-
Vol. 47, No. 4APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1984,
p. 850-8570099-2240/84/040850-08$02.00/0Copyright C) 1984, American
Society for Microbiology
General Method for Determining Anaerobic
BiodegradationPotentialt
DANIEL R. SHELTON't AND JAMES M. TIEDJE12*Departments of Crop
and Soil Sciences' and Microbiology and Public Health,2 Michigan
State University, East Lansing,
Michigan 48824Received 3 October 1983/Accepted 10 January
1984
A simple, generalized method was refined and validated to test
whether an organic chemical wassusceptible to anaerobic degradation
to CH4 + CO2. The method used digested sewage sludge diluted to10%
and incubated anaerobically in 160-ml serum bottles with 50 pug of
C per ml of test chemical.Biodegradation was determined by the net
increase in gas pressure in bottles with test chemicals over
thepressure in nonamended sludge bottles. Gas production was
measured by gas chromatography and by apressure transducer. The
latter method is recommended because of its speed, accuracy, and
low cost.Sewage sludge from municipal digesters with 15- to 30-day
retention times was found to be suitable. Thesludge could be stored
anaerobically at 4C for up to 4 weeks with satisfactory test
results. p-Cresol,phthalic acid, and ethanol are suggested as
reference chemicals to confirm sludge activity and
methodreliability. A revised anaerobic salts medium was developed
which minimizes problems of abiological gasproduction (CO2), avoids
precipitation, and meets the requirements of the anaerobic
microbiota. When>75% of the theoretical gas production was
observed, the chemical was judged to be degradable, and when30 to
75% of the expected gas was produced, it was termed partially
degradable. This method has beentested on more than 100 chemicals
of various physical properties and found to reproducibly
determineanaerobic biodegradation potential. Of the chemicals
tested, 46 were found to be anaerobically degraded.Sludges from
nine different municipal treatment plants were surveyed for their
ability to degrade ninechemicals which differed in susceptibility
to degradation. As expected, the sludges varied in whichsubstrates
they degraded, but this could not be correlated with influent waste
properties of the particulartreatment plants.
Most manufactured chemicals will pass through
anoxicenvironments, and in some cases they will reside in
thesehabitats for long periods of time. These anoxic
habitatsinclude sediments of all types, anaerobic waste
treatmentsystems, gastrointestinal tracts, poorly drained or
floodedsoils, and some landfills and groundwaters. To make
anenvironmental risk assessment for a chemical, it may beimportant
to determine the chemical's susceptibility to an-aerobic
biodegradation. Furthermore, information on anaer-obic
degradability is also requested under EnvironmentalProtection
Agency guidelines implementing the U.S. ToxicSubstances Control Act
and is of interest to the Organizationfor Economic Cooperation and
Development. To obtain thisinformation, a simple general screening
method for assessinganaerobic biodegradability is needed.Owen et
al. (12) provided the first description of such a
test method, drawing on previous gas measurement (11)
andincubation bottle (10) methods. They based their method
onmeasurement of the excess gas volume (CH4 + C02) pro-duced after
addition of a test chemical to an anaerobic seedincubated in sealed
bottles. The gas volume was measuredfrom the displacement of the
piston in a glass syringe whoseneedle had been inserted into the
bottle. Subsequently,Gledhill improved the method with the goal of
defining asimple protocol that could be established by the
American
* Corresponding author.t Published as Journal Article no. 11032
of the Michigan Agricul-
tural Experiment Station.t Present address: Department of Soil
and Environmental Sci-
ences, University of California, Riverside, CA 92521.850
Society for Testing Materials (ASTM) as a standard method(4). He
introduced the use of a pressure transducer tomeasure the gas
pressure, recommended 50 mg of carbon perliter to be the test
chemical concentration, and used 10%anaerobic sludge as described
by Healy and Young (5). Hedid not, however, evaluate each aspect of
the method todetermine whether it was optimum, evaluate the
variabilityand reproducibility, or evaluate the method against
othermethods, all of which are necessary to validate a
standardmethod. In this paper we report on this evaluation
anddescribe the further refinements which we feel are benefi-cial.
Furthermore, we report on chemicals which we foundto be degraded
anaerobically, on an improved anaerobicmedium, and on the
differences among municipal anaerobicsludges in their degradation
capacities.
