Bioresource Technology 100 (2009) 45644571

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

The effect of enzyme addition on anaerobic digestion of Jose Tall Wheat Grass
Rowena T. Romano a, Ruihong Zhang a,*, Sarah Teter b, Jeffery A. McGarvey c
a
b
c

Department of Biological and Agricultural Engineering, University of California, 2030 Bainer Hall, One Shields Avenue, Davis, CA 95616-5294, USA
Novozymes Inc., Davis, CA 95616, USA
USDA/ARS/FCR, Albany, CA 94710, USA

a r t i c l e

i n f o

Article history:
Received 26 November 2007
Received in revised form 22 December 2008
Accepted 25 December 2008
Available online 24 May 2009
Keywords:
Cellulase
Wheat grass
Anaerobic digestion
16S rRNA

a b s t r a c t
The effects of the addition of enzyme products containing cellulase, hemicellulase, and b-glucosidase to
anaerobic digestion systems were studied using Jose Tall Wheat Grass (wheat grass) as a model substrate.
Anaerobic digestion tests were performed using batch reactors operated at 50 C. The application of
enzyme products in three digestion congurations were simulated and investigated: (1) enzyme addition
to a single-stage digester, (2) pre-treatment of wheat grass with enzymes followed by a single-stage
anaerobic digestion, and (3) enzyme addition to the rst stage (hydrolysis and acidication) of a twostage digestion system. The enzyme products showed positive effects on the solubilization of wheat grass
when used alone to treat the wheat grass. However, no signicant differences in biogas and methane
yields, and volatile solids reduction resulted when the enzyme products were tested in the anaerobic
digestion systems. This reveals that the microorganisms present in the inoculum were effective in carrying out the digestion of wheat grass. The types of microorganisms present in the inoculum were identied using 16S rRNA sequence analysis. A comparison of the sequences between the different inocula
revealed that the prevalent operational taxonomic units were similar, but that the acidied inoculum
contained a higher percentage of the species Thermotogae.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Anaerobic digestion is an attractive technology for the treatment of organic waste (e.g., manure, food-processing wastes and
green wastes). The microorganisms within an anaerobic digester
work synergistically to convert organic matter into biogas (methane (CH4) and carbon dioxide (CO2) that has an energy content
ranging from 18,630 to 26,081 kJ/m3 depending on the CH4 content. Biological degradation of lignocellulosic material is normally
facilitated by enzymes, such as cellulases and hemicellulases,
which are produced by the microorganisms. The rate-limiting step
for anaerobic digestion of lignocellulosic material is the hydrolysis
of cellulose and hemicellulose. Increasing the hydrolysis rate is
critical in order to improve the biomass-conversion efciency of
anaerobic digestion.
Previous studies have reported that adding enzymes into anaerobic digesters treating food-processing wastes resulted in improved digestion. Akao et al. (1992) reported enhanced anaerobic
digestion of citrus peels with an enzyme solution from Aspergillus
sp. A-1. The enzyme solution was reported to have cellulase and
pectinase activities that allowed the anaerobic digester to operate

* Corresponding author. Tel.: +1 530 752 9530; fax: +1 530 752 2640.
E-mail address: rhzhang@ucdavis.edu (R. Zhang).
0960-8524/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2008.12.065

at a higher organic loading rate. In another study, Sonakya et al.

(2001) pretreated wheat grains with Trizyme (cellulase, a-amylase, and protease) prior to anaerobic digestion and observed an increase in methane production by 714%. The addition of cellulase
(Celluclast 200L) and b-glucosidase (Novozyme 188) to anaerobic
digesters treating sewage sludge was reported by Higgins and
Swartzbaugh (1986). The system consisted of an enzyme pretreatment stage followed by anaerobic digestion and resulted in increased biogas and methane yields of 12% and 15%, respectively.
In another study, Nagle et al. (1992) treated previously-digested
sewage sludge with protease K (type XI, from Tritirachium album),
thermolysin (type X from Bacillus thermoproteolyticus), trypsin
(type XI, from bovine pancreas), and lysozyme (from chicken egg
white), which resulted in a 22% increase in soluble chemical oxygen demand (SCOD) after a 4.1 h treatment. Rintala and Ahring
(1994) investigated addition of Pulpzyme HA (xylanse and cellulase), Alcalase 2.5 LB (protease), and Resinase A 2X (lipase) to the
anaerobic digestion of source-separated household solid waste. In
batch treatments, enzymes were added individually or as a mixture, and the specic methane activity (SMA) was measured (dened as the slope of the cumulative methane production from
the initial 2030 h of digestion per volatile solids (VS) content of
the inoculum). Only the lowest concentration of protease applied
(0.5 ml/kg VS) resulted in signicantly higher (11%) SMA. However,
the authors reported no signicant differences in biogas yield or

