1 Introduction

Coffee is a very important international commercial product and the most popular soft drink in the world (Wariyo et al 2021). About 70% of the coffee consumed worldwide belongs to the Arabica species (Coffea arabica L.) and its estimated global production is close to 170,560,000 60–kg bags (Lemma and Megersa 2021). Coffee cultivation is responsible for several by-products, including fresh pulp, which accounts for about half of the weight of fresh coffee cherries (Reichembach and de Oliveira-Petkowicz 2020). These wastes are rich in cellulose, proteins, minerals, polyphenols, tannins, and caffeine that can cause pollution to the environment if not disposed of properly (Elida et al. 2020). Thus, it is recommended for large processing plants to have in place systems to process coffee pulp. One of these methods is composting, which ensures that coffee pulp undergoes biological degradation minimizing waste disposal and transforming the pulp into a usable amendment (Kufa et al 2008).

The use of biological waste to produce biochar (BC) can play an important environmental and economic role in agriculture (Owsianiak et al 2021). BC is manufactured by the thermochemical decomposition of biomass under conditions of limited oxygen supply (pyrolysis) and its use is an alternative to improve the physical and chemical conditions of soil (Coomes and Miltner 2016). BC application has been reported to help in the alkalization of soils by increasing the pH (Sorrenti et al 2016), cation exchange capacity, total carbon, and water holding (Chan et al 2008; Kuzyakov et al. 2009; Van Zwieten et al 2010), and reducing the exchangeable acidity and bulk density (Ch'ng et al 2015; Al-Wabel et al. 2017). The transformation of coffee pulp through the pyrolysis process may be an alternative for the management of this residue due to its potential for use as a soil amendment, providing numerous benefits to crop yield and nutrition. The application of BC can also be an affordable alternative for farmers to improve nutrient availability in the soil (Yazhini et al 2020). It has been reported that nutrient availability improves with the application of BC, finding increases of around 48% in organic carbon and 22.6% in available phosphorus in the soil at a dose of 20 t ha−1 BC of wheat straw. Increases of 2.41% in nitrogen, 38.8% in phosphorus, and 88% in potassium have also been registered in soils treated at a dose of 22 t ha−1 BC of peanut shells (Gaskin et al 2010; Zhang et al 2016).

The use of BC also helps certain physical properties of the soil. For example, applications of 7.18 t ha−1 BC from corn stover feedstock increased the aggregate stability values (mean weight diameter, MWD), showing the highest records when the pyrolysis temperature was higher (BC at 350 °C = 2.88 and BC at 550 °C = 3.10) compared to soils treated with 0 t ha−1 BC (MWD = 2.78) (Herath et al 2013). Additionally, applications of BC manufactured with corn crop residues at a rate of 10% (w/w of the soil) increased water holding (43%) and reduced the bulk density (ρa = 1.01 g cm−3) compared to soils not treated with BC (water holding = 0.30 and ρa = 1.26 g cm−3, respectively) (Igalavithana et al 2017).

Labor and other production costs are high, and it is, therefore, important to study alternative technical agronomic strategies to increase crop yield in middle-income countries (Owsianiak et al 2021). BC can offer several potential co-benefits in least-developed countries (LDC) because it can positively affect crop yields and soil health at a relatively low cost (Hansson et al 2021). In this sense, BC is a promising soil amendment able to improve soil health and promote crop production since the application of BC in combination with chemical fertilizers is necessary to achieve the synergic beneficial effects of this amendment on crop growth (Guo 2020).

Biochar has been obtained from different crop species such as cereal, fruit, or perennial crops (Sánchez-Reinoso et al. 2020). The composition of coffee pulp has turned it into a by-product of interest for research and the development of multiple goods. It can be used as a biofertilizer, biological control agent, animal feed concentrate, nutritional supplement for human consumption, and in the production of different bioproducts and bioenergy (Carmen et al 2020). The study of agricultural by-products such as coffee wastes has recently gained importance to improve soil conditions in different regions around the world (Kiggundu and Sittamukyoto 2019; Milian-Luperón et al. 2020). However, studies of the use of BC on soil health are still scarce in coffee regions. It has been reported that the application of BC (obtained from grass and Eupatorium sp. residues) to soils for coffee cultivation increased the pH and organic matter of the soil, improving soil fertility and the nutrient cycle (Gautam et al 2017). In consequence, the pyrolysis of coffee wastes such as coffee pulp may be a promising agronomic strategy to improve the sustainability and fertility of soils in coffee production systems. Therefore, the objective of this study was to evaluate how the application of different doses of biochar obtained from coffee pulp (0, 4, 8, and 16 t ha−1) and chemical fertilization levels (0, 33, 66, and 100% of the nutritional requirements) in mature coffee trees influenced the physical (soil water holding, bulk density, stable aggregates), chemical (effective cation exchange capacity, nutrient availability, pH), and biological (microbial respiration) properties of an Andisol soil.