MATERIALS AND METHODS
Source and characteristics of anaerobic sludge. Sludgesamples
were collected from primary or secondary anaero-bic digesters in 1-
or 2-liter jars, tightly capped, and stored at4C until use. Sludges
were from waste treatment plants inthe following mid-Michigan
communities: Adrian, Ann Ar-bor, Chelsea, Holt, Ionia, Jackson,
Mason, Portland, and St.Johns. Percent organic matter (total solids
x volatile solids)varied from 0.89% (Holt) to 1.99% (Jackson) with
a medianvalue of 1.53%. Average retention times varied from
17(Ionia) to 39 (Holt) days with a median value of 20 days.Inflow
varied from 1.6 x 106 (Chelsea) to 68 x 106 (Jackson)liters/day
with a median value of 4.4 x 106 liters/day.
Preparation of test bottles. Serum bottles of 160-ml capaci-ty
(described as 125-ml Wheaton serum bottles; American
on April 24, 2015 by 61721891
http://aem.asm
.org/D
ownloaded from
-
ASSESSING ANAEROBIC BIODEGRADABILITY 851
Scientific Products, McGraw Park, Ill.) were amended asfollows
with 50 ,ug of carbon per ml of the test compound:liquids were
dispersed via microsyringe; water-soluble solidswere dissolved in
water and then dispensed; water-insolublesolids were dissolved in
diethyl ether and dispensed, and theether was allowed to evaporate
(-2 h); insoluble polymerswere weighed out and added to serum
bottles as solids. Therevised anaerobic mineral medium (RAMM)
developed forthis study consisted of (per liter): phosphate buffer,
0.27 g ofKH2PO4 and 0.35 g of K2HPO4 (adjusted to pH 7.0);
mineralsalts, 0.53 g of NH4CI, 75 mg of CaCl2 * 2H2O, 100 mg ofMgCl
* 6H20, and 20 mg of FeCl2 * 4H20; and trace metalsmodified from
Zehnder and Wuhrmann (20), 0.5 mg ofMnCl2 * 4H,O, 0.05 mg of H3B03,
0.05 mg of ZnCl,, 0.03 mgof CuCl2, 0.01 mg of NaMo4 2H20, 0.5 mg
ofCoCl2 * 6H20, 0.05 mg of NiCl2 6H20, and 0.05 mg ofNa-SeO3. The
medium was autoclaved for 5 to 10 min todrive off 02 and then
cooled to approximately 35C whilesparging with a 10% CO-90% N2 gas
mixture passedthrough copper filings at 300C to remove traces of 02
(7).After cooling, 1.2 g of NaHCO3 and 0.5 g of Na2S *
9H2O(optional) per liter were added to the medium. A 10%
dilutedsludge was prepared by adding 1 part of sludge
filteredthrough one layer of cheesecloth to 9 parts of cooled
mineralmedium while stirring. The diluted sludge was dispensed
intothe serum bottles while sparging with an 02-free 10% CO-90% N2
gas mixture. Methods for anaerobic gassing ofbottles and for
preparation of 02-free gases were essentiallythose of Hungate (7).
Serum bottles were sealed with 1-cm-thick butyl rubber stoppers
(Bellco Glass, Inc., Vineland,N.J.) and capped with aluminum crimp
seals (AmericanScientific Products). All compounds were tested in
triplicatewith the exception of the compound survey, for
whichcompounds were tested in duplicate. Bottles were
incubatedstationary and in the dark at 35C.Measurement of gas
production. After the medium had
equilibrated to 35C, the bottles were vented by syringeneedle to
atmospheric pressure (generally there was 1- to 3-ml overpressure).
Total gas production (CH4 + CO2) wasmeasured by a UniMeasure
pressure transducer (GrantsPass, Ore.) equipped with a P-8 adapter
(bellows) capable ofmeasuring up to 8 lb/in2 of gas pressure (Fig.
1). The needlewas inserted through the stoppers of the serum
bottles andthe signal (in milliohms) from the transducer was
quantifiedwith a Fluke multimeter (Mountlake Terrace, Wash.).