4565

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

VFA concentration in the reactor efuent between enzyme-treated

and non-enzyme treated continuous digester systems.
The previous studies suggest that the addition of exogenous enzymes can improve the performance of anaerobic digestion systems. However, enzyme activity can be affected by many factors
including the substrate, incubation time, system conguration,
and environmental conditions (e.g. temperature and pH). More research is needed to determine if and when the addition of enzymes
such as cellulases and hemicellulases to the anaerobic digestion
system will improve digestion rates and biogas yields of lignocellulosic biomass. For example, enzymes could be added into a singlestage anaerobic digester or could be used to pre-treat the biomass
material prior to anaerobic digestion. In a two-stage anaerobic
digestion system, enzymes could be added to the hydrolysis stage
prior to biogasication. The research described in this paper was
aimed to address these questions using Jose Tall Wheat Grass
(wheat grass) as a model substrate because it is a salt-tolerant biomass crop in the San Joaquin Valley of California and has been largely used as animal feed and recently in the production of biofuels
and biobased products (Zheng et al., 2006). The objectives of this
study were to (1) investigate different methods of enzyme addition
to anaerobic digestion systems, (2) determine the effect of adding
enzyme products containing cellulase, hemicellulase and b-glucosidase to the anaerobic digestion of wheat grass, and (3) identify
the bacterial populations in anaerobic digesters capable of digesting wheat grass via 16S rRNA analysis.
2. Methods
2.1. Preparation and characteristics of wheat grass
Wheat grass was obtained from a farm near Five Points, California, and was cut, eld dried, and baled with an average straw
length of 50 cm and stored indoors at ambient temperature at
the Biomass Laboratory of the University of California, Davis until
use. The wheat grass was milled to 0.230.38 mm using a laboratory hammer mill (Model C269OYB, Franklin Co. Inc., Buffton, IN)
equipped with a 0.32-cm rejection screen and a sieve shaker (RO
TAP, The W.S. Tyler Company, Cleveland, OH) with corresponding
sieves (Newark Wire Cloth Co.), and stored in sealed 8-L zip-locked
plastic bags at 4 C prior to the anaerobic digestion tests. The total
solids (TS) and volatile solids (VS) contents were measured using
standard methods, and were 92% and 932 g/kg (dry solids), respectively (APHA, 1998). The cellulose, hemicellulose, and lignin contents (dry basis) 31.1%, 20.4%, and 20.4%, respectively, as
analyzed by the Department of Natural Resources Analytical Laboratory of UC, Davis (http://danranlab.ucdavis.edu/). The carbon to
nitrogen ratio (C/N) of the wheatgrass was 8.2, as analyzed by
Hazen Research, Inc. (Golden, CO).

end of the digestion period. The high loading of wheat grass allowed hydrolysis and acidication to occur, decreasing the sludge
pH from an initial 8.0 to 5.5. Citratephosphate buffers (10
20 mM) (Stoll and Blanchard, 1990) were prepared at pH 5 and
pH 7. All digestion experiments were conducted in triplicate, incubated in a 50 C chamber and mixed once a day.
Two enzyme samples, both multicomponent mixtures, were obtained from Novozyme Inc. (Davis, CA), and were selected based on
their ability to hydrolyze plant cell wall material, and the enzyme
pH proles. The rst enzyme set was Novozyme 342 (N342), which
contains cellulase and hemicellulase activities with an optimum
activity at pH 7. The N342 product was derived from Humicola
insolens and had an activity of 90 EGU/g (EGU, endoglucanase
units), 470 FXU/g (farvet xylan units), and 45 FBG/g (fungal b-glucanase units). The second enzyme set was a mixture of 85% Celluclast 1.5L (C15L) by weight and 15% Novozyme 188 (N188) by
weight. The C15L contained cellulase produced from Trichoderma
reesei and had an activity of 700 EGU/g and 70 FPU (lter paper
units). The N188 contained b-glucosidase from Aspergillus niger
and had an activity of 250 CBU/g (cellobiase units). Both C15L
and N188 had optimum activities at pH 5 and 50 C.
2.3. Experimental design
Four experiments were conducted to determine the effect of the
enzyme products on the digestion of wheat grass. First, wheat
grass was treated with the enzymes alone to determine their
hydrolytic effectiveness. Based on the results, the enzymes were
applied to different anaerobic digestion congurations to determine the effect of the enzymes on biogas and methane yields,
and volatile solids reduction. The three anaerobic digestion congurations tested were direct enzyme addition to a one-stage digester, enzyme pretreatment prior to anaerobic digestion, and enzyme
addition to the hydrolysis stage of a two-stage digester system. The
one-stage digester conguration (Fig. 1a) studied the synergistic
effects of the enzymes and microbial inoculum on the digestion
of wheat grass. The second conguration studied the pretreatment
of the wheat grass with enzymes (Fig. 1b) prior to anaerobic digestion. The two-stage conguration studied the synergistic effects of
the enzymes with the acidied microbial inoculum in the hydroly-

Enzyme
and
Inoculum
(pH ~7)

Enzyme
in buffer
(pH ~ 7)

One-Stage
Anaerobic
Digestion

Pre-treatment followed by
One-Stage Anaerobic
Digestion

Two-Stage Anaerobic
Digestion

(a)

(b)

(c)

Inoculum
(pH ~ 7)

Enzyme and
Inoculum
(pH ~ 5)

Inoculum
(pH ~ 7)

2.2. Anaerobic reactors, microbial inoculum, and enzymes

Batch digestion tests were conducted using 250-mL and 1000mL glass bottles hermetically sealed with a rubber stopper
clamped down with a plastic screw cap equipped with gas sampling ports. Anaerobic sludge taken from a single-stage thermophilic anaerobic digester located in Oakland, CA was the
microbial inoculum used in the anaerobic digestion tests. The
sludge was stored at 4 C and acclimated to 50 C prior to use. To
simulate the hydrolysis stage of a two-stage digester system the
anaerobic sludge was acidied in a batch digester treating wheat
grass at a loading of 70 g VS/L for seven days. The wheat grass
was held in nylon stockings, submerged into the sludge using
weights, and squeezed to remove any air in the stockings. The nylon stockings were used for easy removal of the wheatgrass at the

Fig. 1. The anaerobic digester congurations: (a) one-stage digestion with N342,
(b) enzyme pre-treatment with N342 followed by one-stage anaerobic digestion,
and (c) enzyme treatment with C15L and N188 added to hydrolysis stage of a twostage digester.

