Waste Management xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Waste Management

journal homepage: www.elsevier .com/locate /wasman

The anaerobic co-digestion of sheep bedding and P50% cattle manureincreases biogas production and improves biofertilizer quality

http://dx.doi.org/10.1016/j.wasman.2015.08.0400956-053X/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: mssmc@ig.com.br (M.S.S.d.M. Costa).

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestion of sheep bedding and P50% cattle manure increases biogas prodand improves biofertilizer quality. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.040

Taiana Cestonaro, Mônica Sarolli Silva de Mendonça Costa ⇑, Luiz Antônio de Mendonça Costa,Marcos Antonio Teofilo Rozatti, Dercio Ceri Pereira, Higor Eisten Francisconi Lorin, Leocir José CarneiroResearch group on Water Resources and Environmental Sanitation, Western Parana State University, Agricultural Engineering Graduate Program, Rua Universitária, JardimUniversitário, 2069, 85819-110 Cascavel, Paraná, Brazil

a r t i c l e i n f o

Article history:Received 1 July 2015Revised 26 August 2015Accepted 28 August 2015Available online xxxx

Keywords:Rice huskLignocellulosic manureOrganic matterMultivariate analysisBatch system

a b s t r a c t

Sheep manure pellets are peculiarly shaped as small ‘capsules’ of limited permeability and thus aredifficult to degrade. Fragmentation of manure pellets into a homogeneous mass is important for decom-position by microorganisms, and occurs naturally by physical shearing due to animal trampling, whensheep bedding is used. However, the high lignocellulose content of sheep bedding may limit decomposi-tion of sheep manure. Here, we evaluated if co-digestion of sheep bedding with cattle manure wouldimprove the yield and quality of the useful products of anaerobic digestion of sheep bedding—biogasand biofertilizer—by providing a source of nutrients and readily available carbon. Mixtures of sheepbedding and cattle manure in varying proportions (0%, 25%, 50%, 75%, or 100% cattle manure) were addedto 6-L digesters, used in a batch system, and analyzed by uni and multivariate statistical tools. PC1, whichexplained 64.96% of data variability, can be referred to as ‘organic fraction/productivity’, because higherrates of organic fraction consumption (COD, cellulose and hemicellulose contents) led to higher digesterproductivity (biogas production, nutrient concentration, and sample stability changes). Therefore,productivity and organic fraction variables were most influenced by manure mixtures with higher(P50%) or lower (625%) ratios of cattle manure, respectively. Increasing the amount of cattle manureup to 50% enhanced the biogas potential production from 142 L kg�1 TS (0% of cattle manure) to 165,171, 160 L biogas kg�1 TS for the mixtures containing 100%, 75% and 50% of cattle manure, respectively.Our results show that the addition ofP50% cattle manure to the mixture increases biogas production andimproves the quality of the final biofertilizer.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Recent literature about the choice of bedding materials in sheepproduction systems mainly consider the aspects related to costs,animal growth, welfare, productive performance and meat quality(Wolf et al., 2010; Teixeira et al., 2013, 2014, 2015). However, thestabilization bioprocess efficiency of the bedding material afterbeing removed from the animal house is not so much studied,given its chemical and physical characteristics.

Because it is a solid waste, the sheep bedding can be stabilizedthrough the composting process, but the high content of lignocel-lulosic components makes the degradation process more difficult,making it necessary co-composting to improve the processefficiency (Costa et al., 2015).

Anaerobic digestion is another way to recycle sheep bedding.But the anaerobic digestion of sheep and goat manure requires alonger period of hydraulic retention and produces less biogas whencompared to that of other farming manure with similar chemicalcharacteristics, such as manure produced by pigs, poultry, andcattle (Orrico Junior et al., 2011). This effect is likely due to thepeculiar shape and packing of sheep and goat pellets, which needto be fragmented for improved anaerobic digestion (Amorimet al., 2004).

The use of bedding for sheep farming modifies the characteris-tics of sheep pellets, which are mechanically transformed byanimal trampling into a homogeneous mass mixed with beddingmaterial. This phenomenon is expected to improve fermentationyield by increasing the manure surface area available for microbialattachment (Motte et al., 2013; Zhang and Banks, 2013; Agyemanand Tao, 2014). However, this expected positive effect of sheepbedding on fermentation efficiency is counteracted by the factthat bedding is rich in lignocelluloses, where cellulose and

uction

Table 1Characteristics of the sheep bedding and the cattle manure used as anaerobic co-digestion substrates.

Parameter Sheep bedding Cattle manure

pH 8.81 8.97EC (mS cm�1) 7.44 5.67Carbon (%) 41.2 43.4TKN (%) 1.73 2.77Phosphorous (g kg�1) 4.55 6.54Potassium (g kg�1) 13.6 38.3Ash (%) 25.8 22.0Cellulose (%) 23.5 18.3Hemicellulose (%) 22.4 19.8HA/FA 0.95 1.34C/N 24 15

EC, electric conductivity; TKN, total Kjedahl nitrogen; HA/FA, humic acids/fulvicacids; C/N, carbon/nitrogen ratio.

2 T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx

hemicellulose are surrounded by a fermentation-resistant layer oflignin and are thus inaccessible for hydrolysis. Al-Masri (2001)reported that increasing the lignocellulosic content from olive cakemixed to sheep and goat manure decreased biogas production dueto an overall reduction in the amount of easily fermentable organicmatter (including volatile solids, neutral detergent fiber, and grossenergy).

To improve the efficiency of anaerobic digestion of sheepmanure, it is important to add an easily fermentable componentto the mixture, such as manure from cattle feedlot.

