Bioresource Technology 169 (2014) 38–44

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Biogas energy production from tropical biomass wastes by anaerobicdigestion

http://dx.doi.org/10.1016/j.biortech.2014.06.0670960-8524/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 330 263 3855; fax: +1 330 263 3670.E-mail address: li.851@osu.edu (Y. Li).

Xumeng Ge a, Tracie Matsumoto b, Lisa Keith b, Yebo Li a,⇑a Department of Food, Agricultural and Biological Engineering, The Ohio State University/Ohio Agricultural Research and Development Center, 1680 Madison Ave., Wooster, OH44691-4096, USAb USDA, ARS, DKI US PBARC, Plant Genetic Resources and Disease Research, 64 Nowelo Street, Hilo, HI 96720, USA

h i g h l i g h t s

� Biogas energy yield of tropical biomass waste was evaluated.� Liquid anaerobic digestion (L-AD) is effective for treating tropical food wastes.� Solid-state anaerobic digestion (SS-AD) is suitable for treating albizia biomass.� Albizia biomass showed higher methane yield than similar lignocellulosic biomass.

a r t i c l e i n f o

Article history:Received 3 May 2014Received in revised form 13 June 2014Accepted 18 June 2014Available online 26 June 2014

Keywords:TropicalBiomass wasteAnaerobic digestionBiogasAlbizia

a b s t r a c t

Anaerobic digestion (AD) is an attractive technology in tropical regions for converting locally abundantbiomass wastes into biogas which can be used to produce heat, electricity, and transportation fuels.However, investigations on AD of tropical forestry wastes, such as albizia biomass and food wastes, suchas taro, papaya, and sweet potato, are limited. In this study, these tropical biomass wastes were evaluatedfor biogas production by liquid AD (L-AD) and/or solid-state AD (SS-AD), depending on feedstock charac-teristics. When albizia leaves and chips were used as feedstocks, L-AD had greater methane yields (161and 113 L kg�1 VS, respectively) than SS-AD (156.8 and 59.6 L kg�1 VS, respectively), while SS-ADachieved 5-fold higher volumetric methane productivity than L-AD. Mono-digestion and co-digestionof taro skin, taro flesh, papaya, and sweet potato achieved methane yields from 345 to 411 L kg�1 VS,indicating the robustness of AD technology.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Tropical regions have high biomass productivity compared toother regions (Parikka, 2004). Large amounts of biomass wasteare generated each year from agricultural, forestry, and food sys-tems in tropical regions, such as Eastern Africa (Ferrey, 2006;Otieno and Awange, 2006; Scheffran, 2010). Albizia moluccana(albizia) is one of the fastest growing tropical and subtropical trees(West, 2014), and is considered an invasive species on islandsacross the Pacific, such as Hawaii (EL Little and Skolmen, 1989).Taro, papaya, and sweet potato, which are native to tropicalregions, are produced as traditional food (Manshardt, 2014;Midmore and Nguyen, 2003) and contribute large quantities offood wastes. For example, about 50% of the fresh papaya grownin Hawaii deteriorate and cannot be sold, creating waste that

requires further treatment (Gill, 2004). As a result, there is anopportunity to improve the sustainability of energy production intropical regions by converting these locally abundant biomasswastes into bioenergy products; however, potential technologiesneed to be evaluated.

Anaerobic digestion (AD) is a widely used technology that canprocess various kinds of organic wastes for biogas production bydecomposing organic matter under oxygen-free conditions (Yuand Schanbacher, 2010). The biogas can be used to produce heat,electricity, compressed natural gas (CNG), and/or liquefied naturalgas (LNG). The digestate, which contains nitrogen and phosphorus,can be used as a soil amendment. AD can be carried out at differenttotal solids (TS) contents. Liquid AD (L-AD) is generally operated ata TS content of less than 15%, while solid-state AD (SS-AD) is usu-ally operated at TS higher than 15%. High methane yields havebeen obtained in L-AD due to the good control of temperature,dilution of inhibitors, and good mass transfer provided by mixing.Compared to L-AD, SS-AD generally has higher volumetric methane

X. Ge et al. / Bioresource Technology 169 (2014) 38–44 39

productivity, fewer moving parts, lower energy requirements forheating and mixing, and an end product that is easier to handle.Floating and stratification of fats and fibers, a problem of L-AD,can be solved in SS-AD. The drawback of SS-AD is the lower meth-ane yield than L-AD, which is caused by inadequate mass transferin the system (Li et al., 2011).