2 Material and Methods

2.1 Plant Material and Growth Conditions

The experiment was carried out in the Luxemburgo farm, located in the municipality of Chaparral in the department of Tolima (Latitude 3°49′39.2″ N, Longitude 75°34′07.1″ W, altitude: 1875 masl) using the same orchard, between August 2018 and August 2020. Figure 1 summarizes the meteorological conditions (radiation, temperature, and rainfall) during the research. Data were obtained from the nearest meteorological station [Chaparral (latitude 3°37′35.193″ N, longitude 75°32′35.483″ W, altitude: 914 masl)] of Meteoblue weather. The trial was carried out on the same 3-year-old coffee (Coffea arabica L.) trees of cultivar Castillo during the experimental time. The trees were spaced at 1.70 m × 1.30 m (4500 trees/ha). The soil was classified as a Typic Hapludand according to the USDA Soil Taxonomy (Agriculture Department 2014) and had the following characteristics: (i) parent material of volcanic ash deposits on granodiorites and quartz monzonites (Ibagué Batholith), the mountain landscape, a type of erosional slope relief; (ii) loam texture (24% clay, 30% silt, and 46% sand); (iii) pH 5.50; iv) electrical conductivity 0.17 dS/m; (v) effective cation exchange capacity (ECEC) 11.51 cmol+/kg. Finally, the studied soil showed the following nutrient concentration: Total N: 0.27%; P: 0.98 mg kg−1; K: 0.50%; Ca: 7.85%; Mg: 2.52%; Cu: 1.40 mg kg−1; Fe: 124 mg kg−1; Mn: 65 mg kg−1; Na: 0.15 meq 100 g−1; B: 0.07 mg kg−1; Zn: 2.90 mg kg−1.

Fig. 1
figure 1

Records of relative humidity (%), rainfall (mm), maximum and minimum air temperature (°C) (A), and solar radiation (W m−2) (B) during the experiment period (from august 2018 until august 2020). Data were obtained from the nearest meteorological station (Chaparral (latitude 3°37′35.193″ N, longitude 75°32′35.483″ W, altitude: 914 masl)) of Meteoblue weather

Full size image

2.2 Characteristics and Dose of Biochar

The BC was applied 4 months after the last harvest of the coffee crop (August 2018) at four different doses. The BC treatments were 0, 4, 8, and 16 t ha−1 (BC0 = 0.0 kg/tree; BC4 = 0.75 kg/tree; BC8 = 1.50 kg/tree, and BC16 = 3.00 kg/tree). The doses were determined based on the physiological responses observed in other crops (Sánchez-Reinoso et al 2020). BC was incorporated at 30 cm from the base of the coffee tree stem and was subsequently covered with a layer of soil. The BC used in this study was manufactured from coffee pulp that was subjected to an intermediate pyrolization of 500 °C for 20 min in a rotary kiln (6 m long × 0.7 m internal diameter) (Tecsol, Bogotá, Colombia). The characteristics of the BC obtained were as follows: pH: 9.42; electrical conductivity: 19.4 dS/m; ECEC: 103 cmol+/kg; ashes: 20.8%; organic carbon (OC): 46.4%; OC/N: 16.5; N: 2.81%; P: 1.11%; K: 4.37%; Ca: 0.97%; Mg: 0.43%; Cu: 53.5 mg kg−1; Fe: 2,795 mg kg−1; B: 99.3 mg kg−1; Zn: 110 mg kg−1.