Thus, the anaerobic digestion of sheep bedding mixed withcattle manure, in a system of ‘co-digestion’, is likely to improvefermentation yield. The benefits of co-digestion include the dilu-tion of toxic compounds found in each mixture component,improved nutrient balance, and a possible synergic effect of themicrobial consortia on fermentation efficiency and yield (Khalidet al., 2011; Ashekuzzaman and Poulsen, 2011; Zheng et al.,2015; Pagés-Díaz et al., 2015; Khoufi et al., 2015).

The benefits of anaerobic co-digestion are also reflected in theagronomic quality (Albuquerque et al., 2012; Pokój et al., 2015)and stability (Molinuevo-Salces et al., 2013) of the biofertilizer,mainly on lignocellulosic material. The increase in agronomic qual-ity of the biofertilizer is related to the chemical characteristics ofthe feedstock (Tambone et al., 2010) and the transformation oflabile carbon into methane, concentrating the nutrients on thebiofertilizer (Molinuevo-Salces et al., 2013). The conversion ofthe readily degradable components directly affects the increaseof the stability of the biofertilizer (Tambone et al., 2009;Albuquerque et al., 2012.).

The aim of our study was to evaluate the anaerobic digestion ofmixtures of sheep bedding and cattle manure in varying propor-tions, to identify the mixtures most efficient at producing biogasand good quality biofertilizers.

2. Materials and methods

2.1. Manure sources

The sheep bedding used in this study was approximately45 days old and was obtained from a feedlot farm located in thecity of Cascavel (Paraná state, Brazil), where sheep were kept ona diet of forage and concentrate feed in the proportion of 70:30.Sheep bedding consisted of rice husk mixed with animal faecesand urine. Cattle manure was obtained from a feedlot in the cityof Santa Tereza do Oeste (Paraná state, Brazil), where cattle werekept on a diet of forage and concentrate feed in the ratio of60:40. Table 1 shows the characteristics of the sheep beddingand the cattle manure used in this study. The values were obtainedfrom a single composite sample analyzed in triplicate.

2.2. Fermentation setup and procedures

Fermentation was performed for 5 months (between May andOctober of 2012) at room temperature, in homemade benchtopdigesters with 6-L capacity, identical to those described by OrricoJunior et al. (2012). The average temperature during the experi-mental period was 18.4 ± 4.1 �C.

Five manure mixtures were tested, containing the followingsheep bedding to cattle manure ratios: 0:100 (CM100), 25:75(CM75), 50:50 (CM50), 75:25 (CM25), and 100:0 (CM0). For fermen-tation, adequate proportions of each manure (based on dry matterweights) were added to fermentation chambers, water was addedto the mixture for a final content of total solids of 5%, and then themanure were manually mixed to homogeneity. It was not usedinoculum in the early of the process.

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

Anaerobic co-digestion was performed in a batch system andcontinued until the biogas production curve declined. The durationof the experiment represented the time period (in days) that themanure remained inside the digester.

2.3. Experimental design

The experimental design was randomized, with five experimen-tal conditions and four replicates per experimental condition, usinga total of 20 experimental units (digesters).

2.4. Analytical methods

Immediately before digester loading and upon unloading,samples of digester content were collected for analysis and eitherkept at 4 �C or pre-dried at 50 �C in a forced air oven until constantweight, then ground using a pestle and mortar, in preparation forfiber series determination and chemical fractionation of theorganic matter. All analysis results from samples dried at 50 �Cwere corrected to dry basis (105 �C).

Electric conductivity and pH were determined from the samesample solution, using a benchtop TEC-3MP potentiometer(TECNAL, SP, Brazil) and a benchtop mCA 150 conductometer(MS Tecnopon Special Equipments, SP, Brazil), respectively. Thelevels of total solids (TS), volatile solids (VS), ash, and chemicaloxygen demand (COD) were determined according to APHA(2012) guidelines. Dissolved COD values were determined fromsample supernatants, after centrifugation at 28g for 10 min.

The carbon (C) content was determined by ignition in a mufflefurnace at 550 �C for 12 h, according to Cunha-Queda et al.(2003). The value of organic matter obtained after sample burningwas divided by 1.8 to obtain the carbon content (C).

The chemical fractionation of organic matter, as well as thedetermination of the carbon content of humic and fulvic acid frac-tions (used to obtain the HA/FA ratio) was performed according toBenites et al. (2003). For this analysis the digestate was dried at50 �C to a constant weight and then ground and sieved at 20 mesh.Next it was done the sample weighing to obtain 60 mg of totalorganic carbono. This sample was submitted to agitation in thepresence of NaOH 0.1 mol L�1 twice in succession, and the precip-itated acidified to pH 1 (± 0.1) and decanted. This supernatant aftervolume scouting for 50 mL originated the AF fraction. The resultingprecipitate of this procedure was dissolved with NaOH 0.1 mol L�1

and after measuring the volume originated the AH fraction. Carbonof the two fractions were digested with potassium dichromate andsulfuric acid and titrated with ferrous sulfate in the presence offerroin indicator. The results are given in mg of carbono in thefraction and are fixed on a dry basis (105 �C).

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040

T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx 3

The total Kjedahl nitrogen (TKN) was estimated using a Kjedahldistiller, according to Malavolta et al. (1997), and was used tocalculate the C/N ratio (C/TKN). The levels of cellulose and hemicel-lulose were estimated by determining the neutral detergent fiber(NDF) and the acid detergent fiber (ADF) content of samples, usingthe sequential method described by Campos et al. (2004) in thefiber digestion device MA-444/CI (Marconi, SP, Brazil). The dryextraction procedure described by Alcarde (2009) was used toestimate phosphorous (P) and potassium (K) levels. P reads wereperformed in a 700 Plus (Femto, Brazil) spectrophotometer,according to Malavolta et al. (1997), and K was determined byflame photometry in a DM-62 (Digimed, Brazil) photometer.