AD of taro, papaya and sweet potato has been reported in sev-eral publications with methane yields ranging from 85 to360 L kg�1 VS. Bindu and Ramasamy (2008) studied biogas produc-tion of taro from AD feeding with solid feedstock, and obtained CH4

yields of 156–360 L kg�1 VS. Yang et al. (1984) reported CH4 yieldsof 85–357 L kg�1 VS during AD of papaya processing wastes.Shiralipour and Smith (1984) investigated CH4 production of dif-ferent storage roots and average CH4 yields about 330 L kg�1 VSwere obtained.

The biogas yield of AD is substantially affected by the composi-tion of the feedstocks (Ahn et al., 2010). Even for the same speciesof biomass, its composition can vary with the geographical loca-tion, variety, and harvesting season (Templeton et al., 2009).Therefore, characterization of feedstock components, such ascellulose, hemicellulose, lignin, and protein, is important for theestimation of methane yield.

To date, no reports on AD of albizia biomass nor on comparisonbetween L-AD and SS-AD of tropical biomass wastes have beenfound. The objective of this study was to evaluate tropical biomasswastes, including albizia leaves, albizia chips, taro skin, taro flesh,papaya, and sweet potato, as feedstocks for biogas production byAD. The compositions of these feedstocks were analyzed. BothL-AD and SS-AD of these biomass wastes were conducted.Degradation of glucan and hemicellulose in albizia biomass duringAD was also investigated.

2. Methods

2.1. Feedstock and inoculum

Albizia trees were cut from the United States Department ofAgriculture (USDA), Agricultural Research Service (ARS), Daniel KInouye (DKI) Pacific Basin Agricultural Research Center (PBARC)in Hilo, Hawaii. Branches containing leaves were trimmed andthe main stem was shredded with a chipper/shredder with a gratesize of 38 mm (Goossen CS1000 PTO Model, Harper, KS, USA).Removed leaves were shredded using a food processor (KitchenAid,St. Joseph, MI, USA), placed in plastic bags and frozen prior to ship-ment. Taro skin and cooked taro flesh were waste products col-lected from a local poi factory. Culled papaya fruits were takenfrom a local papaya packing house. Sweet potato tuberous rootswere remnants collected from a field after harvest. The collectedfood wastes (taro, papaya, and sweet potato) were cut and pro-cessed in a food processor (KitchenAid, St. Joseph, MI, USA), storedin plastic bags, and completely frozen prior to shipment.

The processed feedstocks were shipped in cooler boxes with iceto the Ohio Agricultural Research and Development Center (Woos-ter, OH, USA) then stored at �20 �C (4 �C for albizia chips), andthawed before use. The following mixtures were prepared as feed-stocks for co-digestion: taro skin/taro flesh (1:1, based on freshweight); papaya/taro flesh (1:1, based on fresh weight); papaya/sweet potato (1:1, based on fresh weight); and taro skin/taroflesh/papaya/sweet potato (1:1:1:1, based on fresh weight).

Effluent from a mesophilic liquid anaerobic digester feedingwith sewage sludge operated by Schmack Bioenergy (Akron, OH,USA) was used as inoculum for L-AD tests. The Schmack digesteris a continuous stirred-tank reactor with a hydraulic retention timeof 28 days. The effluent was centrifuged at 4000 rpm for 30 min toincrease the total solids (TS) content from 7.6% to 17.7% for SS-AD

tests. The inoculums were stored at 4 �C, and acclimated at 37 �Cfor 3 days before use. Feedstocks and inoculum samples were driedat 40 �C to reach a moisture content of lower than 10% (w/w) andthen ground to 20-mesh prior to composition analysis.

2.2. Anaerobic digestion

L-AD lab-scale systems were set up by mixing the feedstockwith inoculum to obtain a feedstock/effluent (F/E) ratio of 0.5(based on volatile solids, VS), and adding tap water to obtain a TScontent of 5%. L-AD trials were carried out in 1-L reactors, eachof which contained about 800 g of mixed feedstock, inoculum,and water, and were sealed with a rubber stopper with an outletfor biogas collection. Reactors were placed on a shaker in a walk-in incubator, and incubated at 37 �C with orbital shaking at100 rpm. Biogas generated was collected in 5-L Tedlar gas bagsconnected to the outlet of the reactor (CEL Scientific Tedlar gasbag, Santa Fe Springs, CA, USA). Biogas composition and volumewere measured every 2 days (or longer during late periods of AD)for 24 days. Triplicate reactors were run at each condition. AD trialswith only inoculum were also conducted as control.

