INTRODUCTION

Agricultural activities generate large amounts of organic waste (OW), which when not properly managed can cause salinization of soils, leaching of nitrates and phosphates to aquifers (^{Flotats and Solé, 2008}), accumulation of lignins, aromatic oils and resins in soil (^{Achten and Hofmann, 2009}), spread pests, weeds, diseases, and phytotoxicity (^{Fornes et al., 2012}). In order to reduce environmental risks, it has been considered that OW must be processed through controlled decomposition techniques (^{Bernal et al., 2009}).

Composting (CP), vermicomposting (VC) and semicomposting (SC) have proved to be useful systems for obtaining products that can be used in agriculture (^{Fornes et al., 2012}). The degradation of OW by these processes allows the formation of stable polymerized molecules (^{Bernal et al., 2009}), but each system of controlled decomposition produces organic fertilizers with profoundly variable characteristics (^{Fornes et al., 2012}).

Composting is the transformation of OW by the action of microorganisms. In the first stage, simple carbon compounds are mineralized and metabolized by thermophilic microorganisms, releasing carbon dioxide, water, ammonia and organic acids, reaching temperatures from 45 to 70 °C (^{Fornes et al., 2012}), which destroys undesirable seeds and pathogenic organisms. In the mesophyll stage, humification of the remaining organic compounds occurs (^{Fornes et al., 2012}). The degree of stability and maturity reached reflects the rate of OW decomposition and phytotoxicity removal (^{Ocaña et al., 2015}).

Vermicomposting is a non-thermophilic process of bio-oxidation and stabilization of organic substrates by the joint action of earthworms and mesophilic microorganisms (^{Bernal et al., 2009}). The metabolism of the annelid generates an organic substrate, in part from the microorganisms of the intestine (^{Fornes et al., 2012}).

Semicomposting is a degradation process that handles volumes of OW lower than that recommended for CP and therefore does not present a thermophilic stage. In this system mesophilic microorganisms are responsible for the degradation of organic matter (OM) (^{Herrera-Ortíz et al., 2015}); however, SC has been poorly documented.

The quality of the final products has been evaluated for safety in agricultural through stability criteria: temperature, C/N ratio, pH, enzymatic activity and phytotoxicity of bio-transformed OM (^{Bernal et al., 2009}). However, the work generated is not conclusive, so it is recommended to evaluate different residues and decomposition systems (^{Ocaña et al., 2015}).

The objective of the present study was to compare, throughout the period of decomposition, physical, chemical and enzymatic parameters of OW subjected to three decomposition systems for the production of organic fertilizers. The results of this study will contribute to the selection of the most suitable OW decomposition system for the producer, according to the availability of resources such as organic material, water or earthworms.

MATERIALS AND METHODS

The study was carried out in the facilities of the Universidad Autónoma de Chihuahua, Chihuahua, Mexico. The duration of the experiment was 183 d, throughout April to October 2014. A mixture of bovine manure and sawdust was subjected to three decomposition systems: Composting (CP), semicomposting (SC) and vermicomposting (VC). For the CP, pallet boxes of a capacity of 1110.9 L, for SC and VC 60 L plastic containers were filled, and for VC, 10 adult Californian red earthworms (Eisenia foetida [Savigny, 1826]) were added per liter of substrate at 6 d after the experiment (DAE) (^{Castillo et al., 2010}). Irrigation maintained moisture between 50% and 60% in CP and SC, and between 75% and 85% for VC (^{Castillo et al., 2010}), which was determined by touch test (^{Soto and Muñoz, 2002}). The CP and SC mixtures were manually removed every 28 d.