Biogas production was monitored by performing regular mea-surements of the vertical displacement of gasometers using a rulerattached to these devices. After each reading, gasometers wereemptied by opening the attached gas tap. During each reading,the pressure applied to the gas by the gasometer was measuredusing a U-tube manometer, the room temperature was determinedusing a portable digital thermometer, and these measurementswere used to re-calculate biogas volume readings for NormalTemperature and Pressure conditions (1 atm, 273 K). The valuesfor biogas production potential were calculated by dividing thetotal biogas yield of each mixture by the amount of total solidsadded (TS added), the amount of volatile solids added (VS added)or degraded (VS degraded), the amount of manure in natura(fresh manure) added (Manure), or the initial mixture volume(Substrate).

2.5. Statistical analyses

Data were analyzed by multivariate statistics (principal compo-nent analysis, or PCA, and Cluster Analysis) and by univariatestatistics (analysis of variance, or ANOVA).

Multivariate analyses were applied to all variables studied,except for pH. PCA was used to interpret the relationships betweenthe variables monitored and between these variables andthe experimental conditions tested. The data used for PCA werethe variations in anaerobic co-digestion parameters calculated asthe differences (not converted to percentages) between parametervalues in the final biofertilizer and those measured in the initialmanure mixture. The principal components (PCs) were extractedfrom the correlation matrix of the original variables, to avoid anyinterference from the measurement units of the variables studied.We selected enough PCs to explain more than 70% of data variance(Ferreira, 2011). Cluster analysis (CA) of the experimental condi-tions tested was performed by average linkage, with hierarchicalaggregation clustering based on the matrix of Euclidean distances.All observations were normalized to a median of zero and avariance of 1. Both multivariate techniques were applied to allthe variables observed except pH.

ANOVA was used to compare biogas production from the differ-ent manure mixtures. If any of the experimental conditions wasfound to influence the response variable, a Turkey test (p < 0.05)was then performed to compare means. Data were analyzed usingthe statistical software R (R Development Core Team, 2012).

3. Results and discussion

3.1. Biogas production

The results of the evaluation of biogas production by the differ-ent manure mixtures are shown in Table 2.

Digestion of mixtures with P50% cattle manure (CM100, CM75,and CM50) yielded significantly higher biogas potentials (p < 0.05)when compared to sheep bedding alone (CM0) (Table 2) unless

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

biogas production was normalized to the amount of VS added todigesters, mixtures with P25% cattle manure (CM100, CM75, CM50

and CM25) produced significantly higher (p < 0.05) amounts ofbiogas. However, there were no statistically significant differencesbetween CM100, CM75, and CM50 regarding biogas production.

Therefore, increasing the amount of cattle manure above 50% didnot result in further increases in biogas production.

The significant biogas yield of the system used here, even formixtures containing 50% of sheep bedding, is justified by the ben-efits of anaerobic co-digestion. In general, anaerobic co-digestionproduces higher amounts of biogas and methane than anaerobicdigestion of the individual substrates separately (Pagés-Díazet al., 2015; Borowski and Kubacki, 2015; Yong et al., 2015). Thelower levels of biogas produced by mixtures containing only sheepbedding – CM0 or 25% of it – CM25 (Table 2) resulted mainly fromthe high content of fibber in sheep bedding (confirming the resultsof Orrico Junior et al., 2012), and also from the presence of ricehusk in this mixture component. The presence of lignin in lignocel-lulosic material found in large quantities in rice husk provides abarrier to chemical and biological degradation, preventing micro-bial and enzymatic attack of cellulose and hemicellulose. Ruptureof this impermeable layer of lignin is required for hydrolysis ofcellulose and hemicellulose, and is often performed by physical,chemical, or biological pre-treatment of lignocellulose-rich manurebefore digester loading (Chandra et al., 2012). No pre-treatmentof sheep bedding was performed here. However, the potential ofbiogas shown in Table 2 refers to sheep bedding, which involvesboth the rice husk and the animal manure. In this sense, physicaland chemical changes occur during the period while the animalsare in contact with the lignocellulosic material, which affectspositively the degradation process. Tait et al. (2009) made thisstatement and showed that rice husk presents little degradable(10%), but its degradability increases after having been used asbedding for pigs (20%). Although specifically references concerninganaerobic digestion of sheep bedding is not available at literature,we can affirm that this material (sheep bedding) can be stabilizedby anaerobic digestion and the efficiency in biogas productionimproves when adding cattle manure in co-digestion (Kalamarasand Kotsopoulos, 2014; Xavier et al., 2015). The use of litter forthe biogas production through anaerobic digestion producessatisfactory results, but results with horse bedding are commonlyfound (Wartell et al., 2012; Böske et al., 2014, 2015).

Regarding biogas production curves (Fig. 1), we observed adelayed start and interruptions in biogas production from mostdigesters, resulting in increased duration of the experiment. Thisphenomenon was due to the absence of inoculum, and theinfluence of factors such as the room temperature. Daily biogasproduction values might be 50% lower at temperatures of 18 �Cwhen compared to those obtained at 25 �C (Alvarez and Lidén,2008). Also, methanogenic bacteria are sensitive to suddentemperature changes, which might result in substantial (30%)decreases in gas production (Deublein and Steinhauser, 2008).

However, the temperature levels observed during the experi-ment are common in southern Brazil and even lower in the winterseason. The results obtained are useful to estimate the biogasproduction in field conditions.