Lab-scale SS-AD of albizia biomass was set up at an F/E ratio of2.3 (based on VS) and a TS content of 20%. SS-AD trials were carriedout with duplicate reactors at 37 �C for 50 days without agitation.Other conditions and operations were the same as those for L-AD.

After AD, the weight of digestate in each reactor was deter-mined, and digestate samples were taken to determine TS. The restof the digestate was dried and ground with the same method asused for the preparation of feedstock sample for compositionanalysis.

2.3. Analytical methods

The TS and VS were determined based on gravimetric analysis(APHA, 2005). Total carbon (TC) and total nitrogen (TN) were mea-sured by an elemental analyzer (Vario Max CNS, Elementar Amer-icas, Mt. Laurel, NJ, USA) to calculate the C/N ratio. Extractives weredetermined using the method reported by Sluiter et al. (2005).Crude protein contents were measured based on the methoddescribed by Hames et al. (2008). Crude lipids were analyzed byextraction from dry solids using Soxhlet extraction with hexaneas solvent. Glucan, lignin, and hemicellulose were analyzed basedon the NREL method (Sluiter et al., 2008). Briefly, extractive freebiomass was hydrolyzed into monomers after a two-step acidhydrolysis, and the concentrations of these monomers were deter-mined by HPLC (Shimadzu LC-20AB, Columbia, MD, USA). Acid-sol-uble lignin was measured by UV–Vis spectroscopy, and acid-insoluble lignin was determined by gravimetric analysis.

The volume of biogas collected in the Tedlar bags was measuredby a drum-type gas meter (Ritter, TG 5, Bochum, Germany) at 25 �Cand ambient pressure. The composition of biogas (CO2, CH4, N2,and O2) was analyzed by a gas chromatograph (GC) (HP 6890, Agi-lent Technologies, Wilmington, DE, USA) equipped with an alu-mina/KCl deactivation column (30 m � 0.53 mm � 10 mm) and aThermal Conductivity Detector (TCD). Helium was used as a carriergas at a flow rate of 5.2 mL/min. Temperatures of the injector, col-umn, and detector were set at 150, 40, and 200 �C, respectively. Themethane production contributed by inoculum was subtracted fromthe measured value of each treatment.

The performance of the AD process was evaluated using CH4

yield and volumetric CH4 productivity. The CH4 yield (L kg�1 VS)was defined as the volume of CH4 produced per VS of feedstockadded. The volumetric CH4 productivity (L L�1 d�1) was definedas the volume of CH4 produced per reactor volume per day.

40 X. Ge et al. / Bioresource Technology 169 (2014) 38–44

2.4. Statistical analysis

Statistical significance was determined by analysis of variance(ANOVA, a = 0.05) using Minitab (Version 16, Minitab, Inc., StateCollege, PA, USA). Experimental data were presented as averagevalues (data points in figures) ± standard deviations (error bars infigures).

3. Results and discussion

3.1. Characteristics of feedstocks and inoculums

As shown in Table 1, albizia leaves, albizia chips, and sweetpotato had relatively high TS contents, ranging from 35.5% to56.2%. Taro skin and taro flesh had relatively low TS contents(14.8% and 21.1%, respectively). Papaya showed the lowest TS con-tent (10.7%), which makes it unsuitable for SS-AD even whenmixed with the centrifuged inoculum. The VS contents of thesetropical biomass wastes were higher than 92%, except for taro skinwhich had a VS content of 83.5%, significantly (p < 0.05) lower thanthose of other feedstocks.

In this research, the glucan content of albizia leaves was 10.9%and 7–9% lower than those of Leucaena leucocephala leaves (12%)(Narayanaswami et al., 1986) and fallen tree leaves (11.7%), respec-tively (Brown et al., 2012). The lignin content (26.9%) of albizialeaves was 16–19% higher than that of fallen tree leaves (22.7–23.1%) and 3.9–5.6 times higher than that of Leucaena leucocephalaleaves (4.1–5.5%). The hemicellulose content (4.7%) of albizialeaves was relatively low, and less than half that of fallen treeleaves (10.6–11.5%). The relatively higher crude protein (7.4%) ofalbizia leaves resulted in a lower C/N ratio (9.8) compared to otherfeedstocks in this study, as well as Leucaena leucocephala leavesand fallen tree leaves in other studies (Brown et al., 2012; Xuet al., 1993). The extractive content of albizia leaves was 33.6%,comparable to those of fallen tree leaves (33.0–35.1%) (Brownet al., 2012).