Temperature of decomposing residues was daily recorded using a maximum and minimum thermometer (Fisher Scientific, Pittsburgh, Pennsylvania, USA). A pistil thermometer with a minimum range of -20 °C and maximum of 110 °C was used. Nine samples were taken at 4, 35, 68, 98, 127, 141, 155, 169 and 183 DAE. A portion of the sample was shade dried at room temperature, sieved in nr 20 mesh (1 mm diameter) for chemical analysis. For the determination of enzymatic urease activity (EUA), fresh samples were reserved at 4 °C. The following analyses were carried out: total nitrogen (Nt) by the Micro-Kjeldahl method (^{Fernández et al., 2006}); organic carbon (OC) with the Walkley-Black dichromate oxidation method (ASTM International, 2010); pH with potentiometer (HI 2210 pH Meter, Hanna Instruments, Woonsocket, Rhode Island, USA) in dilution with 1:10 (w/v) deionized water; EUA was determined using the method of ^{Kandeler et al. (1999)}. From 141 and up to 183 DAE, electrical conductivity (EC) was evaluated using a conductivity meter (Solubridge Beckman, Cedar Grove, New Jersey, USA) in dilution with 1:10 (w/v) deionized water, and phytotoxicity with the mentioned germination index (GI) by ^{Zucconi et al. (1981)}.

The experiment consisted of three treatments with three replicates and three analytical replicates per sample. A split plot design was used with observations over time, and the main plot effect was defined by OW decomposition system. ANOVA was performed using the statistical software SAS version 9.0 (SAS Institute, Cary, North Carolina, USA; ^{Stokes et al., 2012}) and the test of comparison of means implemented in Pro Mixed was used.

RESULTS AND DISCUSSION

Temperature revealed differences (P < 0.05) between treatments with time (Figure 1). In CP the thermophilic phase was presented from the fourth DAE to the 85 DAE, at which time the temperature dropped and the cooling and stabilization phase of OW started. The temperature observed in the CP system agree with that found by ^{Ocaña et al. (2015)}, who evaluated composting with municipal bio-waste and wood ash, finding that the temperature reached 70 °C in the thermophilic phase from days 9 to 12. Likewise ^{Bustamante et al. (2008)} reported that in decomposing bovine manure, the thermophilic curve was detected from days 7 to day 84. The rise in temperature per day 100 DAE in this experiment coincided with that reported by ^{Boulter-Bitzer et al. (2006)}, who at 116 d, after intense work of turning, observed elevation of temperature up to 60 °C.

Vermicomposting (―), semicomposting (---) and composting (-(-(-). Reference value (((((). ˃ 45 °C thermophilic phase and ˂ 45 °C mesophilic phase (^{Jiménez et al., 2008}).

Figure 1: Variation over time of the maximum and minimum ambient temperature and temperature of organic waste subjected to three decomposition systems. 

In SC the influence of the ambient temperature on the temperature of the transformed OW was observed, registering in a range of 18 to 34 °C. The thermophilic phase was not presented, which coincides with other studies in which internal temperature was related to the available surface and the total volume of the residues (^{Herrera-Ortíz et al., 2015}). In VC no thermophilic temperatures were present and the recorded temperature ranged from 18 to 34 °C. This coincides with ^{Castillo et al. (2010)}, who worked with dairy cattle manure and pine sawdust and recorded a range of 20 to 40 °C. In VC systems a controlled temperature interval of 20 and 30 °C is fundamental for the survival of the worm and its efficiency to process the residues (^{Sinha et al., 2009}). In this experiment, an average of 26.3 °C was recorded, the optimal temperature for worm development.

Percentages of Nt showed differences between time and treatments (Figure 2a) with tendency to increase. Differences in Nt during the first stages of the study could be related to the lack of homogeneity of the mixtures. In VC and CP the Nt increased at 68 DAE (1.76 ± 0.05% and 1.71 ± 0.13%, respectively), while in SC an increase in Nt was observed at 35 DAE (1.70 ± 0.02%). In all three treatments Nt had increased by 183 DAE, which coincides with ^{Sangamithirai et al. (2015}) who found an increase in Nt at the end of the SC process with OW from garden and tea leaves, as well as with ^{Fornes et al. (2012)} who recorded increases of Nt in VC and CP.

Vermicomposting (―), semicomposting (---) and composting (-(-(-). Reference value (((((). a) Nt 1.2 to 2.0%* (^{Romero, 2004}); b) OC < 30%* (Romero, 2004); c) C/N ratio < 15 (^{Jiménez et al., 2008}); (e) EUA < 200 μmol N-NH₄ ⁺ g^-1* (^{Liu et al., 2011}), (*) based on DM.

