI) for nitrogen. Three treatments were considered: one control (without waste), ammonium nitrate as a mineral reference, and both organic wastes, at a dose of 450 mg of Kjeldahl nitrogen per pot, in triplicate. The characterization results were favorable to the wastes as nitrogen sources for agriculture. However, the AEIs obtained were low (0.5 and 7.9% for the expired coffee powder and the coffee silverskin, respectively) compared to that of the mineral reference (92%). Although coffee silverskin and expired coffee powder showed nitrogen levels were well above the minimum require by normative (0.5 % m/m), its use as organic fertilizers (as exclusive sources of N) is not recommended since less than 10% of N was available as showed by the AEI index.


INTRODUCTION
The presence of plant nutrients, mainly nitrogen, is associated with a high content of organic matter in many organic wastes.For this reason, its use as nitrogen source in agriculture are particularly interesting, as it can be both environmentally and economically valuable (PAULA et., 2013).Based on this relationship, organic fertilizers must have at least 0.5 % (m/m) of total Nitrogen in its content, according to the Brazilian normative Nº25 from the Agriculture Ministry.Also, a minimum content of 15% (m/m) 1,2,35 Instituto Agronômico -Avenida Barão de Itapura, 1.481 -Botafogo -13.020-902 -Campinas -SP -rcarnier10@gmail.com,berton@iac.sp.gov.br,aline@iac.sp.gov.br,jezuanazzi@outlook.com 4 Embrapa Meio Ambiente -Rodovia SP 340 -Km 127,5 -Tanquinho Velho -1 3.820-000 -Jaguariúna -SP -adriana.pires@embrapa.br of carbon is required.If the minimum amount of N is not reached, organic wastes with potential use in agriculture might be registered as soil conditioners.For this use, only its carbon content and cation exchange capacity are regulated by the Brazilian normative Nº35 (BRASIL, 2006).The presence of any other mineral nutrient at material used as organic fertilizer requires minimum contents addressed also by Brazilian normative Nº25.
Brazil is the largest coffee producer, being responsible for 30.1% of coffee beans grown worldwide, and has a trade value of more than