Compared to various woody biomass, such as Eucalyptus, BlackAlder, Cottonwood, Hybrid poplar, Loblolly pine and Sycamore,that have been reported for AD studies (Jerger et al., 1982), albiziachips showed comparable glucan and lignin contents of 36.0% and28.1%, respectively, but a considerably lower hemicellulose con-tent (10.5%).

Taro and sweet potato, which are starchy biomass, showed highglucan contents. Papaya, a fruit, had a high content of extractives(77.6%), which contained a considerable amount of free glucose

Table 1Composition of feedstocks and inoculums.

Compositiona Feedstock

Albizia leaves Albizia chips Taro skin Ta

TS (%) 36.5 ± 0.1 56.2 ± 1.6 14.8 ± 1.0 21VS (%) 93.7 ± 0.1 98.3 ± 0.2 83.5 ± 0.5 98TC (%) 49.7 ± 0.6 45.1 ± 1.1 47.5 ± 0.9 47TN (%) 5.1 ± 0.2 0.9 ± 0.2 2.0 ± 0.3 1C/N 9.8 ± 0.2 52.2 ± 14.3 23.8 ± 4.4 47Extractives (%) 33.6 ± 0.2 8.5 ± 0.5 31.5 ± 5.0 41Free glucoseb (%) 0.0 ± 0.0 0.0 ± 0.0 5.5 ± 0.0 1Glucan (%) 10.9 ± 0.0 36.0 ± 0.9 19.0 ± 0.8 50Hemicellulose (%) 4.7 ± 0.0 10.5 ± 0.8 4.4 ± 0.4 0Lignin (%) 26.9 ± 0.4 28.1 ± 0.1 13.0 ± 0.4 3Crude lipid (%) 5.9 ± 0.0 0.2 ± 0.0 0.2 ± 0.0 0Crude protein (%) 7.4 ± 0.9 3.1 ± 0.0 3.5 ± 0.2 1

N/D: undetected; values are present as average ± standard deviationa Based on total dry weight except for TS.b One component of extractives.

(24% based on total extractives). Therefore, these food wastesshould be highly digestible.

3.2. L-AD of tropical biomass wastes

L-AD of taro skin, taro flesh, papaya, and sweet potato obtainedCH4 yields from 345 to 384 L kg�1 VS, which were significantly(p < 0.05) higher than those of albizia leaves (161 L kg�1 VS) andalbizia chips (113 L kg�1 VS) (Fig. 1a). Compared to lignocellulosicbiomass, food wastes usually have high amounts of organic solu-bles, thus can achieve higher methane (CH4) yields (Li et al., 2011).

During L-AD of albizia leaves, CH4 yield increased rapidly in thefirst 4 days to about 100 L kg�1 VS, and then increased slowly inthe following 20 days to a steady state (Fig. 1a). During L-AD ofalbizia chips, a lag-phase was observed in the first 8 days, and thenCH4 yield increased linearly to about 110 L kg�1 VS in the following14 days (Fig. 1a). During L-AD of taro skin, taro flesh, and sweetpotato, no lag phase was observed, and the CH4 yield increasedrapidly to about 200 L kg�1 VS in the first 6–8 days, and then grad-ually increased to steady state (about 350 L kg�1 VS). CH4 yield ofpapaya increased slowly in the first 4 days, but still reached a levelcomparable to those of other food wastes in 24 days.

The CH4 content of food wastes (70–80%) was much higher thanthose of albiza leaves and albizia chips (60–70%) at steady state.During L-AD of albizia leaves and albizia chips, CH4 contentsincreased to 60–70% in 10 days, and then maintained at that levelwith slight variations. During L-AD of taro skin, taro flesh, papaya,and sweet potato, CH4 contents reached maximum levels (70–80%)in a shorter time (8 days). After that, CH4 contents maintained atthe maximum levels until the termination of L-AD (Fig. 1b).