2.3 Level of Fertilization

The treatments of CF levels were carried out in a fractional manner (two applications per evaluation year) according to the recommendations of Sadeghian (2013). Fertilization was performed using a simple 46N fertilizer (Urea, Yara, Colombia) and a 17N–6P–18 K–2 Mg compound fertilizer (Nutrimon–Café Producción, Monómeros S.A., Colombia). The fertilization levels were as follows: (a) 0% (0 kg ha−1 N, 0 kg ha−1 P2O5, 0 kg ha−1 K2O, 0 kg ha−1 MgO-S-B and 0 kg ha−1 Zn); (b) 33% (38 kg ha−1 N, 8 kg ha−1 P2O5, 24 kg ha−1 K2O, 3 kg ha−1 MgO-S-B and 0.3 kg ha−1 Zn); (c) 66% (75 kg ha−1 N, 16 kg ha−1 P2O5, 48 kg ha−1 K2O, 5 kg ha−1 MgO-S-B and 0.5 kg ha−1 Zn); (d) 100% (114 kg ha−1 N, 24 kg ha−1 P2O5, 73 kg ha−1 K2O, 8 kg ha−1 MgO-S-B, and 0.8 kg ha−1 Zn) of the nutritional requirements. The abovementioned fertilization treatments were provided with the following commercial fertilizers: F0 = 0 g/tree 17N-6P-18 K-2 Mg and 0 g/tree urea; F33 = 30 g/tree 17N-6P-18 K-2 Mg and 7 g/tree urea; F66 = 60 g/tree 17N-6P-18 K-2 Mg and 14 g/tree urea, and F100 = 90 g/tree 17N-6P-18 K-2 Mg and 22 g/tree urea. The doses were determined based on the physical and chemical analysis of the soil and according to the nutritional requirements reported by Salamanca-Jimenez (2017). Fertilizers were applied every 6 months (May and October for each year of evaluation).

The experimental treatments were arranged in a split-plot design in completely randomized blocks where the main plots corresponded to the BC doses (0, 4, 8, and 16 t ha−1), and the subplots corresponded to the CF levels (0, 33, 66, and 100% of the nutritional requirements). However, the data of the registered variables were analyzed in two ways: (i) the soil physical variables were analyzed using a completely randomized block design since we wanted to know the direct effect of the different BC doses (0, 4, 8, and 16 t ha−1), and (ii) the soil chemical variables and microbial respiration were analyzed using a split-plot design. The experimental unit was made up of five trees surrounded by 12 guard trees and each treatment was repeated four times (four blocks), using 320 trees in the development of the entire experiment.

2.4 Soil Physical Properties

At the end of the experiment (August 2020), soil samples were collected from the biochar treatments that did not have fertilizer application (NF 0%) to estimate soil physical variables such as bulk density (ρa), aggregate stability, soil water holding at saturation point, and field capacity.ρa was determined using the cylinder method described by Pla (1983).

$${\uprho }_{a}=\frac{a}{V(1+w)}$$
(1)

where ρa is the bulk density; \(a\) is the weight of the soil extracted from the hole in the field; V is the volume of the metal cylinder, and w is the percentage moisture obtained from the oven-dried sample.

The estimation of water-stable aggregates was carried out by the Yoder wet sieving method (Yoder 1936). The percentage of water-stable aggregates was determined using 130 g of aggregates from dry soil at room temperature of equivalent diameter between 6 and 8 mm. The aggregates were subjected to upward and downward movements in a mechanical agitator using a set of sieves with 6.30, 4.00, 2.00, 1.00, 0.50, and 0.25 mm mesh openings. Additionally, the mean weight diameter (MWD) of the aggregates sieved in water was estimated according to the methodology described by Kemper and Rosenau (1986), using Eq. (2).

$$MWD= \sum_{i=1}^{n}{D}_{i} x {P}_{i}$$
(2)

Soil water holding at the saturation point (SP), field capacity (FC), and permanent wilting point (PWP) were estimated through the pressure cooker methodology. Soil samples were collected from each treatment directly in the field using cylindrical metal rings (5 cm diameter and 2 cm height). The samples in the cylinders were saturated with water by capillary action (SP) and then subjected to − 33 kPa (FC) and − 1500 kPa (PWP). The water content at SP, FC, and PWP was determined using Eq. (3):

$$WC \left(\%\right)=\left(\frac{{M}_{ws}- {M}_{ds}}{{M}_{ds}}\right)x 100$$
(3)

where WC is the water content; Mws is the mass of wet soil, and Mds is the mass of dry soil at 105 °C (24 h).