The presence of inhibitory compounds might also have con-tributed to suppressing biogas production in the first month ofour experimental period (Fig. 1). In the work reported by Hansenet al. (1998), 1.1 g/L was the lowest concentration of free ammonia(NH3) capable of inhibiting methanogenic activity. In our experi-ments, it is possible that ammonia concentrations in this rangewere found during fermentation, given the high pH values ofthe initial manure mixture (8.9–9.2) (Table 3), and also due tothe release of ammonia by protein and aminoacid degradation, asdescribed by Deublein and Steinhauser (2008). By the end of the

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040

Fig. 1. Biogas production curves and average and minimum temperature during the experimental period.

Table 2Biogas production from anaerobic co-digestion of sheep bedding with cattle manure in varying proportions.A

Biogas production (L kg�1) CV (%) p-value Cattle manure and sheep bedding mixtures

CM100 CM75 CM50 CM25 CM0

BP/TS 6.5 0.014 165 ± 12.2a 171 ± 11.0a 160 ± 6.8a 154 ± 10.1b 142 ± 10.6bBP/VS 6.5 0.043 212 ± 14.1a 221 ± 8.8a 211 ± 14.6a 206 ± 13.4a 188 ± 14.1bBP/dVS 53.5 0.069 1741 ± 194 3673 ± 2271 1679 ± 863 2264 ± 370 1282 ± 482BP/Res 6.5 <0.001 64 ± 4.8a 65 ± 2.6a 60 ± 4.1a 53 ± 3.4b 36 ± 2.7cBP/Subst 6.5 0.006 8.1 ± 0.59a 8.4 ± 0.33a 8.0 ± 0.55a 7.3 ± 0.47b 6.9 ± 0.52b

BP/TS – Biogas potential per kg of total solids added; BP/VS – Biogas potential per kg of volatile solids added; BP/dVS – Biogas potential per kg of volatile solids degraded;BP/Res – Biogas potential per kg of manure in natura (fresh manure) added; BP/Subst – Biogas potential per kg of substrate mixture added.CM0–CM100, cattle manure and sheep bedding mixtures digested, where the numbers represent the proportion of cattle manure in the mixture (in %); CV, coefficient ofvariation. Values in the same row followed by different letters were considered statistically different (ANOVA, Scott Knott test p < 0.05). The data distribution was normal(Shapiro–Wilk test, p < 0.05), and variances were homogeneous (Bartlett test, p < 0.05).

A Mean ± standard error.

4 T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx

co-digestion period (Table 3), the pH of samples had reachedadequate values, possibly due to the formation of organic acids.

3.2. Biofertilizer characteristics

The duration of the experiment observed for all experimentalconditions was 159 days. Table 3 shows the characteristics of theinitial manure mixture and of the final biofertilizer.

Increased proportions of sheep bedding in the mixture resultedin a biofertilizer with reduced TKN, P, and K (Table 3). Two impor-tant factors correlating with the chemical composition of thebiofertilizer were the initial characteristics of the manure (Table 1)and the conversion efficiency of the organic fraction into biogas(Table 2), which altered the concentration of nutrients in the med-ium. The diet fed to cattle also affected the characteristics of themanure. Orrico et al. (2007) reported that feeds with increasedproportions of concentrate relative to the amount of forage resultin manure with increased digestibility, and therefore faeces inwhich higher nutrient and carbon levels could be easily assimi-lated. Thus, the lower amount of concentrate in the sheep feed(30%) compared to that found in the cattle feed (40%) given tothe animals used in this study might have contributed to thedecreased levels of nutrients found in the biofertilizers fromexperimental conditions CM25 and CM0. The presence of rice huskin sheep bedding also contributed to decreasing the amounts ofnutrients in the mixtures containing higher proportions of thismanure, since rice husk has a high C/N ratio (109) and low levelsof nutrients such as K (<1% in the ash, which contains 65% silica)(Leconte et al., 2009; Ma et al., 2012).

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

Our results also show that increasing the amount of cattlemanure in the fermenting mixture resulted in higher HA/AF ratios(Table 3).

In general, biofertilizers are used as sources of nutrients ratherthan to modify the physical properties of the soil. Nevertheless, itis important to evaluate biofertilizer stability to ensure that, whenapplied to the soil, the biofertilizer will not behave as a manure innatura (fresh manure). The increase of the HA/FA during thedecomposition process of organic matter may occur either byincreasing the C content of the HA fraction as the decrease ofthe C content of FA fraction, depends on the characteristicsof the organic material. Anyway, it is indicative of humificationof organic matter (Senesi, 1989). We can observe by the data ofTable 3 that the larger the amount of cattle manure in themixture, the greater the difference between the ratio HA/FAconsidering the beginning and end of the anaerobic co-digestion.This fact may be related to the greater amount of labilesubstances in cattle manure compared to sheep bedding and theseto decompose more quickly contribute to the ratio AH/AFincreases forming a product, biofertilizer, more stable and withhigher quality. Although polymerization index (HA/FA) is mostcommonly used in composting studies (Senesi, 1989; Bernalet al., 2009), the results of this experiment are in agreement withthe observations made by other authors who used spectroscopictechniques to evaluate the transformations of organic matterduring anaerobic digestion (Tambone et al., 2009; Cuetos et al.,2010). Marcato et al. (2009) reported that, during anaerobicdigestion, organic matter is stabilized by fermentation of themost labile fractions (aliphatic structures, lipids, amides, and

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040

Table 3Characteristics of the initial manure mixture and of the final biofertilizer obtained by anaerobic co-digestion of sheep bedding and cattle manure.