Compared to CH4 yields of fallen tree leaves (81.0 L kg�1 VS)(Brown et al., 2012), albizia leaves achieved higher CH4 yields atthe same temperature (37 �C). In general, CH4 yields of food wastesobtained in this study were higher or comparable to those reportedin other publications (85–360 L kg�1 VS), although a higher tem-perature (37 �C) was used in this study compared to these reports(30–35 �C) (Bindu and Ramasamy, 2008; Shiralipour and Smith,1984; Yang et al., 1984). Besides, differences in other factors, suchas mixing conditions and source of inoculum, might also contributeto the difference in CH4 yield between this study and the others.

3.3. SS-AD of tropical biomass wastes

During SS-AD of albizia leaves and albizia chips, about 156.8and 59.6 L kg�1 VS of CH4, respectively, were produced in 50 days(Fig. 2a), while maximum CH4 contents of around 65% were

Inoculum

ro flesh Papaya Sweet potato Original Centrifuged

.1 ± 0.5 10.7 ± 0.2 35.5 ± 0.3 7.6 ± 0.0 17.7 ± 1.1

.0 ± 0.1 92.4 ± 0.2 97.6 ± 0.1 50.0 ± 0.0 56.9 ± 1.2

.9 ± 1.0 52.0 ± 5.5 44.8 ± 0.0 48.7 ± 2.1 35.5 ± 0.9

.0 ± 0.1 4.2 ± 0.3 1.0 ± 0.0 11.5 ± 0.4 7.7 ± 0.0

.3 ± 3.8 12.4 ± 0.5 46.4 ± 1.0 4.4 ± 0.2 4.6 ± 0.1

.4 ± 0.2 77.6 ± 2.9 12.1 ± 0.2 10.4 ± 0.0 10.5 ± 0.0

.3 ± 0.9 18.7 ± 0.1 0.2 ± 0.0 N/D N/D

.3 ± 1.1 6.3 ± 0.1 77.3 ± 0.6 1.6 ± 0.0 1.8 ± 0.0

.4 ± 0.1 1.5 ± 0.1 0.4 ± 0.1 1.5 ± 0.1 1.5 ± 0.0

.1 ± 0.1 5.1 ± 0.1 3.1 ± 0.0 N/D N/D

.0 ± 0.0 2.9 ± 0.1 0.3 ± 0.0 N/D N/D

.8 ± 0.1 4.6 ± 0.4 1.7 ± 0.1 N/D N/D

0

100

200

300

400

500

0 5 10 15 20 25

Cum

ulat

ive

CH

4yi

eld

(L k

g-1V

S)

Time (d)

Albizia Leaves Albizia ChipsTaro Skin Taro FleshPapaya Sweet potato

0

20

40

60

80

100

0 5 10 15 20 25

CH

4co

nten

t (%

)

Time (d)

Albizia Leaves Albizia ChipsTaro Skin Taro FleshPapaya Sweet potato

a b

Fig. 1. CH4 yields (a) and contents (b) during L-AD of tropical biomass wastes.

0

40

80

120

160

200

0 5 10 15 20 25 30 35 40 45 50

Cum

ulat

ive

CH

4yi

eld

(L k

g-1V

S)

Time (d)

SS-ADAlbizia Leaves

Albizia Chips

0

20

40

60

80

100

0 5 10 15 20 25 30 35 40 45 50

CH

4co

nten

t (%

)

Time (d)

SS-ADAlbizia Leaves

Albizia Chips

a b

Fig. 2. CH4 yields (a) and contents (b) during SS-AD of albizia leaves and chips.

X. Ge et al. / Bioresource Technology 169 (2014) 38–44 41

achieved after 10 days (Fig. 2b). The CH4 yield of albizia leaves washigher than that obtained in SS-AD of pretreated or un-pretreatedfallen tree leaves (70–82 L kg�1 VS) (Brown et al., 2012; Liew et al.,2011). Compared to albizia leaves, albizia chips showed a lowerCH4 yield during SS-AD. Besides the difference in surface areaand composition, one reason for the difference might be that albi-zia chips had a more recalcitrant structure than albizia leaves. LowCH4 yields (15–50 L kg�1 VS) during SS-AD of other woody biomasswere also observed by researchers at 35–37 �C (Brown et al., 2012;Jerger et al., 1982). Zhao et al. (2014a, 2014b) reported on an envi-ronmentally friendly fungal pretreatment technology thatincreased the CH4 yield of yard trimmings by 154%. It is reasonableto envisage that, with proper pretreatment technologies, the CH4

yield of albizia chips could be increased.