Finally, the contents of clays (Cl), silts (Si), and sands (Sa) were determined by densimetry using the Bouyoucos method (Bouyoucos 1927) by dispersion with sodium hexametaphosphate and the USDA textural classification triangle.

2.5 Soil Chemical Characteristics

Samples of 1.5 kg of soil were used for the analysis of chemical variables and nutrient content. The material was collected 1 month before the harvest (the second week of March) in the same area of trees used in each evaluation year. A composite sample of the three central trees of each experimental unit was collected, taking soil from the first 25 cm deep at 30 cm from the base of the coffee tree stem. The soil samples were processed at the Water and Soil Laboratory of the Facultad de Ciencias Agrarias, Universidad Nacional de Colombia, Bogotá. The pH was recorded by potentiometric determination according to the soil:water suspension method in a 1:1 w/v ratio. The OC was determined using the Walkley–Black colorimetric method, and the estimated nitrogen (Nest) was calculated from the OC using the factor 0.0862.

The total major element bases (Ca, K, Mg, and Na) were determined by atomic absorption using the extraction method with 1 M ammonium acetate at pH 7. The exchangeable acidity (EA) was measured by volumetric determination through the extraction method in 1 M KCl. The effective cation exchange capacity (ECEC) was estimated by the sum of the bases and exchangeable acidity. The amount of available phosphorus (P) was measured by colorimetric determination with the Bray II method. The microelements (Cu, Fe, Mn, and Zn) were estimated by atomic absorption through the DTPA extraction method. Boron (B) was quantified by colorimetric determination according to the method of extraction with monocalcium phosphate.

2.6 Soil Microbial Respiration

Soil basal respiration was quantified using the modified closed flask technique (Alef and Nannipieri 1995). A subsample of 100 g of soil from each experimental unit was passed through a 2-mm sieve. The samples were incubated at a temperature of 25 °C for 24 h in the dark in hermetically sealed 2-L glass containers. Inside each container, three beakers were placed as follows: (i) a beaker with 15 mL of a 0.05 M NaOH solution storing the respired C; (ii) a beaker with 15 mL of distilled water to maintain the humidity in the container; and (iii) a beaker with 30 g of soil sample previously sieved. Additionally, three flasks were used as blank, which did not contain soil samples. After the incubation time had elapsed, the beaker with NaOH was extracted, and the reaction was stopped by adding 3 mL of a 0.5 M BaCl2 solution to precipitate the C present. Subsequently, three drops of a titration indicator (0.5 g phenolphthalein dissolved in 50 mL distilled water and 50 mL ethanol) were added, and titration was carried out with a 0.05 M HCl solution until color change. The rate of soil respiration in mg of CO2 per soil dry weight in the incubation time in hours (CO2 (mg)/SW/t) was determined according to Eq. (4).

$${CO}_{2}\left(mg\right)=\frac{\left({V}_{0}- V\right) x 1.1 }{dwt}$$
(4)

where SW is the amount of soil in dry weight; t corresponds to the incubation time in hours; V0 is the volume of HCl used for the titration of the blank; V is the volume of HCL used for the titration of the soil sample; 1.1 is the conversion factor (1 mL of 0.05 M NaOH = 1.1 mg CO2), and dwt is the dry weight of 1 g of wet soil.

2.7 Data Analysis

Two separate data analyses were carried out to quantify the effect of BC on the different soil properties studied. A randomized block design with four BC doses (0, 4, 8, and 16 t ha−1) was used for the analysis of the soil’s physical variables. Additionally, an analysis using a design in split plots (BC levels vs. CF levels) was established in random blocks for the chemical variables and soil microbial respiration. The results, with values expressed as percentages, were transformed into arcsine, and an analysis of variance was then performed. When significant differences were obtained, Tukey’s comparative test of means was applied at P ≤ 0.05. A polynomial contrast analysis was also performed to identify trends in the levels of BC and CF. Data were analyzed using Statistix v 9.0 (analytical software, Tallahassee, FL, USA). Additionally, a principal component analysis (PCA) was carried out using the software InfoStat 2016 (Di Rienzo et al 2016).