4566

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

sis stage prior to the biogasication (Fig. 1c). The experiments were
carried out at 50 C to reect the optimal temperature of the enzymes, and are lower than thermophilic conditions generally studied at 6065 C.
2.3.1. Direct treatment of wheat grass with enzymes
Two preliminary tests were carried out to determine the effect
of the enzymes on the hydrolysis of wheat grass at pH 5.0 and 7.0.
Treatment 1 tested the effect of enzymes C15L and N188 at pH 5.0,
and Treatment 2 tested the effect of enzyme N342 at pH 7.0. Control reactors were prepared containing no enzymes. The reactors
were lled with 140 mL of citratephosphate buffer prepared at
the required pH (5.0 or 7.0), loaded with 1.4 g VS of wheat grass
(10 g VS/L) and 35 mg of enzyme product (25 mg of enzyme product per gram of wheat grass VS). Signicant differences between
the treatment and controls were determined using the Data Analysis Tool Pak available on Microsoft EXCEL program.
2.3.2. Enzyme addition to different anaerobic digestion congurations
Treatment 3 examined the effect of the addition of enzyme
N342 to a one-stage digester operating at pH 7.0. The one-stage
digesters were prepared by adding 100 mL of the microbial inoculum (7.4 g VS/L), 100 mL of citratephosphate buffer, 2.0 g VS of
wheat grass (10 g VS/L), and 50 mg of enzyme N342 (25 mg/
g VS). The digestion was carried out for 14 days. Treatment 4 tested
the effect of enzyme pre-treatment of wheat grass at pH 7.0 prior
to anaerobic digestion. The pretreatment was performed by adding
2.0 g VS of wheat grass and 50 mg of enzyme N342 to 100 mL of
citratephosphate buffer. After 7 days of pretreatment, 100 mL of
microbial inoculum (7.4 g VS/L) was added to the reactors to induce anaerobic digestion for 14 days. Treatment 5 tested for the effect of the addition of enzymes C15L and N188 to the hydrolysis
stage of a two-stage digester conguration. Hydrolysis digesters
were prepared by adding 1.0 g VS of wheat grass (10 g VS/L) and
25 mg of enzymes (25 mg/g VS) to 100 mL of the acidied microbial inoculum (12 g VS/L), and the reaction was carried out for
7 days. After hydrolysis, the digester contents were transferred to
1-L reactors containing 400 mL of microbial inoculum, which was
not acidied and had a pH 7.0, to simulate biogasication for another 14 days. Due to the buffer capacity of the sludge, no pH
adjustment was necessary. The treatments are summarized in Table 1. Signicant differences between the treatments was determined using single-factor ANOVA and Turkey test using SAS
Statistics software (Cary, NC).
2.4. Chemical analyses
The amount of soluble compounds released from the wheat
grass was measured as soluble chemical oxygen demand (SCOD).
Samples of digester liquid were analyzed in duplicate using the
HACH COD method (Loveland, CO). The wheat grass and microbial
inoculum were analyzed for total solids (TS), volatile solids (VS),

xed solids (FS), and pH using standard methods (APHA, 1998).

Daily biogas production was determined by measuring headspace
pressure using a pressure transducer (WAL-BMP-Test system
3150, Oldenburg, Germany) and converting the pressure into volume using the ideal gas law corrected to standard conditions (temperature of 20 C and pressure of 1 atm). Biogas was analyzed for
CO2 and CH4 content using gas chromatography/thermal conductivity detection (GC/TCD) with an Alltech Carbosphere column
(Deereld, IL) and helium as the carrier gas. The GC oven temperature was 125 C, and the injector and detector temperatures were
125 C.

2.5. Microbial analyses

Bacterial 16S rRNA sequence analysis was performed on the two
starting microbial inocula (original sludge and acidied sludge) to
produce two sequence libraries. DNA was extracted from the cultures using the Mo Bio Soil Extraction Kit (Mobio, Carlsbad, CA).
The extracted DNA was amplied via PCR using the primers 27f
(50 AGAGTTTGATCCTGGCTCAG 3) and 1392r (50 GACGGGC
GGTGTGTAC 30 ) (Lane, 1991). PCR products were puried by ethanol precipitation, cloned using TOPO TA Cloning Kit (Invitrogen,
Carlsbad, CA), transformed into Escherichia coli and plated onto
LB agar plates containing kanamycin, isopropyl-b-D-thiogalactopyranoside, and 5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside
(X-gal). Transformed white E. coli colonies were grown in LB broth
containing kanamycin, and the plasmids were amplied using the
TempliPHI Amplication Kit (Amersham Biosciences, Sunnyvale,
CA). Sequencing reactions were performed using the primer
1392r and the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Forster City, CA). The sequencing reaction products were puried using the DYEX 96 Kit (Qiagen, Valencia, CA)
and sequenced using an Applied Biosystems 3730XL Genetic Analyzer (Applied Biosystems, Foster City, CA). For each sample 192
16S rRNA sequences were analyzed.
The 16S rRNA sequences were edited using Chromas (Technelysium Pty Ltd., 2006). Each 16S rRNA sequence was about
500 base pairs long. The diversity within the two libraries was
measured using the Shannon index (H), species richness (S),
and evenness (E) (Chang BioScience, 2004). Rarefaction analysis
was performed using the program aRarefactWin (Holland,
1998). The 16S rRNA sequence libraries were compared using Library Compare software (Ribosomal Database Project II (RDP))
(Michigan State University, 2006)) which estimates the likelihood
that the frequency of membership in a given taxon is the same
for the two libraries. The closest matches for the sequences in
GenBank were identied using the Basic Local Alignment Search
Tool (BLAST) available at the National Center for Biotechnology
Information webpage (NCBI) (NCBI, 2006). Genbank accession
numbers EU212999EU213008 were assigned for the 16S rRNA
sequences reported here (Table 8).