3.4. Degradation of glucan, hemicellulose, and extractives of albizia byAD

For albizia feedstocks, degradation of extractives was signifi-cantly (p < 0.05) higher than degradation of hemicellulose and glu-can, except for L-AD of albizia chips (Fig. 3). The reason for thedifference may be that extractives usually contain compounds

such as free sugars, oligomers, and organic acids (Chen et al.,2007), which are readily degradable (Zheng et al., 2009), whilehemicellulose is easier to degrade than cellulose, which is a maincomponent of glucan in lignocellulosic biomass (Table 1).

During L-AD, degradation of glucan, hemicellulose, and extrac-tives in albizia leaves was higher than that of albizia chips, butonly the difference in extractives was significant (p < 0.05)(Fig. 3). During SS-AD, degradation of glucan, hemicellulose, andextractives in albizia leaves was significantly (p < 0.05) higherthan that of albizia chips (Fig. 3). Leaves and chips have differentlignocellulosic structures, which may have contributed to the dif-ference in degradation. In addition, the larger surface area of theprocessed albizia leaves may have been another reason for thehigher degradation.

Positive linear correlations between CH4 yield and degradationof glucan and hemicellulose in albizia biomass were observed asillustrated in Fig. 4. This result is consistent with Liew et al.(2011) study on AD of fallen tree leaves in which higher CH4 yieldswere observed in reactors with higher glucan and hemicellulosedegradation. The correlations between CH4 yield and degradationof glucan and hemicellulose were fitted with linear equations withR2 higher than 0.87 (Fig. 4a and b).

0

20

40

60

80

100

L-AD L-AD SS-AD SS-AD

Deg

rada

tion

(%)

Glucan Hemicellulose Extractives

Albizia leaves Albizia chips Albizia leaves Albizia chips

Fig. 3. Degradation of glucan, hemicellulose and extractives by L-AD and SS-AD of albizia leaves and albizia chips.

R² = 0.874

40

80

120

160

200

0 20 40 60

CH

4yi

eld

(L k

g-1V

S)

Degradation (%)

Glucan

R² = 0.898

40

80

120

160

200

0 20 40 60 80

CH

4yi

eld

(L k

g-1V

S)

Degradation (%)

Hemicellulosea b

Fig. 4. Correlation between methane yield and degradation of glucan (a) andhemicellulose (b) by AD of albizia.

42 X. Ge et al. / Bioresource Technology 169 (2014) 38–44

3.5. Co-digestion of tropical food wastes

CH4 yields of co-digestion with different combinations of taroskin, taro flesh, papaya, and sweet potato were comparable to eachother with values ranging from 371 to 411 L kg�1 VS (Fig. 5a). CH4

contents during co-digestion reached about 70% in 8 days, and

0

100

200

300

400

500

600

0 5 10 15 20 25

Cum

ulat

ive

CH

4yi

eld

(L k

g-1V

S)

Time (d)

Taro Skin/Taro FleshPapaya/Taro FleshPapaya/Sweet potatoTaro Skin/Taro Flesh/Papaya/Sweetpotato

a b

Fig. 5. CH4 yields (a) and contents (b) during li

then maintained at that level with variations between 70% and80% (Fig. 5b). The CH4 yields by co-digestion of food wastes in thisstudy are higher or comparable with those reported in other pub-lications, such as co-digestion of fruit and vegetable wastes withother substrates at 30–35 �C (Bouallagui et al., 2009; Garcia-Peñaet al., 2011; Lin et al., 2011).

The CH4 yields and contents of these co-digestions were not sig-nificantly (p > 0.05) different from those of the individual feed-stocks, indicating the robustness of the AD process in treatingdifferent types of feedstocks. Although these tropical food wasteshad different compositions, they mainly contain easily degradableorganic components such as extractives and starch (Table 1).Therefore, they showed comparable CH4 yields during L-AD.

3.6. Volumetric CH4 productivity of AD using tropical biomass wastes

L-AD of albizia biomass, especially albizia chips, showed higherCH4 yields than SS-AD (Figs. 1a and 2a), while SS-AD of albizialeaves and albizia chips achieved 5 times higher (p < 0.05) volu-metric CH4 productivities than L-AD of albizia leaves and chips(Fig. 6).