Experimentalconditions

Samplea Initial manure mixture and final biofertilizer characteristics

TS VS pH ECb TKNc Pd Kd CODdissd HA/FA Cell.c Hemicell.c

CM100 Initiale 4.9 77.6 9.2 6.4 0.14 0.43 2.3 10.6 1.42 16.3 15.7Finalf 4.6 ± 0.1 72.1 ± 0.5 7.6 ± 0.05 11.2 ± 0.4 0.18 ± 0.004 0.5 ± 0.02 2.7 ± 0.041 8.2 ± 0.4 3.0 ± 0.1 10.3 ± 1.0 12.3 ± 0.6

CM75 Initiale 4.9 77.3 9.2 7.0 0.13 0.43 2.0 10.3 1.38 20.0 15.7Finalf 4.8 ± 0.2 72.7 ± 0.6 7.6 ± 0.08 11.1 ± 0.4 0.16 ± 0.02 0.6 ± 0.02 2.1 ± 0.07 7.7 ± 0.6 2.6 ± 0.1 12.4 ± 1.2 13.2 ± 0.6

CM50 Initiale 5.0 75.8 9.1 7.8 0.12 0.38 1.6 8.1 1.19 21.6 16.65Finalf 4.6 ± 0.3 71.1 ± 0.6 7.5 ± 0.08 12.0 ± 0.2 0.14 ± 0.01 0.5 ± 0.01 1.8 ± 0.05 6.6 ± 0.3 2.3 ± 0.03 14 ± 0.6 14.0 ± 1.0

CM25 Iniciale 4.7 75.1 9.0 8.2 0.10 0.33 1.3 7.5 1.04 22.0 18.1Finalf 4.5 ± 0.1 71.2 ± 0.4 7.4 ± 0.07 11.7 ± 0.2 0.11 ± 0.003 0.4 ± 0.003 1.4 ± 0.004 6.3 ± 0.2 1.5 ± 0.2 16.1 ± 0.8 15.8 ± 1.1

CM0 Initiale 4.9 75.5 8.9 9.3 0.08 0.30 0.9 6.2 1.00 23.2 18.2Finalf 4.3 ± 0.3 70.2 ± 0.5 7.3 ± 0.10 12.0 ± 0.4 0.09 ± 0.004 0.3 ± 0.02 1.0 ± 0.02 5.2 ± 0.5 1.3 ± 0.04 17.6 ± 0.9 16.1 ± 1.1

CM0–CM100, digested cattle manure and sheep bedding mixtures, where the number represents the proportion of cattle manure in the mixture (in %). EC, electric con-ductivity; NTK, total Kjedahl nitrogen; P, phosphorous; K, potassium; CODdiss., dissolved chemical oxygen demand; HA/FA, humic acids/fulvic acids; Cell., cellulose; Hemicell.,hemicelullose.

a ‘Initial’ samples were those taken from the manure mixture at the onset of digester loading, and ‘final’ samples were those taken from the final biofertilizer, at the end ofthe digestion period.

b In mS cm�1.c Concentration in % (w/v).d Concentration in g L�1.e The same initial mixture was used for all replicates of each experimental condition, and thus initial values represent single measurements with no standard deviation.f Mean ± Standard error.

T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx 5

polyssaccharides), leading to a relative increase in the levels of themost stable compounds.

3.3. Principal component analysis

Table 4 shows the correlation coefficients between the variablesused for principal component analysis (Fig. 2).

There were numerous correlations of moderate to high magni-tude between the variables, highlighting the inter-dependencebetween them and the importance of using a statistical analysistool such as PCA, which is capable of addressing such a complexpattern of variable inter-dependency (Lattin et al., 2011).

Two principal components (PCs) selected by PCA were capableof explaining 78.30% of data variability. Fig. 2 illustrates the asso-ciations between these PCs and the variables and experimentalconditions.

In PC1, the most relevant variables, which explained 64.96% ofdata variability, were: P, EC, HA/FA, TKN, K, biogas productionpotentials (negatively correlated), COD, and cellulose andhemicellulose contents (positive correlation) (Fig. 2). PC1 was

Table 4Linear correlation values. The bold value means the correlations of moderate to high mag

EC TKN P K COD HA/FA

EC 1TKN 0.68 1P 0.80 0.83 1K 0.80 0.61 0.61 1COD �0.59 �0.60 �0.66 �0.39 1HA/FA 0.87 0.77 0.86 0.81 �0.72 1Cell. �0.50 �0.29 �0.54 0.01 0.38 �0.37Hemicell. �0.36 �0.45 �0.31 �0.48 0.35 �0.36BP/TSa 0.62 0.57 0.71 0.41 �0.65 0.56BP/VSa 0.56 0.48 0.63 0.34 �0.57 0.47BP/dVSa 0.58 0.48 0.64 0.37 �0.60 0.50BP/Res.a 0.84 0.68 0.84 0.63 �0.71 0.82BP/Subst.a 0.66 0.62 0.77 0.44 �0.66 0.64

BP/TS, biogas production potential per kg of total solids (TS) added; BP/VS, biogas propotential per kg of degraded VS; BP/Res., biogas production potential per kg of manuresubstrate mixture added.

a Biogas production potential indicators.

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

characterized by the relationship between the organic fraction con-sumption (represented by the variables with positive correlation)and the phenomena trigged by this consumption (represented byvariables with negative correlation), such as biogas production,nutrient concentration, and sample stability changes. Thus, PC1can be referred to as ‘organic fraction/productivity’, because higherrates of organic fraction consumption led to higher digester pro-ductivity. All variables related to digester productivity displayedstrong correlation with PC1 (Fig. 2). Moreover, COD was theonly organic fraction consumption variable strongly correlatedwith PC1. Therefore, amongst the organic fraction consumptionvariables (COD, and cellulose and hemicellulose levels) COD aloneexplained most of the productivity of digesters, while cellulose andhemicellulose levels, which displayed moderate correlation tosome (but not all) productivity variables, explained only part ofthe biogas productivity in our experiments.

The experimental replicates of CM100, CM75, and CM50 (numbers1–12 in Fig. 2) displayed the highest weights among productivityvariables, while experimental replicates of CM25 and CM0

(numbers 13–20 in Fig. 2) displayed higher weights for organic

nitude between the variables.