Table 1
Experimental design of batch digestion treatments.
Treatments

Conguration

Enzyme products

pH

Treatment 1

Enzyme hydrolysis

Treatment 2
Treatment 3
Treatment 4

Enzyme hydrolysis
One-stage anaerobic digestion
Enzyme pretreatment and anaerobic digestion

Celluclast 1.5 L and Novozyme

188
Novozyme 342
Novozyme 342
Novozyme 342

Treatment 5

Two stage anaerobic digestion with enzyme added into the rst stage
(hydrolysis)

Celluclast 1.5 L and Novozyme

188

7
7
7 (Pretreatment) 7 (anaerobic
digestion)
5 (Hydrolysis) 7 (biogasication)

4567

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

3. Results and discussion

1
0.9

The initial pH values were 5.3 for Treatment 1 and 7.5 for Treatment 2. The nal pH of the treatments after one week incubation
was 4.7 for Treatment 1 and 6.7 for Treatment 2, and is not significantly different from their controls (p > 0.05). Both Treatments 1
and 2 yielded signicantly greater SCOD than their respective controls (p < 0.05) (Fig. 2). Treatment 1 had a 94% difference in SCOD
yield from its control, and Treatment 2 had a 31% difference in
SCOD yield from its control. Therefore addition of the enzymes increased the release of the SCOD from wheat grass.

Biogas Y ield (L/ gVS)

3.1. Direct treatment of wheat grass with enzymes

0.8

With Enzyme

0.7

Without Enzyme

0.6
0.5
0.4
0.3
0.2

3.2. Enzyme addition to one-stage anaerobic digestion

Direct enzyme addition to a one-stage anaerobic digester system is an attractive design because it eliminates the need of additional reactors or equipment. Batch anaerobic digestion tests
showed biogas production increased rapidly within the rst eight
days and then decrease as digestion neared completion (Fig. 3).
The initial and nal pH of the treatment and control was 7.6 and
volatile solids reduction was below 50%. Biogas and methane yields
and volatile solids reduction were not signicantly different between Treatment 3 and the control (p > 0.05) (Table 2). The methane yields from wheat grass were similar to those of woody
biomass and other grasses, ranging from 0.15 to 0.30 L CH4/g VS
(Yu et al., 2002; Turick et al., 1991). For comparison, the methane
yields reported from food processing wastes such as fruits and vegetables digested in various anaerobic digestion systems range from
0.25 to 0.58 L CH4/g VS (van den Berg and Lentz, 1977; Knol et al.,
1978; Hills and Roberts, 1982; Lane, 1984). Higgins et al. (1986) reported that the solid reductions were similar between enzymetreated and non-enzyme-treated systems.
3.3. Enzyme pretreatment prior to anaerobic digestion
Enzyme treatment prior to anaerobic digestion was performed
to allow the enzymes to rst hydrolyze the wheat grass without
possible inhibitors from the microbial inoculum. During the pretreatment phase (07 days), biogas production was low, composed
mainly of CO2, and cumulative biogas production was not signi-

0.1
0
0

10

12

14

Fig. 3. Cumulative biogas production in the one-stage digestion conguration

(Treatment 3) with enzyme product N342.

Table 2
One-way ANOVA for biogas yield in the one-stage system.
Groups

Count

Sum

Average

Variance

Control 3 (no enzyme)

Treatment 3

3
3

1.56
1.50

0.52
0.50

0.00027
0.00093

Source of variation

SS

df

MS

P-value

F crit

Between groups
Within groups
Total

0.00048
0.00239
0.00287

1
4
5

0.00048
0.00059

0.808

0.420

7.709

cantly higher than the control (p > 0.05) (Fig. 4). The pH of the reactors after pretreatment was 6.4 and 7.4 for Treatment 4 and the
control, respectively. After addition of the anaerobic digester
sludge to the reactors the average pH increased to 7.0 and 7.7 for
Treatment 4 and the control, respectively. From days 8.5 to 10, biogas production in the treatment was signicantly higher than in

0.40

PRETREATMENT

ANAEROBIC DIGESTION

0.9

0.35

0.8

With Enzyme

Biogas Y ield (L/ gVS)

0.30

SCOD Yield (g / gVS)

Time (days)

0.25
0.20
0.15
0.10

0.7

Without
Enzyme

0.6
0.5
0.4
0.3
0.2

0.05

0.1

0.00
Treatment 1
(pH 5.0)

Control 1 (pH
5.0)

Treatment 2
(pH 7.0)

Control 2 (pH
7.0)

Reactor

0
0

12

15

18

21

Time (days)
Fig. 2. The SCOD yield from wheat grass treated with cellulosic enzymes with
optimal pH of 5.0 (Treatment 1 with C15L and N188) and optimal pH of 7.0
(Treatment 2 with N342).

Fig. 4. Cumulative biogas production for the pretreatmentdigestion conguration

(Treatment 4) with enzyme product N342.