Serumbottles were shaken vigorously before pressure measure-ments
were taken, and excess gas pressure was ventedafterwards through
the three-way valve to avoid cumulativegas pressures beyond the
response range of the P-8 adapter.The milliohm response was related
to milliliters of gasproduced by a standard curve constructed by
adding knownquantities of gas to serum bottles by syringe; the r2
was>0.999%. Net gas production was calculated by subtractinggas
produced in unamended bottles from that produced intest bottles.
Degradation is expressed as percentage oftheoretical gas production
based on the stoichiometry ofmineralization to CH4 + CO2 and
correcting for gas solubili-ties.
Methane production was quantified by injecting 0.3 ml
ofheadspace gas from serum bottles into a gas chromatographequipped
with a flame ionization detector. Net methaneproduction was
calculated by subtracting background meth-ane production in
unamended bottles from that in testbottles. Degradation is
expressed as percentage of theoreti-cal methane production based on
the stoichiometry of degra-dation.
-D
Qe
FIG. 1. Gas production was measured with a pressure transduc-er
(A) equipped with a P-8 bellows (B). The transducer wasconnected
via 10 cm of 1/16-in. (ca. 0.16-cm) stainless-steel tubingand a
1/16-in. Swagelok union to a Hamilton three-way valve (C)with a
20-gauge needle attached. The signal from the transducer
wasquantified with a multimeter (D).
Experimental. To determine the effect of length of sludgestorage
on assay results, sludge from primary digesters inJackson, Holt,
and Chelsea sewage treatment plants wasstored, sealed at 4C.
Incubations were begun by taking newcontainers of sludge from
storage after 0, 1, 2, 3, and 4weeks. Ethanol, p-cresol, phthalic
acid, and di-n-butylphtha-late at 50 p.g of C per ml were used as
test compounds.The effect of oxygen intrusion on the assay was
investigat-
ed by replacing equal volumes of headspace gas with oxygenadded
by syringe. Oxygen concentrations were quantifiedwith a gas
chromatograph equipped with a thermal conduc-tivity detector.
Calculations. Since all substrates are provided at 50 pg ofC per
ml, the theoretical gas yield from 100 ml of medium forall
substrates is 10.5 ml. This gas will be divided betweenCO2 and CH4
based on the stoichiometry of the reaction,which can be calculated
by the Buswell equation (17):
a bCnHaOb + n ----H.O0
n a b n_ _
- + -CO,+_2 8 4 2
a b+ -- - CH48 4
Knowledge of the mole fraction of each gas is necessary ifonly
CH4 is measured or if total gas is measured because ofthe high
water solubility of CO2. To correct for solubility,multiply the
theoretical milliliters of CH4 x 0.95 and those ofCO2 x 0.35 to
obtain the expected gas in the headspace.These constants were
determined empirically and are accu-rate only for the temperature
(35C), aqueous and gasvolumes, and medium composition of the method
recom-mended here. If the substrate has carboxyl groups that
wereneutralized when added to the test bottles, 1 CO2
perneutralized carboxyl group should be subtracted from
thestoichiometry given by the Buswell equation since this groupdoes
not contribute to the gas phase; if the group(s) was inthe acid
form, it should remain in the equation.
VOL. 47, 1984
on April 24, 2015 by 61721891
http://aem.asm
.org/D
ownloaded from
-
852 SHELTON AND TIEDJE
RESULTSPreliminary experiments were initiated to obtain
degrada-
tion data for a wide variety of organic compounds to
selectcompounds for future testing of assay parameters and
todetermine the minimum length of incubation. The assayconditions
were as described except that sludges fromsecondary digesters in
Adrian and Jackson were used.Compounds were initially incubated for
4 weeks; however,this proved to be inadequate. Subsequently, all
compoundswere incubated for 8 weeks, or until net methane
productionhad ceased. Of 94 compounds tested, 27 were
mineralized(>75% of theoretical methane production) in at least
onesludge (Table 1); of these, 8 had lag times of >2 weeks.