Cell. Hemicell. TS_ada VS_ada VS_reda Res.a Sub.a

1�0.23 1�0.50 �0.41 1�0.51 �0.39 0.99 1�0.50 �0.41 0.99 0.99 1�0.52 �0.40 0.87 0.83 0.82 1�0.58 �0.42 0.97 0.95 0.96 0.87 1

duction potential per kg of volatile solids (VS) added; BP/dVS, biogas productionin natura (fresh manure) added; BP/Subst., biogas production potential per kg of

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040

Experimental conditions

Euc

lidia

n D

ista

nce

CM0CM25CM75CM50CM100

5,29

3,53

1,76

0,00

Fig. 3. Dendrogram showing the results of cluster analysis of the differentexperimental conditions, considering all biogas production variables as well asP, EC, TKN, HA/FA ratio, K, hemicellulose and cellulose levels, and COD.

BP/Subst

-2 -1 0 1 2 3

-2

-1

0

1

2

3

PC1 Scores

PC

2 S

core

s

12

3

45

6 78

9

10

11

12

13

14

15

161718

19

20

-0.5 0.0 0.5 1.0

-0.5

0.0

0.5

1.0

ECTKN

P

K

COD

HA_FA

Cellulose

HemicelluloseBP/TS

BP/VSBP/dVS

BP/Res

CM100: 1 - 4CM75: 5 - 8 CM50: 9 - 12CM25:13-16CM0:17-20

Fig. 2. PCA results relative to parameter variations during anaerobic co-digestion ofsheep bedding with cattle manure. BP/TS, BP/VS, BP/dVS, BP/Res, and BP/Substcorrespond to the potential biogas productions per kg of TS added, VS added,VS degraded, manure, and substrate.

6 T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx

fraction variables. Therefore, productivity and organic fractionvariables were most influenced by manure mixtures with higher(P50%) or lower (625%) ratios of cattle manure, respectively. Inparticular, manure mixtures with P50% cattle manure resultedin increased organic fraction consumption, biogas production,nutrient concentration, and biofertilizer stability when comparedto mixtures with 625% of this manure.

The most relevant variables for PC2, which explained 13.34% ofdata variability, were K and cellulose (negatively correlated); andbiogas production potential per unit of VS added, VS degraded,ST added, substrate, and hemicellulose levels (positivelycorrelated). In this component, cellulose levels were negativelycorrelated with variables describing biogas production potential,suggesting that cellulose consumption had greater influence onbiogas production when compared to that of hemicellulose.Yue et al. (2013) described that differences in the consumptionof cellulose and hemicellulose fractions during anaerobic digestionresult mainly from variations in the composition of the original

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

substrates. Also, rupture of the lignin matrix by animal digestionfacilitates enzymatic access to cellulose in animal manure (Trioloet al., 2011) and, if the lignin layer is no longer present, cellulasesrapidly degrade the easily accessible cellulose (Gallert andWinter, 2005). Interestingly, hemicellulose levels were negativelycorrelated with the concentration of potassium (K) (Fig. 2). At thispoint we have no explanation for this finding.

3.4. Cluster analysis (CA)

Fig. 3 shows a dendrogram produced by cluster analysis of theexperimental conditions tested.

Cluster analysis (CA) showed two separate groups, the firstformed by CM100, CM75, and CM50 (where CM75 and CM50 weremost similar to each other) and the second represented by CM25

and CM0 (Fig. 3). Overall, these results confirm the conclusion thatthe anaerobic co-digestion of sheep bedding with P50% cattlemanure is more efficient (CM100, CM75, and CM50) than the diges-tion of mixtures containing higher amounts of sheep bedding(CM25 and CM0).

4. Conclusions

PCA were effective in identifying that all variables related todigester productivity displayed strong correlation with PC1characterized by the relationship between the organic fractionconsumption (COD, cellulose and hemicellulose contents) andthe phenomena trigged by this consumption, such as biogasproduction, nutrient concentration, and sample stability changes.Therefore, amongst the organic fraction consumption variablesCOD alone explained most of the productivity of digesters.

Associated multivariate analysis (PCA and CA) were effective inidentifying that the addition of P50% cattle manure to sheepbedding allowed increased biogas production and resulted in highquality biofertilizer.

Anaerobic co-digestion caused little transformation of the ricehusk fraction in sheep bedding.

Funding

We thank to Coordination for the Improvement of HigherEducation Personnel (CAPES) for the assignment of the master’sscholarship. This research received no specific grant from anyfunding agency in the public, commercial, or not-for-profit sectors.

Acknowledgement

We would like to thank Coordination for the Improvement ofHigher Education Personnel (CAPES) for granting researchscholarships.

References

Alcarde, J.C., 2009. Manual de Análise de Fertilizantes. FEALQ, Piracicaba.Al-Masri, M.R., 2001. Changes in biogas production due to different ratios of some

animal and agricultural wastes. Bioresour. Technol. 77, 97–100.Alvarez, R., Lidén, G., 2008. The effect of temperature variation on biomethanation

at high altitude. Bioresour. Technol. 99, 7278–7284.American Public Health Association – APHA, AWWA, WEF, 2012. Standard methods

for the examination of water and wastewater, 22 ed. APHA, Washington.Amorim, A.C., de Lucas Junior, J., Resende, K.T., 2004. Biodigestão anaeróbia de

dejetos de caprinos obtidos nas diferentes estações do ano. Eng. Agríc. 24,16–24.

Ashekuzzaman, S.M., Poulsen, T.G., 2011. Optimizing feed composition forimproved methane yield during anaerobic digestion of cow manure basedwaste mixtures. Bioresour. Technol. 102, 2213–2218.