4568

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

the control (p < 0.05) indicating a higher rate of biogas production

in Treatment 3; however, the difference decreased over time and
became non-signicant after day 10 (p > 0.05) (Table 3). The nal
pH of Treatment 4 decreased to 6.7 and the control decreased to
7.4. At the end of the digestion period (14 days), the nal biogas
and methane yields and volatile solids reduction did not signicantly differ between the treatment and control. The higher rate
of biogas production in Treatment 3 suggests that pretreatment
had a signicant effect on hydrolysis; however, it did not produce
a signicant effect in biogas and methane yields and volatile solids
reduction from the control.
3.4. Enzyme addition to the rst stage of a two-stage system
The two-stage system consisted of a hydrolysis stage followed
by a biogasication stage. Enzymes (C15L and N188) were added
with the acidied microbial inoculum to the hydrolysis stage of
wheatgrass. In the hydrolysis stage, biogas production was negligible after the second day of hydrolysis, and after seven days the
cumulative biogas production from Treatment 5 and the control
were signicantly different (p < 0.05) (Fig. 5). The nal pH of Treatment 5 and control after hydrolysis were both 5.6. After the hydrolysis stage, the contents of the reactors were transferred to the
biogasication stage, and the pH of the reactors increased to 7.6.
After the transfer biogas production was initially signicantly different up to day 8 (p < 0.05) and then again from day 11 to 13 biogas production was signicantly different (p < 0.05). However, by
the end of digestion there was no signicant difference in biogas
and methane yields and volatile solids reduction between the

Table 3
One-way ANOVA analysis for biogas yield in the pre-treatment system.
Groups

Count

Control 4 (no enzyme)

Treatment 4
Source of variation
SS

3
3

Between groups
Within groups
Total

Sum

df

0.00078
0.00053
0.00131

1
4
5

3.5. Microbial diversity

The lack of difference between the enzyme treatments and control digesters in the biogas and methane yields and volatile solids
reduction reveals that the microbial population alone was capable
of digesting a similar amount of wheat grass. However a recalcitrant fraction of the wheat grass remains in the digester and therefore other methods are necessary for further digestion of wheat
grass. Identication of the bacteria in the anaerobic digester sludge
would be helpful to elucidate other biochemical enhancement
strategies for anaerobic digestion. To characterize the bacterial

Variance

0.00018
0.00009
P-value
F crit

Table 4
One-way ANOVA analysis for biogas yield in the two-stage system.

5.93

0.072

Groups

Count

Sum

Average

Variance

Control 5 (no enzyme)

Treatment 5

3
3

1.33
1.61

0.44
0.53

0.00006
0.00473

7.71

Source of variation

SS

df

MS

P-value

F crit

Between groups
Within groups
Total

0.0130
0.0096
0.0226

1
4
5

0.0130
0.0024

5.43

0.080

7.71

nd

st

2 STAGE:
BIOGASIFICATION

0.8

Biogas Y ield (L/ gVS)

0.36
0.38

0.00078
0.00013

1 STAGE:
HYDROLYSIS

0.9

Average

1.08
1.15
MS

treatment and control (p > 0.05) (Table 4). In the biogasication

stage the treatment had a higher rate of biogas production in the
beginning over the control, indicating an effect of treating wheat
grass with enzymes. The biogas yield, methane yield and solids removal are summarized in Table 5.
The low biogas and methane yields and volatile solids reduction
implies the need for harsher treatment methods for wheat grass
prior to anaerobic digestion. In the production of sugars from lignocellulosic biomass for ethanol production, pretreatment methods prior to enzymatic hydrolysis are used to disrupt the cell
wall structure, making cellulose in particular more accessible to
enzymes. Pretreatment methods include steam explosion, liquid
hot water, ammonia explosion, aqueous ammonia recycle, controlled pH, dilute acid and lime treatments (Wyman et al., 2005,
Hu and Wen, 2007; Silverstein et al., 2007). Hu and Wen (2007)
reported that the total amount of sugar (glucose and xylose)
released from switchgrass was about four times higher when
pretreated with a microwave-assisted sodium hydroxide pretreatment prior to enzymatic hydrolysis. White rot fungi have also been
used to pre-treat bamboo culms to improve enzymatic hydrolysis
(Zhang et al., 2007).

With Enzyme
Table 5
Summary of biogas yield, methane yield, and volatile solids reduction for each
anaerobic digestion conguration studied. Values are presented as average coefcient of variation.

0.7
Without Enzyme

0.6
0.5

Treatment

Conguration

Biogas
yield
(L/g VS)

Methane
yield
(L/g VS)

Volatile
solids
reduction (%)

0.3

Treatment 3

0.35 6%

0.16 9%

44 7%

0.2

Control 3

0.36 3%

0.17 4%

46 12%

0.1

Treatment 4

One-stage digester
with enzymes
One-stage digester
without enzymes
Pretreatment with
enzymes
Pretreatment without
enzymes
Two-stage digester
with enzymes
Two-stage digester
without enzymes

0.38 3%

0.16 10%

45 4%

0.36 4%

0.15 7%

48 4%

0.53 18%

0.29 13%

32 2%

0.44 4%

0.22 3%

34 5%

0.4

Control 4

12

15

18

21

Time (days)
Fig. 5. Cumulative biogas production for two-stage digestion conguration (Treatment 5) with enzyme products C15L and N188.

Treatment 5
Control 5

4569

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

Table 6
Indices for microbial diversity in the original sludge (library S1) and acidied sludge
(library S2).

Table 7
Microbial populations in the original sludge (library S1) and the acidied sludge
(library S2).