Tencompounds were partially mineralized (>30 to
-
ASSESSING ANAEROBIC BIODEGRADABILITY 853
TABLE 2. Effect of substrate concentration on gas production in
10% sludge from the Jackson digesterMean % of theoretical
degradation t SD
Substrate concn(~Lg of carbon per ml) Phenol p-Cresol Benzoic
Phthalicgocarbnperml)Penol
-Cresoacid acid25 100 9.9 98 92 18.0 105 4.850 104 8.0 86 13.5
92 6.0 104 + 18.7100 106 6.3 98 5.5 96 4.7 109 3.7200 113 3.1 95
4.5 98 3.6 100 1.5
Background gas production 31.4 31.6 24.5 28.4(ml)
Theoretical gas production 7.4 7.6 6.5 5.7from 50 ,ug of carbon
perml (ml)"a Corrected for gas solubilities.
times, and extent of degradation in Jackson, Ionia, and Holt
virtually all of the excess gas production was observed in
theprimary sludges. An inhibition of background gas production
first week of incubation, well before the end of the typical lagwas
observed in the ASTM medium (Table 4). Further period for this
compound (Fig. 3).experimentation indicated that an inverse
correlation existed Sludge storage had no significant effect on
extent ofbetween added sulfide and background gas production (data
degradation, but lag times before degradation began werenot shown).
There was no significant effect on lag times with affected for
compounds that were degraded more slowly.any of the media; however,
the extent of degradation did This is illustrated for p-cresol
(Fig. 4), where lag timesvary (Table 4). Percent theoretical gas
production was increased from 4 to 7 weeks for Holt sludge, 2 to 3
weeks forconsistently higher in the ASTM medium than in the Jackson
sludge, and 4 to 4.5 weeks for Chelsea sludge, lagRAMM, whereas the
supplemental medium generally yield- times for ethanol, which is
readily degraded, were unaffecteded the lowest percent theoretical
gas production. The effect (data not shown).was particularly
pronounced with phthalic acid and m- Effect of oxygen on gas
production. The rate of oxygenchlorobenzoic acid (both amended into
sludges as the mon- consumption by 10% sludge was investigated
since oxygenoacid), where percent theoretical gas production in the
could inadvertently enter the test bottles and affect testASTM
medium exceeded 150%. With m-chlorobenzoic acid results. Adrian
sludge consumed approximately 1 ml of 02
TABLE 3. Comparison of mineral salts and metals in anaerobic
media versus 10% sludgeConcn (mM, minerals; ,uM, metals) in:
MineraUmetal ASTM Survey of Survey of sludges'RAMM medium'
anaerobic
media' Range Median Mean
MineralK 6.0 3.5 2.0-36.2 0.02-2.8 0.3 0.5NH4+ 10.0 4.3 3.2-42.4
0.03-15.4 0.4 2.1P042- 4.0 0.3 0.3-23.6 0.7-18.9' 0.9' 4.4"Ca 0.5
0.3 0.05-1.0 1.9-20.5 5.0 5.9Mg 0.5 1.8 0.05-1.8 0.05-2.3 0.8 1.0Fe
0.1 1.85 0.007-1.85 0.07-11.2 0.9 1.2S2- 0.5 2.0
MetalMn 2.53 101.0 0.15-101.0 4.0-530 21 30Zn 0.37 14.7
0.35-15.4 7.0-1,740 120 210Cu 0.22 17.2 0.06-20.1 5.0-650 60 90Co
2.10 126.0 0.42-126 0.2-1.2 0.5 0.6Ni 0.21 0.08-0.21 0.1-240 6.0
30B 0.81 97.1 0.81-97.1 5.0-280 14.0 37Mo 0.04 41.7 0.04-41.7
1.0-1.3 1.3 1.2Se 0.29 0.29-119.4
NaHCO3 1.2 g/liter 2.64 g/literResazurin 1 mg/liter 1
mg/literCO,-N2 10%:90% 30%:70%a From reference 4.b From references
3, 5, 6, 9, 13, 15, 18, 19, 20, and 21.' Calculations based on a
median total solids of 4.1%.d Total phosphorus.