Agyeman, F.O., Tao, W., 2014. Anaerobic co-digestion of food waste and dairymanure: effects of food waste particle size and organic loading rate. J. Environ.Manage. 133, 268–274.

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040

T. Cestonaro et al. /Waste Management xxx (2015) xxx–xxx 7

Albuquerque, J.A., de la Fuente, C., Alicia Ferrer-Costa, A., Carrasco, L., Cegarra, J.,Abad, M., Bernal, M.P., 2012. Assessment of the fertiliser potential of digestatesfrom farm and agroindustrial residues. Biomass Bioenergy 40, 181–189.

Benites, V.M., Madari, B., Machado, P.L. de A., 2003. Extração e FracionamentoQuantitativo de Substâncias Húmicas do Solo: um Procedimento Simplificadode Baixo Custo. Embrapa Solos, Rio de Janeiro (Comunicado Técnico 16).

Bernal, M.P., Albuquerque, J.A., Moral, R., 2009. Composting of animal manures andchemical criteria for compost maturity assessment. A review. BioresourceTechnology 100, 5444–5453.

Borowski, S., Kubacki, P., 2015. Co-digestion of pig slaughterhouse waste withsewage sludge. Waste Manage. 40, 119–126.

Böske, J., Wirth, B., Garlipp, F., Mummeb, J., Van den Weghe, H., 2014. Anaerobicdigestion of horse dung mixed with different bedding materials in an upflowsolid-state (UASS) reactor at mesophilic conditions. Bioresour. Technol. 158,111–118.

Böske, J., Wirth, B., Garlipp, F., Mummeb, J., Van den Weghe, H., 2015. Upflowanaerobic solid-state (UASS) digestion of horse manure: thermophilic vs.mesophilic performance. Bioresour. Technol. 175, 8–16.

Campos, F.P.de, Nussio, C.M.B., Nussio, L.G., 2004. Métodos de Análise de Alimentos.FEALQ, Piracicaba.

Chandra, R., Takeuchi, H., Hasegawa, T., 2012. Methane production fromlignocellulosic agricultural crop wastes: a review in context to secondgeneration of biofuel production. Renew. Sustain. Energy Rev. 16, 1462–1476.

Costa, M.S.S.de.M., Cestonaro, T., Costa, L.A.de.M., Rozatti, M.A.T., Carneiro, L.J.,Pereira, D.C., Lorin, H.E.F., 2015. Improving the nutrient content of sheepbedding compost by adding cattle manure. J. Cleaner Prod. 86, 9–14.

Cuetos, M.J., Gómez, X., Otero, M., Morán, A., 2010. Anaerobic digestion of solidslaughterhouse waste: study of biological stabilization by Fourier Transforminfrared spectroscopy and thermogravimetry combined with massspectrometry. Biodegradation 21, 543–556.

Cunha-Queda, A.C.F., Vallini, G., Bruno de Sousa, R.F.X., Almeida Duarte, E.C.N.F.,2003. Estudo da evolução de actividades enzimáticas durante a compostagemde resíduos provenientes de mercados horto-frutícolas. Anais do InstitutoSuperior de Agronomia 49, 193–208.

Deublein, D., Steinhauser, A., 2008. Biogas from Waste and Renewable Resources:An Introduction. WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Ferreira, D.F., 2011. Estatística Multivariada, second ed. UFLA, Lavras.Gallert, C., Winter, J., 2005. Bacterial metabolism in wastewater treatment

systems. In: Jördening, H.-J., Winter, J. (Eds.), Environmental Biotechnology:Concepts and Applications. WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim,pp. 1–48.

Hansen, K.H., Angelidaki, I., Ahring, B.K., 1998. Anaerobic digestion of swinemanure: inhibition by ammonia. Water Res. 32, 5–12.

Kalamaras, S.D., Kotsopoulos, T.A., 2014. Anaerobic co-digestion of cattle manureand alternative crops for the substitution of maize in South Europe. Bioresour.Technol. 172, 68–75.

Khalid, A., Arshad, M., Anjum, M., Mahmood, T., Dawson, L., 2011. The anaerobicdigestion of solid organic waste. Waste Manage. 31, 1737–1744.

Khoufi, S., Louhichi, A., Sayadi, S., 2015. Optimization of anaerobic co-digestion ofolive mill wastewater and liquid poultry manure in batch condition and semi-continuous jet-loop Reactor. Bioresour. Technol. 182, 67–74.

Lattin, J.M., Carroll, J.D., Green, P.E., 2011. Análise de Dados Multivariados. CengageLearning, São Paulo.

Leconte, M.C., Mazzarino, M.J., Satti, P., Iglesias, M.C., Laos, F., 2009. Co-compostingrice hulls and/or sawdust with poultry manure in NE Argentina. Waste Manage.29, 2446–2453.

Ma, X., Zhou, B., Gao, W., Qu, Y., Wang, L., Wang, Z., Zhu, Y., 2012. A recyclablemethod for production of pure silica from rice hull ash. Powder Technol. 217,497–501.

Malavolta, E., Vitti, G.C., Oliveira, S.A., 1997. Avaliação do Estado Nutricional dasPlantas: Princípios e Aplicações, second ed. POTAFOS, Piracicaba.

Marcato, C., Mohtar, R., Revel, J., Pouech, P., Hafidi, M., Guiresse, M., 2009. Impact ofanaerobic digestion on organic matter quality in pig slurry. Int. Biodeter.Biodegrad. 63, 260–266.

Molinuevo-Salces, B., Gómez, X., Morán, A., García-González, M.C., 2013. Anaerobicco-digestion of livestock and vegetable processing wastes: fibre degradationand digestate stability. Waste Manage. 33, 1332–1338.