MATERIAL AND METHODS
Coffee silverskin and the expired coffee powder were obtained from a sewage sludge treatment plant where these materials are used for sewage sludge composting.The samples were air-dried, homogenized and sieved through 2-mm.All analyses were run in triplicate.Chemical characterization followed the standard established by the National Environment Council resolution No. 375 (CONAMA, 2006).Moisture was determined according to EPA-SW 846, with drying to constant weight at 65°C.The pH was measured using 2 g of moist sample and 20 ml of deionized water, and the mixture was stirred for 5 min at 220 rpm and rested for 30 min.For inorganic N, 5-g samples was distilled with 50 ml of 1.0 mol/L KCl, 0.2 g of MgO, and 0.2 g of Devarda alloy.The distillate was collected in 5 mL of 20 g/L H 3 BO 3 and titrated with 0.0025 mol/L H 2 SO 4 (BREMNER, 1996).Nitrite and nitrate were determined according to Mulvaney (1996).For organic N (N-Kjeldahl), 0.05 g of oven-dried samples were mixed with 3 mL of concentrated H 2 SO 4 and placed in a digester block (360°C) for 3 h; the sample was then distilled with 20 mL of 10 mol/L NaOH.The distillate was also collected in 20 mL of 20 g /L H 3 BO 3 and then titrated with 0.0025 mol/L H 2 SO 4 (APHA, 2017).Total solids and volatile solids were obtained by drying samples at 105°C and then combusting at 550°C for 4 h (APHA, 2017).Organic carbon was determined by the K 2 Cr 2 O 7 method (NELSON; SOMMERS, 1996).Ca, K, P, Mg, S, Cu, Fe, Ni, Mn, Mo, Si, Zn, Al, As, Ba, Cd, Cr, Pb, Hg, and Na were extracted in a microwave oven (EPA, 2018).K and Na were quantified by flame photometry and the other elements by ICP-AES.The cation exchange capacity (CEC) was measured following the procedure described by Normative No. 17 (BRASIL, 2007).The electrical conductivity (EC) was measured in a 1:5 (10 g of waste: 50 mL of water) extract.From triplicate data, the average and standard deviation were calculated.The C:N ratio was obtained considering N-Kjeldahl and organic C.
The experiment was carried out in a greenhouse at the Agronomic Institute of Campinas, State of São Paulo,Brazil (22º53'37.27" S;47º3'54" W), from December 2016 to February 2017.The soil sample, collected from the 0-20 cm depth of a Rhodic Ferralsol soil, was air-dried, homogenized and sieved through 2-mm.The soil characteristics are shown in Table 1 and were determined according to Raij et al. (2001).20.6 billion Brazilian reais (CONCEIÇÃO et al., 2017).However, within the coffee bean agricultural chain, approximately 50% of the volume of products generated are wastes, such as the coffee husks, pulp, parchment and coffee silverskin (MATIELLO, 1991;ESQUIVEL;JIMÉZEZ, 2012).Several of these wastes may be used in agriculture as fertilizers or soil conditioners if meet the normative requirements listed above.The reuse of coffee silverskin, the main waste product of the coffee-roasting industry, could be also a significant alternative to its environmental disposal (TOSCHI et al., 2014).
In contrast to mineral fertilization, the use of organic materials as fertilizers depends on organic matter mineralization for the plant's nutrient release (ECKHARDT et al, 2016).For this reason, nitrogen availability in the form of inorganic nitrogen is often lower than that in mineral fertilizers.These characteristics can protect the environment and meet the plant's needs.Organic matter mineralization is affected by several factors, such as pH, C/N ratio and water availability (PAULA et al., 2013).For this reason, only the presence of the minimum amounts of nitrogen, according to the Brazil's Agriculture Ministry normative cited earlier, in new fertilizers based on recycling of organic wastes may not be enough, and the evaluation of nitrogen availability in tests with soils by means of the agronomic efficiency index (AEI) is advisable.
In addition, besides to the agronomic efficiency index (AEI) testing for nitrogen, the chemical characterization of organic wastes that shall be used as new organic fertilizers is mandatory as a first step in its evaluation.The chemical characterization of organic wastes is the basis for the agronomic recommendation and is usually performed following the requirements described in the CONAMA'S resolution No. 375.The characterization of the wastes is important not only to identify the concentration of macro and micronutrients but also to indicate other attributes and relationships among attributes in the waste composition.Furthermore, it can also show the presence of heavy metals, that limits are regulated by the Brazilian normative instruction Nº7 (BRASIL, 2016).After the characterization, the data was compared with Brazil's Agriculture Ministry normative for organic fertilizers to verify if it is in accordance.
The aim of this work was to evaluate the use of coffee silverskin and expired coffee powder as organic fertilizers.
The greenhouse experiment was performed using 2-kg pots vases with maize, as the test plant.Five plants were kept per vase.The treatments were control (without residue), ammonium nitrate as mineral reference, and both organic wastes, at a dose of 450 mg of Kjeldahl nitrogen per pot.The corresponding experimental design was a completely randomized block design, with four treatments, in triplicate, with a total of 12 vases.
The pots were prepared as follows: 1-the soil base saturation was settled to 70% by the addition of 2.7 g of lime per pot, and the moisture was adjusted to 60% of soil water retention, with incubation for 10 days; 2-to each pot, the calculated amount of waste or ammonium nitrate was added and incubated for 15 days; and 3-all pots were fertilized by the addition of 0.45, 0.4, 0.46, 0.27 and 0. 17 g per pot of K, P, Ca, Mg and S, respectively, plus micronutrients (2.0, 1.0, 0.28, 1.0, 6.3 and 8.0 mg per pot of B, Mn, Mo, Fe, Zn and Cu, respectively).After fertilization, 10 healthy and homogeneous seeds of maize (AG8676PR02 variety) were sown uniformly distributed on the surface of each pot at a depth of 2 cm.After emergence, five seedlings were kept per pot.Thirty-seven days after seeding, the shoots were measured and cut.Fresh weight was registered, followed by washing with water and oven drying until a constant weight at a 65 ºC ± 5 ºC.The dry matter was milled, and its N Kjeldahl determined (BREMNER, 1996).
Soil and plant data were submitted to ANOVA and Tukey's test for average comparison of treatments (p ≤ 0.05).To perform these analyses, the statistical program used was the SISVAR ® -5.6 version.
Shoot nitrogen extraction was calculated as follows (equation 1): Where: SNE is the shoot nitrogen extraction (mg pot -1 );DM is dry matter (g pot -1 ); N content is the nitrogen found in shoots (g kg -1 ).The agronomic efficiency index (AEI) was calculated as described by Farinelli e Lemos (2010) (equation 2): Where: SNE0 is the shoot nitrogen extraction in the control; SNE1 is the shoot nitrogen extraction in the nitrogen sources (wastes or ammonium nitrate) and N applied that was 450 mg of N per pot.
The relative agronomic efficiency index (RAEI) was calculated as follows (equation 3):