Library

Phylum

Library S1 (%)

Library S2 (%)

Proteobacteria
Bacteriodetes
Thermotogae
Actinobacteria
Firmicutes
TM7
Unclassied

14.5
21.7
8.6
11.8
11.8
0.7
30.9

7.2
5.3
22.4
1.3
54.6
0.0
9.2

S1
S2

Number of
clones

Richness (Number
of OTUs)

Shannon
index

Evenness

152
152

44
45

2.66
3.04

0.70
0.80

Coverage
(%)
71
70

populations within the inocula, two libraries of 16S rRNA sequences were constructed: S1 was derived from the anaerobic
sludge that was used for the one-stage and pre-treatment reactors,
and S2 was derived from acidied anaerobic sludge that was used
for the hydrolysis stage of the two-stage system. A comparison of
the two libraries was made to determine what changes occurred
in the predominant phyla when the bacterial population is induced
to hydrolyze and ferment wheat grass.
Statistical analysis of the two libraries revealed that they were
very similar in regards to the richness and evenness; however, S2
had a larger Shannon index indicating greater diversity (Table 6).
Coverage was approximately 70% for both libraries, which is supported by the rarefaction analysis that suggests our sampling
was not exhaustive, but that most of the predominant OTUs (operational taxonomic units) were likely identied as the slopes of the
curves decrease greatly towards the end points (Fig. 6). Statistical
analysis of the 16S rRNA sequences revealed that both libraries
contained members of the phyla Proteobacteria, Bacteroidetes,
Thermotogae, Actinobacteria, and Firmicutes (Table 7). Both libraries
also contained sequences that were unclassiable and were likely
derived from novel organisms. At the 80% condence threshold,
30.9% of the sequences in the S1 library were unclassiable followed by sequences representing the phyla Bacteroidetes (21.7%),
Proteobacteria (14.5%), Actinobacteria (11.8%), Firmicutes (11.8%),
and Thermotogae (8.6%). The S2 library was signicantly different
from library S1 (p < 0 .01) for those sequences representing the
phyla Bacteroidetes, Thermotogae, Actinobacteria, TM7 and Firmicutes. Library S2 had a lower amount of unclassied sequences
(9.2%) compared to library S1. The most prevalent 16S rRNA sequences in library S2 resembled those of the phyla Firmicutes
(54.6%), followed by Thermotogae (22.4%) and Proteobacteria
(7.2%). Acidication of the thermophilic sludge decreased the
amount of Proteobacteria, Actinobacteria, TM7 and Bacteroidetes
and increased the number of Thermotogae and Firmicutes.
The top ve most prevalent OTUs in the libraries S1 and S2 are
presented in Table 8. S1-1 was identical to the uncultured bacterial
sequence EF559054 previously identied in a thermophilic anaerobic solid waste digester (NCBI, 2007). S1-2 was 99% similar to that
from the uncultured bacterial sequences EF559030 and CT573979
previously identied in anaerobic solid waste digesters (NCBI,

2007 and Chouari et al., 2005). The third most common sequence
S1-3 was 99% similar to that from the uncultured bacteria
EF559211 and 97% similar to that from Anaerobaculum mobile
(AJ243189). EF559211 was isolated from a mesophilic anaerobic
digester (NCBI, 2007) and AJ243189 from a wool-scouring wastewater treatment lagoon and is known to ferment glucose, organic
acids and glycerol to acetate, hydrogen and carbon dioxide (Menes
and Muxi, 2002). The fourth most common OTU, S1-4, was identical to Gordonia amarae (X80635) and AF020331. AF020331 was obtained from foaming activated sludge and anaerobic digesters (De
los Reyes et al., 1998). The fth most common OTU, S1-5, had a sequence 99% similarity to that of an uncultured bacterium
(AJ853533) that was reported in the leachate of a municipal solid
waste landll and was identied as a species of Bacillus (Huang
et al., 2005).
Several OTUs identied in S1 were also present in S2. For
example, S2-1 was identical to S1-2 and S2-3 was identical to
S1-1. S2-2 was 97% similar to Clostridium thermopalmarium
(X72869), which has been reported as a spore-forming anaerobic
thermophilic bacterium capable of producing butyrate (Rainey
et al., 1993). The S2-4 sequence was 99% similarity to the uncultured bacterium (DQ817799) reported in the gut microbiotas of
zebrash and mice (Rawls et al., 2006). The S2-5 was only 90%
similar to that from an uncultured Clostridiales (DQ069222),
reported in the deep subsurface of a sill composed of iron and
magnesium (NCBI, 2005).
The remaining OTUs for the S1 library had less than eight clones
each, and the S2 library had less than six clones each. The top sequences of both libraries had high similarity to uncultured sequences. Thermotogae was one of the most prevalent phyla in
both libraries; however, acidication of the sludge resulted in
greater numbers of these bacteria. Interestingly all cultured members of the Thermotogae are thermophilic or hyper-thermophilic
anaerobic bacteria (i.e., capable of growth only at temperatures between 45 and 80 C or above 80 C). However, recent studies have
identied Thermotogae 16S rRNA genes in mesophilic environments, including mesophilic anaerobic digesters (Sasaki et al.,

Fig. 6. Rarefaction curves of library S1 (original sludge) and library S2 (acidied sludge).