Please cite this article in press as: Cestonaro, T., et al. The anaerobic co-digestiand improves biofertilizer quality. Waste Management (2015), http://dx.doi.or

Motte, J.C., Escudié, R., Bernet, N., Delgenes, J.P., Steyer, J.P., Dumas, C., 2013.Dynamic effect of total solid content, low substrate/inoculum ratio and particlesize on solid-state anaerobic digestion. Bioresour. Technol. 144, 141–148.

Orrico, A.C.A., de Lucas Junior, J., Orrico Junior, M.A., 2007. Caracterização ebiodigestão anaeróbia dos dejetos de caprinos. Eng. Agríc. 27, 639–647.

Orrico Junior, A.A.P., Orrico, A.C.C., de Lucas Junior, J., Sampaio, A.A.M., Fernandes, A.R.M., de Oliveira, E.A., 2012. Biodigestão anaeróbia dos dejetos da bovinoculturade corte: influência do período, do genótipo e da dieta. R. Bras. Zootec. 41,1533–1538.

Orrico Junior, M.A.P., Orrico, A.C.A., de Lucas Junior, J., 2011. Produção animal e omeio ambiente: uma comparação entre potencial de emissão de metano dosdejetos e a quantidade de alimento produzido. Eng. Agríc. 31, 399–410.

Pagés-Díaz, J., Westman, J., Taherzadeh, M.J., Pereda-Reyes, I., Horváth, I.S., 2015.Semi-continuous co-digestion of solid cattle slaughterhouse wastes with otherwaste streams: interactions within the mixtures and methanogenic communitystructure. Chem. Eng. J. 273, 28–36.

Pokój, T., Bułkowska, K., Gusiatin, Z.M., Klimiuk, E., Jankowski, K.J., 2015. Semi-continuous anaerobic digestion of different silage crops: VFAs formation,methane yield from fiber and non-fiber components and digestate composition.Bioresour. Technol. 190, 201–210.

Development Core Team, R., 2012. R: A Language and Environment for StatisticalComputing. R Foundation for Statistical Computing, Vienna.

Senesi, N., 1989. Composted materials as organic fertilizers. The Science of the TotalEnvironment 81–82, 521–542.

Tait, S., Tamis, J., Edgerton, B., Batstone, D.J., 2009. Anaerobic digestion of spentbedding from deep litter piggery housing. Bioresour. Technol. 100, 2210–2218.

Tambone, F., Genevini, P., D’imporzano, G., Adani, F., 2009. Assessing amendmentproperties of digestate by studying the organic matter composition and thedegree of biological stability during the anaerobic digestion of the organicfraction of MSW. Bioresour. Technol. 100, 3140–3142.

Tambone, F., Scaglia, B., D’Imporzano, G., Schievano, A., Orzi, V., Salati, S., Adani, F.,2010. Assessing amendment and fertilizing properties of digestates fromanaerobic digestion through a comparative study with digested sludge andcompost. Chemosphere 81, 577–583.

Teixeira, D.L., Miranda-de la Lama, G.C., Pascual-Alonso, M., Aguayo-Ulloa, L.,Villarroel, M., María, G.A., 2013. A note on lamb’s choice for different types ofbedding materials. J. Veterinary Behav. 8, 175–179.

Teixeira, D.L., Villarroel, M., María, G.A., 2014. Assessment of different organicbeddings materials for fattening lamb. Small Ruminant Res. 119, 22–27.

Teixeira, D.L., Miranda-de la Lama, G.C., Villarroel, M., OLLETA, J.L., García-Belenguer, S., Escós, J., María, G.A., 2015. Effects of alternative beddingsubstrates on lamb welfare, productive performance, and meat quality duringthe finishing phase of fattening. J. Veterinary Behav. 10, 171–178.

Triolo, J.M., Sommer, S.G., Møller, H.B., Weisbjerg, M.R., Jiang, X.Y., 2011. A newalgorithm to characterize biodegradability of biomass during anaerobicdigestion: influence of lignin concentration on methane production potential.Bioresour. Technol. 102, 9395–9402.

Wartell, B.A., Krumins, V., Alt, J., Kang, K., Schwab, B.J., Fennell, D.E., 2012. Methaneproduction from horse manure and stall waste with softwood bedding.Bioresour. Technol. 112, 42–50.

Wolf, B.T., Molloy, H.R.B., Trayte, M.J., Rose, M.T., 2010. Behaviour of growing lambshoused on straw or woodchip bedding materials and their preference for floortype. Appl. Animal Behav. Sci. 124, 45–50.

Xavier, C.A.N., Moset, V., Wahid, R., Møller, H.B., 2015. The efficiency of shreddedand briquetted wheat straw in anaerobic co-digestion with dairy cattle manure.Biosyst. Eng. 139, 16–24.

Yong, Z., Dong, Y., Zhang, X., Tan, T., 2015. Anaerobic co-digestion of food waste andstraw for biogas production. Renew. Energy 78, 527–530.

Yue, Z., Chen, R., Yang, F., Maclellan, J., Marsh, T., Liu, Y., Liao, W., 2013. Effects ofdairy manure and corn stover co-digestion on anaerobic microbes andcorresponding digestion performance. Bioresour. Technol. 128, 65–71.

Zhang, Y., Banks, C.J., 2013. Impact of different particle size distributions onanaerobic digestion of the organic fraction of municipal solid waste. WasteManage. 33, 297–307.

Zheng, Z., Liu, J., Yuan, X., Wang, X., Zhu, W., Yang, F., Cui, Z., 2015. Effect of dairymanure to switchgrass co-digestion ratio on methane production and thebacterial community in batch anaerobic digestion. Appl. Energy 151, 249–257.

on of sheep bedding and P50% cattle manure increases biogas productiong/10.1016/j.wasman.2015.08.040