0

20

40

60

80

100

0 5 10 15 20 25

CH

4co

nten

t (%

)

Time (d)

Taro Skin/Taro FleshPapaya/Taro FleshPapaya/Sweet potatoTaro Skin/Taro Flesh/Papaya/Sweetpotato

quid co-digestions of tropical food wastes.

0.0

0.1

0.2

0.3

0.4

0.5

Vol

umet

ric

CH

4pr

oduc

tivi

ty (

L L

-1d-1

)

SS-AD L-AD

Fig. 6. Volumetric CH4 productivity of AD of tropical biomass wastes.

X. Ge et al. / Bioresource Technology 169 (2014) 38–44 43

Taro skin, taro flesh, papaya, sweet potato, and their mixturesachieved CH4 yields ranging from 345 to 411 L kg�1 VS, about3-fold higher than those of albizia biomass (Figs. 1a and 2a).However, the volumetric productivity of SS-AD of albizia leaveswas also significantly (p < 0.05) higher than those obtained duringL-AD of taro, papaya, sweet potato, and their mixtures (Fig. 6). Thismeans that more methane can be produced by SS-AD of albiziathan by L-AD of albizia or other tropical biomass wastes usingdigesters with the same size.

4. Conclusions

Compared to leaves and woody biomass of other plants thathave been used in AD studies, albizia leaves and chips showedcomparable glucan and lignin contents, but significantly low hemi-cellulose content. CH4 yields obtained from tropical biomasswastes were comparable or higher than those reported in literatureusing similar feedstocks. The highest CH4 yield was obtained fromL-AD of tropical food wastes, while the highest volumetric produc-tivity was obtained from SS-AD of albizia biomass. The AD systemshowed robustness in treating different tropical food wastes,including individual feedstock and their mixtures.

Acknowledgement

This project was funded by the United States Office of NavalResearch and USDA NIFA Hatch Program. The authors would liketo thank Mrs. Mary Wicks (Department of Food, Agricultural andBiological Engineering, OSU) for reading through the manuscriptand providing useful suggestions. The authors also would like tothank Dr. Dennis Gonsalves for his vision and scientific discussionon the utilization of feedstock from Hawaii.

References

Ahn, H.K., Smith, M.C., Kondrad, S.L., White, J.W., 2010. Evaluation of biogasproduction potential by dry anaerobic digestion of switchgrass-animal manuremixtures. Appl. Biochem. Biotechnol. 160, 965–975.

APHA, 2005. Standard Methods for the Examination of Water and Wastewater.American Public Health Association, Washington, DC.

Bindu, T., Ramasamy, E.V., 2008. Recovery of energy from Taro (Colocasia esculenta)with solid-feed anaerobic digesters (SOFADs). Waste Manage. 28, 396–405.

Bouallagui, H., Lahdheb, H., Ben Romdan, E., Rachdi, B., Hamdi, M., 2009.Improvement of fruit and vegetable waste anaerobic digestion performanceand stability with co-substrates addition. J. Environ. Manage. 90, 1844–1849.

Brown, D., Shi, J., Li, Y., 2012. Comparison of solid-state to liquid anaerobic digestionof lignocellulosic feedstocks for biogas production. Bioresour. Technol. 124,379–386.

Chen, S.F., Mowery, R.A., Scarlata, C.J., Chambliss, C.K., 2007. Compositional analysisof water-soluble materials in corn stover. J. Agric. Food Chem. 55, 5912–5918.

EL Little, J., Skolmen, R.G., 1989. Common Forest Trees of Hawaii (Native andIntroduced). United States Department of Agriculture, Washington, DC.

Ferrey, S., 2006. Renewable Power in Developing Countries: Winning the War onGlobal Warming. PennWell Books, Tulsa.

Garcia-Peña, E.I., Parameswaran, P., Kang, D.W., Canul-Chan, M., Krajmalnik-Brown,R., 2011. Anaerobic digestion and co-digestion processes of vegetable and fruitresidues: process and microbial ecology. Bioresour. Technol. 102, 9447–9455.

Gill, A.T., 2004. Prospective direct use enterprises in Kapoho, Hawaii. Trans.Geotherm. Resour. Counc. 28, 85–89.

Hames, B., Scarlata, C., Sluiter, A., 2008. Determination of protein content inbiomass. National Renewable Energy Laboratory Technical Report NREL/TP-510-42625.