4570

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

Table 8
Summary of the most prevalent OTUs from the original sludge (S1-1S1-5) and acidied sludge (S2-1S2-5).
OTU

Assigned accession
number

Percent of total
clones (%)

Phylum (condence
threshold %)

Best match
in NCBI

Similarity (%)

Reported conditions (temperature,

system isolated from, carbon sources)

S1-1
S1-2
S1-3
S1-4
S1-5
S2-1
S2-2
S2-3
S2-4
S2-5

EU212999
EU213000
EU213001
EU213002
EU213003
EU213004
EU213005
EU213006
EU213007
EU213008

38.8%
9.2%
5.9%
5.3%
5.3.%
22.4%
17.8%
9.2%
5.3%
3.9%

Bacteriodetes (90)
Thermotogae (100)
Firmicutes (100)
Actinobacteria (100)
Firmicutes (80)
Thermotogae (100)
Firmicutes (100)
Bacteriodetes (91)
Firmicutes (100)
Firmicutes (100)

EF559054
EF559030
EF559211
X80635
AJ853533
EF559030
X72869
EF559054
DQ817799
DQ069222

100
99
99
100
99
99
97
100
100
90

T/AD/OM
T/AD/OM
M/AD/OM
T/WWTP/WW
/LF/MSW
T/AD/OM
T//
T/AD/OM
/gut microbiota/
/mac sill/

Temperature regimes: M, mesophilic; T, thermophilic. Systems: AD, anaerobic digestion; LF, landll; WWTP, wastewater treatment plant. Carbon sources: OM, organic
matter; MSW, municipal solid wastes; WW, wastewater.

2007; Leven et al., 2007, and Chouari et al., 2005). Nesbo et al.
(2006) identied 2 clades of mesophilic Thermotogae 16S rRNA sequences, one of which is closely related to the genus Petrotoga, to
which our sequences are homologous. All of the characterized Petrotoga have been isolated from oil wells and are able to ferment a
variety of simple and complex carbohydrates.
Analysis of the 16S rRNA gene sequences revealed the types of
bacteria capable of degrading lignocellulosic material present in
Jose Tall Wheat Grass. Our results showed that a signicant change
in the microbial populations occurred when the wheat grass was
loaded at a high concentration. Although our data is phylogenetic
and we have no physiological data on the organisms found under
these conditions, we speculate that these bacteria are selected
for due to their ability to degrade the cellulosic material present
in the digester. Future studies should include culturing these
microorganisms, such as the Thermotogae species identied, and
studying their abilities to degrade Jose Tall Wheat Grass in pure
cultures. If these organisms prove to be efcient at hydrolyzing
Jose Tall Wheat Grass, they may be used to enhanced biofuel
production.
4. Conclusions
The addition of the enzymes studied in this research to the
anaerobic digestion of wheat grass did not produce signicant effects on the biogas production after a 14-day digestion time. When
the enzymes were used to pre-treat wheatgrass prior to anaerobic
digestion or when added to the hydrolysis stage of a two stage digester, the rate of biogas production was signicantly affected,
however at the end of the digestion period there were no signicant improvements in the biogas yield, methane yield or solids
reduction. Although the enzyme products were capable of solubilizing wheat grass when used directly, when added with anaerobic
digester inoculum, the biogas yield, methane yield, and solids
reduction were not enhanced. The biogas and methane yields for
all the congurations ranged from 0.35 to 0.53 L/g VS, and 0.15
to 0.29 L/g VS, respectively, which are similar to values reported
for other woody biomass and grasses. The 16S rRNA sequence analysis determined that there were signicant differences between
the anaerobic inocula used to seed the reactors. The acidied inocula had a larger population of the species Thermotogae, and thus
further study into the activity and enzyme production of these bacteria may lead to enhanced biochemical degradation processes for
lignocellulosic material.
Acknowledgements
The authors would like to thank Novozyme, Inc. (Davis, CA) for
gift of enzymes, Jeremy Lathrop from USDA/ARS/FCR, Albany, CA

for his help in DNA sequencing, and Yi Zheng from UC Davis for
his help in processing the wheat grass.
References
Akao, T., Mizuki, E., Saito, H., Okumura, S., Murao, S., 1992. The methane
fermentation of citrus unshu peel pretreated with fungus enzymes.
Bioresource Technology 41, 3539.
American Public Health Association (APHA), 1998. Standard Methods for the
Examination of Water and Wastewater. American Public Health Association,
Washington, DC.
Chang BioScience, 2004. Shannon-Wiener Diversity Index/Shannon Entropy
Calculator. <http://www.changbioscience.com/genetics/shannon.html>.
Chouari, R., Paslier, D., Daegelen, P., Ginestet, P., Weissenbach, J., Sghir, A., 2005.
Novel predominant archael and bacterial groups revealed by molecular
analysis of an anaerobic sludge digester. Environmental Microbiology 7 (8),
11041115.
De los Reyes, M., De los Reyes III, F., Hernandez, M., Raskin, L., 1998. Quantication
of Gordona amarae strains in foaming activated sludge and anaerobic digester
systems with oligonucleotide hybridization probes. Applied and Environmental
Microbiology 64 (7), 25032512.
Higgins, G., Swartzbaugh, J., 1986. Enzyme addition to the anaerobic digestion of
municipal wastewater primary sludge. USEPA Water Engineering Research
Laboratory, Ofce of Research and Development. EPA/600/2-86/084, Cincinnati,
OH.
Hills, D., Roberts, D., 1982. Conversion of tomato, peach and honeydew solid waste
into methane gas. Transactions of the American Society of Agricultural
Engineers (ASAE) 25 (3), 820826.
Holland, S., 1998. aRarefactWin. University of Georgia, Athens GA, <http://
www.uga.edu/~strata/AnRareReadme.html>.
Hu, Z., Wen, Z., 2007. Enhancing enzymatic digestibility of switchgrass by
microwave-assisted alkali pretreatment. Biochemical Engineering Journal 38
(3), 369378.
Huang, L., Zhu, S., Zhou, H., Qu, L., 2005. Molecular phylogenetic diversity of bacteria
associated with the leachate of a closed municipal solid waste landll. FEMS
Microbiology Letters 242 (2), 297303.
Knol, W., Vander Most, M., Waart, J., 1978. Biogas production by anaerobic digestion
of fruit and vegetable wastes, a preliminary study. Journal of the Science of Food
and Agriculture 29 (9), 822830.
Lane, A.G., 1984. Laboratory-scale anaerobic digestion of fruit and vegetable solid
wastes. Biomass 5 (4), 245259.
Lane, D.J., 1991. 16S/23S rRNA sequencing. In: Stackebrandt, E., Goodfelllow, M.
(Eds.), Nucleic Acid Sequencing Techniques in Bacterial Systematics. Wiley and
Sons, New York, NY, pp. 115175.
Leven, L., Eriksson, A., Schnurer, A., 2007. Effect of process temperature on bacterial
and archaeal communities in two methanogenic bioreactors treating organic
household waste. FEMS Microbiology and Ecology 59, 683693.
Menes, R., Muxi, L., 2002. Anaerobaculum mobile sp. nov., a novel anaerobic,
moderately
thermophilic,
peptide-fermenting
bacterium
that
uses
crotonate as an electron acceptor, and emended description of the genus
Anaerobaculum. International Journal of Systematic and Evolutionary
Microbiology 52, 157164.
Michigan State University, 2006. Ribosomal Database Project II. <http://rdp.cme.
msu.edu/>.
Nagle, N., Rivard, C., Adney, W., Himmel, M., 1992. Efcacy of hydrolytic enzyme
addition and thermochemical pretreatments for increased secondary anaerobic
digestion of treated municipal sewage sludges. Applied Biochemistry and
Biotechnology (34/35), 737750.
National Center for Biotechnology Information, 2005. GenBank: DQ069222.1
Uncultured Clostridiales bacterium clone SRB47 16S ribosomal RNA gene,
partial sequence. <http://www.ncbi.nlm.nih.gov/Genbank/index.html>.
National Center for Biotechnology Information, 2006. Basic Local Alignment Search
Tool (BLAST). <http://blast.ncbi.nlm.nih.gov/Blast.cgi>.