3 Results

3.1 Effect of BC Application on the Physical Properties of Soil

Table 1 summarizes the effects of the different application rates of biochar (BC) on the physical properties of the soil at the end of the experiment (year 2020). In general, a positive linear trend was evidenced regarding the progressive addition of BC. An increase in BC doses caused higher MWD, aggregation status, SP, and FC. A negative linear trend was generally registered for SA, ρa, and PWP, with these variables decreasing at a higher dose of BC. Finally, no significant differences were found in the content of clays (Cl), silts (Si), and sands (Sa). The evaluated soil showed a loam texture (Lo) according to the USDA textural classification triangle.

Table 1 Physical variables of an ANDISOL SOIL treated with different doses of biochar (0, 4, 8, and 16 t ha−1) obtained from coffee pulp
Full size table

3.2 Soil Chemical Properties

Differences were only found between the BC or CF treatments on the chemical properties of the soil (Tables 2, 3, and 4). Regarding the dose of BC, three groups of variables of the nutrient content did not show significant differences: (i) Ca, K, Na, P, and Cu in both evaluation years (2019 and 2020); (ii) Mg and ECEC in 2019, and (iii) OC, Nest, Fe, and Mn in 2020. On the other hand, a positive linear trend was also observed for variables such as pH (4.96), OC (4.41%), Nest (0.38%), Zn (2.33 and) in 2019 and B (0.48 and 0.38 mg kg−1 in 2019 and 2020, respectively) when the BC application rate increased, especially in soils with 16 t ha−1 BC. A negative linear trend was observed for the EA (0 t ha−1 BC = 0.58 and 0.54 meq/100 g vs. 16 t ha−1 BC = 0.18 and 0.22 meq/100 g in the evaluation of 2019 and 2020, respectively) with the increase in the application rate of BC. On the other hand, a quadratic trend was observed for the content of Mg for the year 2020 (Table 3) and the content of Fe and Mn in 2019 (Table 4), observing the highest values, in general, with applications at 8 t ha−1 BC. Finally, a cubic trend was registered for the ECEC of the year 2020, reaching the highest records in soils with the application of 16 t ha−1 BC (11.08 cmol+/kg).

Table 2 Chemical variables and nutrients content of an Andisol soil treated with different doses of biochar obtained from coffee pulp (0, 4, 8, and 16 t ha−1) and chemical fertilization levels (0, 33, 66, and 100% of the nutritional requirements) during 2 years of experiments (2019 and 2020)
Full size table
Table 3 Nutrients content of an Andisol soil treated with different doses of biochar obtained from coffee pulp (0, 4, 8, and 16 t ha−1) and chemical fertilization levels (0, 33, 66, and 100% of the nutritional requirements) during 2 years of experiments (2019 and 2020)
Full size table
Table 4 Nutrients content of an Andisol soil treated with different doses of biochar obtained from coffee pulp (0, 4, 8, and 16 t ha−1) and chemical fertilization levels (0, 33, 66, and 100% of the nutritional requirements) during 2 years of experiments (2019 and 2020)
Full size table

Regarding the effects of fertilization levels, Fe did not show significant differences with the gradual application of fertilizers in 2020. However, a positive linear trend was observed with the progressive increase of fertilizers in all the chemical variables studied in both years of evaluation (2019 and 2020). However, the OC in 2019 showed a quadratic trend and the Mg of the year 2020 obtained linear and quadratic trends (Table 3). In summary, the application of fertilizers generated an increase in the values of the different chemical properties in both years of evaluation, especially when the level of fertilization was 100% of the nutritional requirements (pH: 4.88 and 4.84; OC: 4.27% and 4.52%; Nest: 0.38% and 0.39%; Ca: 6.57 and 6.73 cmol+/kg; Mg: 3.49 and 2.70 cmol+/kg; K: 1.79 and 1.63 cmol+/kg; Na: 0.03 and 0.03 cmol+/kg; EA: 0.44 and 0.39 cmol+/kg; ECEC: 11.50 and 10.97 cmol + /kg; P: 9.89 and 9.40 mg kg−1; Cu: 1.01 and 1.06 mg kg−1; Fe: 164.23 and 217.56 mg kg−1; Mn: 42.90 and 58.32 mg kg−1; Zn: 2.46 and 3.02 mg kg−1; B: 0.47 and 0.43 mg kg−1 in 2019 and 2020, respectively).