Jerger, D.E., Dolenc, D.A., Chynoweth, D.P., 1982. Bioconversion of Woody Biomassas a Renewable Source of Energy. Institute of Gas Technology, Chicago, IL.

Li, Y., Park, S.Y., Zhu, J., 2011. Solid-state anaerobic digestion for methaneproduction from organic waste. Renew. Sust. Energy Rev. 15, 821–826.

Liew, L.N., Shi, J., Li, Y., 2011. Enhancing the solid-state anaerobic digestion of fallenleaves through simultaneous alkaline treatment. Bioresour. Technol. 102, 8828–8834.

Lin, J., Zuo, J., Gan, L., Li, P., Liu, F., Wang, K., Chen, L., Gan, H., 2011. Effects of mixtureratio on anaerobic co-digestion with fruit and vegetable waste and food wasteof China. J. Environ. Sci. 23, 1403–1408.

Manshardt, R., 2014. History and Future of the Solo Papaya, Genetics and Genomicsof Papaya. Springer, NY, pp. 95–113.

Midmore, D., Nguyen, V., 2003. Select Markets for Taro, Sweet Potato and Yam.Rural Industries Research and Development Corporation, Barton, ACT.

Narayanaswami, V., Sankar, K., Sekaran, P.M., Lalitha, K., 1986. Biomethanation ofLeucaena leucocephala: a potential biomass substrate. Fuel 65, 1129–1133.

Otieno, H.O., Awange, J.L., 2006. Alternative energy technology for East Africa. In:Otieno, H.O., Awange, J.L. (Eds.), Energy Resources in East Africa: Opportunitiesand Challenges. Springer Berlin Heidelberg, NY, pp. 101–204.

Parikka, M., 2004. Global biomass fuel resources. Biomass Bioenergy 27, 613–620.Scheffran, J., 2010. The global demand for biofuels: technologies, markets and

policies. In: Vertes, A.A., Qureshi, N., Blaschek, H., Yukawa, H. (Eds.), Biomass toBiofuels: Strategies for Global Industries. Blackwell Publishing Ltd., Oxford, UK,pp. 27–54.

Shiralipour, A., Smith, P.H., 1984. Conversion of biomass into methane gas. Biomass6, 85–92.

Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008.Determination of structural carbohydrates and lignin in biomass. NationalRenewable Energy Laboratory Technical Report NREL/TP-510-42618.

Sluiter, A., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2005.Determination of extractives in biomass. National Renewable EnergyLaboratory Technical Report NREL/TP-510-42619.

Templeton, D.W., Sluiter, A.D., Hayward, T.K., Hames, B.R., Thomas, S.R., 2009.Assessing corn stover composition and sources of variability via NIRS. Cellulose16, 621–639.

44 X. Ge et al. / Bioresource Technology 169 (2014) 38–44

West, P.W., 2014. Growth Rates and Wood Quality, Growing Plantation Forests.Springer, NY, pp. 25–44.

Xu, Z.H., Myers, R.J.K., Saffigna, P.G., Chapman, A.L., 1993. Nitrogen cycling inleucaena (Leucaena leucocephala) alley cropping in semi-arid tropics. Plant Soil148, 73–82.

Yang, P.Y., Weitzenhoff, M.H., Moy, J.H., 1984. Biogasification of papaya processingwastes. Trans. ASAE 27, 840–843.

Yu, Z., Schanbacher, F.L., 2010. Production of Methane Biogas as Fuel ThroughAnaerobic Digestion. In: Singh, O.V., Harvey, S.P. (Eds.), SustainableBiotechnology. Springer, NY, pp. 105–127.

Zhao, J., Ge, X., Vasco-Correa, J., Li, Y., 2014a. Fungal pretreatment of unsterilizedyard trimmings for enhanced methane production by solid-state anaerobicdigestion. Bioresour. Technol. 158, 248–252.

Zhao, J., Zheng, Y., Li, Y., 2014b. Fungal pretreatment of yard trimmings forenhancement of methane yield from solid-state anaerobic digestion. Bioresour.Technol. 156, 176–181.

Zheng, Y., Pan, Z., Zhang, R., El-Mashad, H.M., Pan, J., Jenkins, B.M., 2009. Anaerobicdigestion of saline creeping wild ryegrass for biogas production andpretreatment of particleboard material. Bioresour. Technol. 100, 1582–1588.