RESULTS AND DISCUSSION
The chemical characterization of the organic wastes is shown in Table 2.The moisture of both materials was quite low (1.2 and 16.6% (m/m) for expired coffee powder and coffee silverskin, respectively) and was consistent with the roasting of the beans to which the wastes were associated.The low moisture of such material is favorable when transport costs for its application are considered.The pH of coffee silverskin was 6.8, and that of expired coffee powder was 5.0.A soil pH in the range 6.0 to 7.0 is considered optimum for plant nutrient availability.Previous studies showed that application of high ratios of expired coffee powder could result in a soil pH below 6.0 (MELO et al., 2008).However, the final soil pH after organic amendment depends on several factors such as the rate of waste application, the waste reaction with the soil and the soil buffering capacity and organic matter content (RONQUIM, 2010).----------------------------------mg kg -1   ---------------------------------- a= Average and standard deviation for three replicates, b= 1:10 (waste:water), c= dry basis, d= lower than the quantification limit Both wastes presented similar concentrations of organic carbon (43% to 46% of Organic Carbon) and organic nitrogen (2.3 to 2.6% of N-Kjeldahl).Based on these values, the C:N ratio of coffee silverskin and expired coffee powder was quite similar and close to 20.The mineralization of organic nitrogen is favorable over its immobilization in this condition (CANTARELLA, 2007), making the agricultural use of such wastes promising.
In addition to this, either the organic carbon, the N-Kjeldahl of the material tested was in accordance with the organic wastes minimum content established by the Agriculture Ministry at Normative No. 25 of 2009, which corresponds to 150 g kg -1 and 5 g kg -1 for C and N, respectively.
The inorganic nitrogen, in the form of NH 4 + and NO 3 -+ NO 2 -ions, was quite different among the wastes, being higher for the expired coffee powder than for coffee silverskin .However, for both wastes, the NO 3 -content was higher than NH 4 + .This can be related to a high oxidation level of the materials since over time, a decrease in NH 4 + and increase in NO 3 -due to the bacterial nitrification are expected (SANCHÉZ-MONEDERO et al., 2001).The agricultural use of wastes with high amounts of NO 3 -needs additional caution since NO 3 -may be easily leached through the soil profile, causing environmental damage (MELO et al., 2008).Since more than 90% of the nitrogen in the wastes is in the organic form, even the nitrate content in the expired coffee powder was not troublesome.Furthermore, the nitrate can be retained by the positive charges due to the presence of Fe and Al oxides in the soil, principally in the deep layers (OLIVEIRA et al., 2000).The CEC, 354 mmol c dm -3 for expired coffee powder and 523 mmol c dm -3 for coffee silverskin, was above the limit for soil conditioner set by Normative No. 35 (BRASIL, 2006).Considering the typical low organic matter levels of tropical soils (FONTANA et al., 2014), the addition of such wastes could significantly improve the fertility of these soils by means of nutrient retention.Aside from that, the humification degree of the coffee silverskin, based on the CEC/OC, was superior to that of the expired coffee powder.The humification degree also indicates the organic matter maturity, which reflects an increase of the material's nutrient retention capacity (MELO et al., 2008;PAIVA et al., 2013).However, further testing would be need to verify such material impact on soil CEC, for soil conditioning.
The electrical conductivity of the wastes was 4.9 and 7.5 dS m -1 for expired coffee powder and coffee silverskin, respectively.Electrical conductivities in the range from 0.64 to 6.85 dS m -1 are commonly used in agriculture (MELO et al., 2008).Thus, the long-term usage of expired coffee powder grounds in agriculture may require soil solution monitoring in order to prevent salinize.
Overall, higher nutrient concentrations were found in coffee silverskin than in expired coffee powder.It is not clear if this difference may be due to the roasting of the coffee bean or if it is related to the coffee tissue's physiology.It is known, however, that the coffee plant demands high potassium fertilization in leaves, comparable to that of nitrogen (MARTINEZ et al., 2014).This was indeed reflected in the wastes composition, with higher concentration in the coffee silverskin The presence of a considerable concentration of K in this waste makes its particularly interesting for tropical soils, which are poor in K, and many crops besides coffee itself, where some wastes are traditionally returned to the fields (LIMA, 2014;MOURA, 2016).The use of coffee silverskin or expired coffee powder as part of the potassium fertilization under those conditions might represent a significant advantage since KCl represents the major input of K for the crops (RODRIGUES et al., 2014).However, further testing would be need to verify this.Expired coffee powder also demonstrated considerable S presence.Sulfur in plants is related to protein synthesis, and it is absorbed in the form of sulfate released during organic matter mineralization (FIORINI et al., 2016;SOARES et al., 2017).
The micronutrients Fe and Mn were found in high concentration in both wastes.This can be attributed to the natural high concentration of such elements in Brazilian soils (BIONDI et al., 2011).Although coffee is not a molybdenum-demanding crop (MARTINEZ et al., 2014), a considerable concentration of molybdenum, higher than the required amounts of B, Cu and Zn, was found in the expired coffee powder.
The potentially toxic elements (Pb, Ni and Cd) were below the detection limits of the characterization method (Table 3).For both wastes, low to trace concentrations of selenium (12.4 ± 0.73 and 5.5 ± 0. 37 mg/kg for coffee silverskin and expired coffee powder, respectively) and barium (35.7 ± 1.8 and 14.0 ± 0 1.6 mg/kg for coffee silverskin and coffee silverskin respectively) were found.Only coffee silverskin contained some arsenic (5.0 ± 0.4 mg/kg) and chromium (1.7 ± 0.4 mg/kg).These results are well below those allowed by Normative No. 7 (BRASIL, 2016).
The maize grown in soil treated with the coffee organic wastes, did not differ from that of the control (p<0.05).For all treatments, plants were shorter than those grown in soil treated with ammonium nitrate, where all the nitrogen added in the inorganic form was available to plants (Figure 1).This behavior may indicate that the nitrogen mineralized from coffee silverskin and expired coffee powder was not reflected in plant growth, or if nitrogen mineralization occurred, the presence of another toxic substance (such as tannins) inhibited plant growth.Since chlorosis symptoms were observed, nitrogen deficiency seems to explain the plants' inferior development (COSTA et al., 2015).
After 52 days of the addition of the coffee organic wastes to the soil (15 days of incubation plus 37 days after maize sowing), the shoot nitrogen content and nitrogen extraction per pot did not differ for treatments that received the coffee organic wastes and the control (Table 4).As observed for the plants height, shoot dry matter weight did not differ.Once more, both the control and treatments with coffee wastes applied differed significantly from the mineral reference treatment.This trend was also reflected by the AEI and RAEI of coffee silverskin and the expired coffee powder when compared to the ammonium nitrate treatment (p ≤ 0.05).Considering the costs involved in the high waste volumes needed to be applied in the field and its low nutrient availability, AEI values lower than 60% are considered low and not economically valuable.The low nitrogen availability, despite a favorable C:N ratio in coffee silverskin and expired coffee powder, may be due to the organic composition of the wastes associated with the presence of recalcitrant organic compounds, such as cellulose and hemicellulose (SANTOS et al., 2012).Mussato et al. (2011) reported that the presence of fibrous tissues in coffee silverskin was associated with those compounds.Lignin is also related due the high stability of vegetable tissues (BRUM, 2007).
In addition to recalcitrant compounds, the limited nitrogen availability of coffee silverskin and expired coffee powder may be related to the presence of caffeine, tannins and polyphenols.These compounds are toxic to microorganisms and may avoid organic matter mineralization (MUSSATTO et al., 2011;HERMOSA, 2014).Inhibition of extracellular enzymes and direct action on microbial metabolism are some mechanisms of microorganism inhibition due to the presence of tannin (SCALBERT, 1991).Besides that, a protein-tannin complex may be formed in the soil (KRAUS, 2003), that are resistant to N release (HOWARD; HOWARD, 1993).
In addition to this, caffeine and polyphenols are allelopathic to several vegetable species (LIMA et al., 2014).Chandra et al. (2013) found allelopathic activity against Cicer arietinum and Triticum aestivum caused by coffee bean extract, with a significant reduction of germination for both species.