3.3 Microbial Respiration and Principal Component Analysis Biplot

Soil microbial respiration showed differences in the BC × CF interaction. In general, microbial respiration was higher in the soils of the coffee crop treated with chemical fertilization compared to the soil without fertilizers (Fig. 2). The increasing addition of BC also generated an increment in this variable, mainly in soils with fertilizers at 66 and 100% of the nutritional needs of coffee trees, reaching values of ~ 187.6 µg CO2 g−1 dry soil day−1 at a dose of 8 t ha−1 in 2019 and ~ 146.4 µg CO2 g−1 dry soil day−1 at a dose of 16 t ha−1 in 2020, respectively (Fig. 2 A and B). However, the application of 16 t ha−1 in 2019 showed similar values in all levels of fertilization compared to the dose of 0 t ha−1 (between ~ 95 and ~ 105 µg CO2 g−1 dry soil day−1).

Fig. 2
figure 2

Soil respiration of an Andisol soil treated with different doses of biochar obtained from coffee pulp (0, 4, 8, and 16 t ha.−1) and chemical fertilization levels (0, 33, 66, and 100% of the nutritional requirements) in 2019 (A) and 2020 (B). Points represent ± standard error (n = 4). Bars represent the mean of four blocks ± standard error. Equal letters indicate that the means are not statistically different according to Tukey’s test at P ≤ 0.05

Full size image

The PCA shows that the treatments are represented by points and the variables by vectors (Fig. 3). In general, PCA1 and PCA2 represented 54.5% and 19.2% of the variation of the different parameters analyzed for the year 2019 (Fig. 3A) while for the year 2020, PCA1 and PCA2 represented 59.7% and 12.1% of the variation (Fig. 3B), respectively. The vectors of Mg, Zn, K, Na, Ca, ECEC, P, Nest, pH, and B have angles close to the origin, showing a high correlation between the soil variables evaluated for the year 2019. The plants with 16 t ha−1 BC in 100% CF were located in the extreme right sector of the biplot analysis (group V). On the contrary, the application of 0 t ha−1 BC and 0% CF (group I) was located in the sector opposite to group V, evidencing a negative effect of the lack of BC and CF applications on the analyzed variables. Three differential effects were also found in the BC and CF doses in the soil of the coffee crop: (a) 0 t ha−1 BC + 33% and 66% CF; 4 t ha−1 BC + 0% and 33% CF; 8 t ha−1 BC + 0% CF (group II) behaved with the same trend as group I; (b) 16 t ha−1 BC + 0%, 33% and 66% CF (group III) showed a less negative effect on the soil variables studied. Finally, (c) soils with 0, 4, and 8 t ha−1 BC + 100% CF; 8 t ha− BC + 33% and 66% CF (group IV) registered a better response showing a trend similar to that of group V.

Fig. 3
figure 3

Principal component analysis (PCA) biplot of different variables of soil for coffee (Coffea arabica L.) cultivation under different doses of biochar and fertilization levels. Respiration, microbial respiration of the soil; pH, soil pH; ECEC, effective cation exchange capacity; EA, exchangeable acidity; nutrient content in the soil; OC, organic carbon; Nest, estimated nitrogen; Ca, calcium; Mg, magnesium; K, potassium; Na, sodium; P, phosphorus; Cu, copper; Fe, iron; Mn, manganese; Zn, zinc; and B, boron. B0, biochar dose of 0 t h−1; B4, biochar dose of 4 t h−1; B8, biochar dose of 8 t h−1; B16, biochar dose of 16 t h.−1. F0, 0% of the nutritional requirements; F33, 33% of the nutritional requirements; F66, 66% of the nutritional requirements; F100, 100% of the nutritional requirements. A Evaluation period of the year 2019; B evaluation period of the year 2020