Figure 2: Total nitrogen (Nt) organic carbon (OC), C/N ratio, pH and enzymatic urease activity (EUA) with time in organic waste under three decomposition systems. 

Organic C decreased 23.55 ± 0.32% in CP after 35 DAE, while in VC it was 68 DAE (21.24 ± 0.6%) and in SC 23.92 ± 2.41% at 98 DAE. There were increases in OC in VC and SC at 155 and 183 DAE respectively, and in CP 21.55 ± 3.08% at 141 DAE (Figure 2b).

In this study, in CP the largest OC decrease corresponded to 37.3% during the thermophilic phase, which is in agreement with that found by ^{Bustamante et al. (2008)}. In VC the percentage of OC dropped sharply by 68 DAE and the subsequent decrease was gradual, which coincides with ^{Garg and Gupta (2011)} for the transformation of plant residues and cow manure with E. foetida. The gradual degradation of OC in SC coincides with that mentioned by ^{Sangamithirai et al. (2015)}, when studying gardening and paper residues.

Carbon/Nitrogen ratio decreased sharply in all three treatments (Figure 2c), but showed differences between time and treatments. It has been claimed that when OW are transformed, the C/N ratio decreases with time by the C loss as carbon dioxide (^{De Guardia et al., 2010}). In CP and SC the C/N ratio decreased at 35 DAE (14.62 ± 1.32 and 15.68 ± 1.18, respectively). For VC, a significant decrease was observed at 68 DAE (15.56 ± 1.5). A similar result was recorded by ^{Hernández-Rodríguez et al. (2012)}, when evaluating the decomposition of sheep manure and sawdust reported an increase in the C/N ratio at 56 d and a steady decline up to 112 d. In this study, the C/N ratio was equal to or less than 15/1 for the three treatments, and there were no differences after 127 DAE in any of the treatments suggesting the stabilization of OM at 68 DAE according to ^{He et al. (2014)}.

Initial pH values were between 8.57 and 8.94 in all three treatments (Figure 2d). There was a difference between VC (8.94 ± 0.05) and SC (8.57 ± 0.14) at the fourth DAE, and a tendency to acidification to 35 DAE was observed, being greater in VC (7.72 ± 0.15), which coincides with other authors (^{Garg and Gupta, 2011}). In CP the initial pH was 8.9 ± 0.2 and increased to 9.5 ± 0.27 by 127 DAE. Alkalization is a consequence of the mineralization of proteins, amino acids, peptides and their transformation to ammoniacal compounds (^{Bustamante et al., 2008}). In other compost experiments with bovine manure, it was observed that the basic pH could be related to the frequency and type of activities such as churning, humidification and leachate of the material (^{Gómez-Brandón et al., 2008}). ^{López-González et al. (2014)} recorded a slightly alkaline final pH when using residues with high levels of cellulose.

Differences in both time and treatments were observed in the EUA (Figure 2e). In this study, all three treatments observed a peak of EUA at 35 DAE and a decrease at 68 DAE. The VC and SC EUA stability was recorded at 68 DAE (409.32 ± 58.75 and 437.87 ± 36.09 μmol N-NH₄ ⁺ g^-1, respectively), whereas CP presented a higher EUA 756.05 ± 270.14 μmol N-NH₄ ⁺ g^-1 at 127 DAE, coinciding with the establishment of the mesophilic phase. This datum coincides with that mentioned by ^{Castaldi et al. (2008)}, that during the composting of municipal solid waste, a rise of urease from the first 3 wk and a peak of EUA at day 56 is explained by the presence of nitrates. ^{Liu et al. (2011)} when evaluating CP of cow dung and rice stubble concluded that the decrease in EUA may be a good indicator of compost stability.

Germination index has been used as an evaluation of OW maturity and phytotoxicity (^{Hernández-Rodríguez et al., 2012)}. The changes recorded in the phytotoxicity test, for all three treatments, showed a GI greater than 96% (Figure 3a). The GI presented a similar gradual decrease for SC and CP; however, for VC an increase of the GI was observed towards 169 DAE (132.78 ± 5.46%), followed by a subsequent decrease at 189 DAE (100.26 ± 9.34%). ^{Sangamithirai et al. (2015)} reported that the GI of different mixtures of OW transformed under SC during 90 d was less than 80%, and therefore considered that their direct application to the crops could be phytotoxic.