CONCLUSIONS
Although coffee silverskin and expired coffee powder showed nitrogen levels were well above the minimum require by regulation (0.5 % m/m), its use as organic fertilizers (as exclusive sources of nitrogen) is not recommended since less than 10% of nitrogen was available as showed by the AEI index.
The use of these wastes as soil conditioners is promising due to their high organic matter content and CEC.In addition, these wastes may be used as complimentary sources of potassium.However, further testing is needed to confirm this

ACKNOWLEDGEMENT
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -Brasil (CAPES) -Finance Code 001.

FIGURE 1 -
FIGURE 1 -Average height of maize plants in soil treated with coffee organic wastes, and controls.Columns with the same letters do not differ significantly by Tukey test (p<0.05).
standard deviation for three replicates, b= lower than the quantification limit, c= no informed
1 Soil organic matter -dichromate oxidation and colorimetry, pH in CaCl 2 , P, K, Ca, and Mg by ion exchange resin and H+Al by SMP buffer.Cu, Fe, Mn and Zn-DTPA pH 7.3.V-base saturation, B.S.-sum of basis; CEC-cation exchange capacity

TABLE 2 -
-Main characteristics of the organic wastes.

TABLE 3 -
Potentially toxic elements from the organic wastes.

TABLE 4 -
Nitrogen absorption and efficiency index of nitrogen for maize.Different letters indicate a significant difference between treatments at the 5% (P<0.05) level by Tukey's test.AEI = agronomic efficiency index, RAEI = relative agronomic efficiency index