Full size image

In 2020, the vectors of Ca, ECEC, K, Zn, Na, Mn, pH, Cu, Nest, B, and P registered angles close to the origin, showing a high correlation between the soil variables analyzed for the year 2020 (Fig. 3B). Similar trends were identified in groups I and V, which were represented by the same treatments observed in 2019. Three differential responses were obtained by the BC and CF doses in the soil of the coffee crop: (a) 0 t ha−1 BC + 33% and 66% CF; 4 and 8 t ha−1 BC + 0% CF (group II) behaved in the same direction as group I; (b) 0 t ha−1 BC + 100% CF; 4, 8 and 16 t ha−1 BC + 33% CF; 4 and 8 t ha−1 BC + 66% CF; 16 t ha−1 BC + 0% and 33% CF (group III) represented a less negative effect on the soil variables evaluated. Finally, (c) soils with 4 and 8 t ha−1 BC + 100% CF and 16 t ha−1 BC + 66% CF (group IV) obtained a better response represented by a trend similar to that of group V. The PCA showed that the treatment of 16 t ha−1 BC and 100% CF had a similar trend in both years of study (2019 and 2020), indicating a greater positive effect on the chemical variables and microbial respiration of the soil.

4 Discussion

The application of BC has been documented as an alternative to improve the physical, chemical, and biological properties of the soil (Al-Wabel et al. 2017). In this study, BC applications to the soil of a commercial coffee farm, mainly at doses of 8 or 16 t ha−1, favored certain physical properties (stable aggregates, bulk density (ρa), water content at saturation point, and field capacity). Additionally, the same doses combined with chemical fertilization levels of 66 or 100% of the nutritional requirements favored some chemical and biological variables of the studied soil (pH, OC, and soil microbial respiration). Other studies also showed that soil with the addition of BC (at doses between 3 and 300 t ha−1) registered higher pH values, improved its water-holding capacity, and increased the organic matter content and nutrient availability (nitrogen, phosphorus, and potassium) (Zhang et al 2015; Mahmood et al 2017). The interaction between the application of BC and CF has also been shown to favor N uptake and cause greater efficiency of the chemical fertilizer and the reduction of nutrient leaching (Sadaf et al 2017; Saha et al 2019; Zhang et al 2019).

BC is known to affect the physical and hydraulic properties of the soil (Razzaghi et al. 2020; Blanco-Canqui 2017). In this study, the applied BC caused two different effects on those properties. A gradual increase in the water holding capacity, MWD, and stable aggregates and a drop in the bulk density (ρa) of the soil at higher BC doses (4, 8, and 16 t ha−1). Herath et al. (2013) also observed that the use of between 10 and 17 t ha−1 BC (of corn residues) caused greater water-holding capacity of the soil due to higher mesoporosity (pores with diameters between 2 and 50 nm). Additionally, Obia et al. (2016) observed that applications between 2 and 6 t ha−1 BC (produced from corn cobs) increased aggregate stability and reduced the ρa in sandy loam soil as a consequence of the higher porosity. Soil bulk density is one of the most studied properties with the use of BC, with a gradual decrease in ρa with an increase in the amount of BC used (Blanco-Canqui 2017). Verheijen et al. (2019) report that the use of BC modifies the soil ρa mainly because small BC particles are located in the soil pores, generating a reduction in the macropores between aggregates. This would imply greater soil volume due to the presence of a material with a lower density (like BC). Our findings confirm that the use of the BC can be justified because a low ρa and high stable aggregates indicate that the use of coffee pulp is a tool for improving soil health since biochar-amended soils can be more friable and less compatible than soils without biochar (Blanco-Canqui 2017).

Positive responses of soil chemical variables due to the application of BC have also been reported (Singh et al 2022). Lower EA and higher pH, OC, and ECEC were recorded in soils treated with 8 and 16 t ha−1 BC. Yuan and Xu (2012) stated that the application of 1% (w/w) of different BC (obtained from various crop residues) increased the pH (between 12 and 32%) and CEC (between 13 and 20%) of the soil. The changes observed in soil pH and CEC were mainly associated with the pH of the BC (values between 8.88 and 10.35) and its base cation content, suggesting that the incorporation of BC helped to improve the soil’s acidity and fertility. Jien and Wang (2013) recorded higher pH, CEC, and OC in silty clay soil when using between 2.5 and 5.0% w/w BC of Leucaena leucocephala wood residues. Additionally, Tanure et al. (2019) report that BC improves the exchangeable acidity and soil fertility due to the presence of hydroxides and oxides of the alkali metals in BC ashes that dissolve easily and increase pH. Also, because of the carboxyl, phenolic, and hydroxyl groups of the negatively charged BC that bind to H+ ions that alkalize the pH and reduce its activity in the soil solution. Finally, a plausible explanation for the observed relationship between increasing pH and CEC for this soil type can be that the higher pH of acidic and slightly acidic soils after the addition of biochar might cause the deprotonation of functional groups from minerals, such as kaolinite, which results in the generation of more negative charges that contribute to a higher CEC (Premarathna et al 2019).