R.T. Romano et al. / Bioresource Technology 100 (2009) 45644571

National Center for Biotechnology Information, 2007. GenBank: EF559054.1
Uncultured bacterium clone A55_D21_L_B_F07 16S ribosomal RNA gene,
partial sequence. <http://www.ncbi.nlm.nih.gov/Genbank/index.html>.
Nesbo, C.L., Dlutek, M., Zhaxybayeve, O., Doolittle, W.F., 2006. Evidence for the
existence of Mesotogas, members of the order Thermotogales adapted to lowtemperature environments. Applied and Environmental Microbiology 72, 5061
5068.
Rainey, F., Ward, N., Morgan, H., Toalster, R., Stackebrandt, E., 1993. Phylogenetic
analysis of anaerobic thermophilic bacteria: aid for their reclassication.
Journal of Bacteriology 175 (15), 47724779.
Rawls, J., Mahowald, M., Ley, R., Gordon, J., 2006. Reciprocal gut microbiota
transplants from zebrash and mice to germ-free recipients reveal host habitat
selection. Cell 127 (2), 423433.
Rintala, J., Ahring, B., 1994. Thermophilic anaerobic digestion of source-sorted
household solid waste: the effects of enzyme additions. Applied Microbiology
and Biotechnology 40, 916919.
Sasaki, K., Haruta, S., Ueno, Y., Ishii, M., Igarashi, Y., 2007. Microbial population in
the biomass adhearing to supporting material in a packed-bed reactor
degrading organic solid waste. Applied Microbiology and Biotechnology 75,
941952.
Silverstein, R., Chen, Y., Sharma-Shivappa, R., Boyette, M., Osborne, J., 2007. A
comparison of chemical pretreatment methods for improving saccharication
of cotton stalks. Bioresource Technology 98, 30003011.

4571

Sonakya, V., Raizada, N., Kalia, V., 2001. Microbial and enzymatic improvement of
anaerobic digestion of waste biomass. Biotechnology Letters 23, 14631466.
Stoll, V., Blanchard, J., 1990. Buffers: principles and practice. Methods in
Enzymology 182, 2438.
Technelysium Pty Ltd., 2006. Chromas. <http://www.technelysium.com.au/
chromas.html>.
Turick, C., Peck, M., Chynoweth, D., Jerger, D., 1991. Methane fermentation of woody
biomass. Bioresource Technology 37, 141147.
van den Berg, L., Lentz, C., 1977. Methane production during treatment of food plant
wastes by anaerobic digestion. In: Loehr, R.C. (Ed.), Food, Fertilizer, and
Agricultural Residues. Ann Arbor Science, Ann Arbor, pp. 381393.
Wyman, C., Dale, B., Elander, R., Holtzapple, M., Ladisch, M., Lee, Y., 2005.
Coordinated development of leading biomass pretreatment technologies.
Bioresource Technology 96, 19591966.
Yu, H., Samani, Z., Hanson, A., Smith, G., 2002. Energy recovery from grass using
two-phase anaerobic digestion. Waste Management 22, 15.
Zhang, X., Yu, H., Huang, H., Liu, Y., 2007. Evaluation of biological pretreatment with
white rot fungi for the enzymatic hydrolysis of bamboo culms. International
Biodeterioration and Biodegradation 60, 159164.
Zheng, Y., Pan, Z., Zhang, R., Jenkins, B., Blunk, S., 2006. Particleboard quality
characteristics of saline Jose Tall Wheat Grass and chemical treatment effect.
Bioresource Technology 98 (6), 13041310.