Vermicomposting (―), semicomposting (---) and composting (-(-(-). Reference value (((((). a) GI ˃ 80% (^{Zucconi et al., 1981}), and b) EC < 3.0 dS m^-1* (^{Soumaré et al., 2002}), (*) based on DM.

Figure 3: Germination index (GI) and electrical conductivity (EC) with time in organic waste under three decomposition systems. 

Electrical conductivity changes in the OW under the three treatments were different (VC, 0.83 ± 0.06; SC, 0.60 ± 0.02; CP, 2.34 ± 0.44 dS m^-1) at 141 DAE (Figure 3b). A decrease in EC from 155 DAE was observed for the three decomposition systems. The EC observed in VC coincided with ^{Lazcano et al. (2008)} and with ^{Sangamithirai et al. (2015)} for SC. This decrease was due to the volatilization of ammoniacal compounds, the fixation by N-assimilating microorganisms or the precipitation of mineral salts, which has been related to degradation of OM and loss of mass (^{Bustamante et al., 2008)}.

At 183 DAE the transformed OW presented the characteristics recorded in Table 1. Ranges reported as acceptable for agricultural use are presented in Figures 2 and 3. The three products of the systems studied presented Nt contents within the range suitable for agricultural use according to ^{Romero (2004)}, however, there was a difference between CP 1.72 ± 0.11% and other systems. The percentage of final OC did not represent difference between VC 23.23 ± 0.59% and SC 23.20 ± 0.54%. The three products reached the recommended range of OC for agricultural use according to ^{Romero (2004)}. The final C/N ratio in the three treatments was VC, 12.91 ± 1.05; SC, 12.36 ± 0.56 and CP 11.17 ± 1.32, values considered within the range of fertilizers suitable for use according to ^{Jiménez et al. (2008)}. In the case of pH there were differences between the three products; OW submitted to VC and SC recorded pH values within the range recommended for agricultural use (7.46 ± 1.01 for VC and 7.94 ± 1.7 for SC) according to ^{Jiménez et al. (2008)}. The EC at 183 DAE in all three systems was within the range considered non-phytotoxic (0.71 ± 0.10, 0.66 ± 0.11, and 1.9 ± 0.03 dS m^-1 for VC, SC and CP, respectively) according to ^{Soumaré et al. (2002)}. EUA values in VC (300.77 ± 47.51 μmol N-NH₄ ⁺ g^-1) and CP (307.63 ± 7.73 μmol N-NH₄ ⁺ g^-1) showed no difference, while in SC (196.57 ± 7.73 μmol N-NH₄ ⁺ g^-1) had lowest value recorded being the closest to what is recommended for agricultural use, with less than 200 μmol N-NH₄ ⁺ g^-1 according to ^{Liu et al. (2011)}. For GI, the final values did not present significant differences between treatments, being greater than 80% recommended for agricultural use (^{Zucconi et al., 1981}).

Table 1: Characterization of the products of the three decomposition systems at 183 d. 

Nt: Total N, OC: organic C, EC: electrical conductivity, EUA: enzymatic urease activity, GI: germination index, VC: vermicomposting, SC: semicomposting, CP: composting, (*) based on DM.

Different letters in the same column indicate significant differences according to test of comparison of means implemented in Pro Mixed (P < 0.05).

CONCLUSION

Under the conditions of this study, all the products of the three systems obtained at 183 d were within the parameters to be considered as organic fertilizers; however, the semicompost product transformed into an organic fertilizer fastest. All three systems reduced the C/N ratio to less than 15/1. Compost product had higher pH, electric conductivity, and enzymatic urease activity values than both vermicompost and semicompost. Vermicompost product had the highest germination index values, while semicompost showed a lower enzymatic urease activity and higher total N in a shorter time, corresponding to stability indicators. It is suggested to follow up this type of work, since the variables studied may differ by decomposition system and type of organic waste available.