Nutrient availability has been positively affected in soils treated with BC (Gupta et al 2020; Karimi et al 2020). The above was recorded in this study, where soils treated with 8 and 16 t ha−1 BC and the application of 66 and 100% CF showed higher values in the nutrient content. Similar trends were also found by Puga et al. (2020), who showed increases in the availability of N in soil treated with BC (obtained from wood residues of eucalyptus wood) at an application rate of 15 g kg−1 of soil. Usman et al. (2016) also observed that applications between 4 and 8% BC (obtained from Conocarpus wood residues) increased the concentration of P, K, Fe, Mn, Zn, and Cu in sandy soil. Those authors state that the use of BC generates greater nutrient availability in the soil due to the following: (i) the increase in pH (liming effect); (ii) the fact that BC acts as a fertilizer, improving the physicochemical properties of the soil; (iii) the increase in the soluble and/or retained amounts of nutrients in response to the charge and surface area of BC; and (iv) the contribution of nutrients from BC in the form of ashes, generating an increase in the content and uptake of plant nutrients.

In this study, the microbial respiration of soil was also conditioned by the BC treatments, mainly with 8 and 16 t ha−1 BC at 66% CF. Similar results were also observed by Karimi et al. (2020) in low-fertility soils treated with 2% w/w BC (obtained from corn residues), which significantly increased respiration and soil microbial biomass. However, contrasting results were reported by Ge et al. (2020), who found that applications of BC (5 and 10 t ha−1) manufactured with bamboo branches in combination with nitrogen fertilization (50 kg ha−1 of NH4NO3) generated a reduction in soil respiration because of a compensation effect of the addition of N, resulting in greater availability of this nutrient. Additionally, Zhou et al. (2017) report that BC applications generally generate increased soil respiration due to the primary effect of BC on the decomposition of C from the soil through abiotic or biotic means, and the mineralization of more labile C that would stimulate the renewal of soil organic matter.

This study shows a series of advantages for the management of the nutrition of the coffee production system. It is important to indicate that the application of BC to the soil in combination with chemical synthesis fertilizers is an interesting alternative for the nutrition of coffee as it has been used in other species to improve their growth conditions (El-Naggar et al 2019). Additionally, this research suggests that the use of BC can improve plant physiology due to a direct effect such as better water holding in the soil (higher water content at field capacity). These results provide information to recommend the transformation of coffee crop residues, such as fresh pulp, through the pyrolysis process to obtain biochar and favor the recycling of nutrients in soils in coffee-growing areas. This would help the sustainability of the productive system since this amendment shows a positive impact on the plant-soil relationship (Ayaz et al 2021).

5 Conclusion

It is important to study alternative technical agronomic strategies to increase crop yield or reduce the production cost in middle-income countries in which biochar (BC) can offer several potential co-benefits at a relatively low cost. In the present study, the use of BC from the coffee pulp (of 8 and 16 t ha−1) generated a positive effect on the behavior of the different properties of soil for coffee production. In addition, our results found that the combination of 66 and 100% chemical fertilization and 8 and 16 t ha−1 BC favored the pH and microbial respiration of the soil. Our findings enlarge the knowledge how the BC improves soil quality, mainly, in tropical soils. Also, this research suggests that the use of BC manufactured with coffee pulp (applications between 8 and 16 t ha−1) can be an alternative to improve the quality of soils of coffee-growing areas and help in the management of solid residues. Finally, BC also provides a complementary option for the nutrition of this crop that would allow reducing the levels of chemical fertilization applications.