VOL. 47, 1984
on April 24, 2015 by 61721891
http://aem.asm
.org/D
ownloaded from
-
854 SHELTON AND TIEDJE
TABLE 4. Effect of three mineral salts media on gas production
in 10% sludgeMean % of theoretical gas production (+ SD)
Sludge/medium p-Cresol Phthalic acid Di-n-butyl- n-Chlorobenzoic
Background gas
Jackson sludgeASTM medium" 102 3.2 156 0 128 3.9 30 + 8.6
13.5RAMM 108 3.9 130 5.0 101 9.7 0 18.8Supplemental medium' 79 7.9
85 7.7 89 14.8 0 17.9
Holt sludgeASTM medium 106 10.3 145 + 19.0 59 0 153 6.1 9.6RAMM
104 4.5 112 13.7 46 3.2 101 3.5 18.1Supplemental medium 89 4.6 96
2.9 19 5.4 65 5.2 19.0
lonia sludgeASTM medium 127 7.1 183 9.5 117 4.7 50 9.4 16.2RAMM
115 9.5 118 11.4 72 + 3.6 15 10.4 20.1Supplemental medium 99 6.7
104 24.8 77 + 16.7 0 21.6" Added 3.6 g of NaHCO3 per liter to be in
equilibrium with the 30% CO2 specified instead of the 2.64 g of
NaHCO3 per liter indicated in ref-
erence 4.bContained 6 mM P0427 9 mM K+, 10 mM HN4', 10 ,uM Co,
10% CO,-90% N2 headspace.
per day (Fig. 5). The upper portion of the sludge suspensionin
serum bottles containing a headspace gas mixture of -8%02 was
lightly pink due to the oxidation of resazurin (Fig. 5).The effect
of 0, intrusion on gas production was tested in
sludge from the Holt primary digester. In serum bottleswithout
substrate, approximately 4 ml less gas was producedin bottles
injected with 4 ml of 0, than in bottles with no O2,due to 02
consumption (Fig. 6). In serum bottles containingbenzoic acid plus
4 ml of 0, approximately 6 ml less gas wasproduced. With serum
bottles receiving no benzoic acid ascontrols, the percent
theoretical degradation of benzoic acidwas 115% in bottles
receiving no 02 and 87% in bottlesreceiving 4 ml of 02. After 16 h,
bottles receiving 4 ml of 02(7% O, in the headspace) were slightly
pink at the gas-waterinterface; however, the pink disappeared if
the bottles wereslightly agitated. No pinkish color was evident
after 40 h.
Capacity of different anaerobic sludges to degrade
selectedcompounds. Sludges from primary digesters at nine
sewagetreatment plants were compared for their ability to
degradenine test compounds (Table 5). Ethanol, polyethylene
glycol
11
9E]ASTM4.. HOLTE0LD 7 /5-RAMM
5-0~
z1/3-D
z 1/ 7 RAMM .- _
20,000, p-cresol, and phthalic acid were degraded in allsludges.
m-Cresol and di-n-butylphthalate were degraded inapproximately half
of the sludges, whereas 2-octanol, m-chlorobenzoic acid, and
propionanilide were degraded inthree sludges or less. Percent
degradation was generally>75% for m-cresol, p-cresol, phthalic
acid, and m-chloro-benzoic acid. The lower percent degradation for
polyethyl-ene glycol 20,000 and 2-octanol may be in part due to
theirslower rates of degradation such that an 8-week incubationwas
not sufficient to allow for complete methane recovery.We suspect
that low methane recoveries for propionanilidewere due to
metabolism of only the propionate moiety.
DISCUSSIONThe only feasible approach to a generalized method
of
screening organic compounds for anaerobic biodegradationis to
measure their common terminal products, CH4 + CO2.
H
6-IoQ H
28 |5 l |0a.
cot;wWLI '
1 2 3 ALENGTH OF STORAGE (weeks)
FIG. 3. Comparison of two mineral salts media (RAMM versusASTM)
on gas production from mn-chlorobenzoic acid in Holt andIonia
sludges.
FIG. 4. Weeks required for 50% of net gas production from
p-cresol as affected by length of storage of sludges from Jakcson
(J),Holt (H), and Chelsea (C).
-14 5WEEKS
APPL. ENVIRON. MICROBIOL.
on April 24, 2015 by 61721891
http://aem.asm
.org/D
ownloaded from
-
ASSESSING ANAEROBIC BIODEGRADABILITY 855
w
< 13.3
10.0 }
COLORLESS0
ui. 6.70
z
w
3.3
w
00 2 4 6 8
DAYSFIG. 5. Rate of consumption of different 0 concentrations
by
10% sludge. The conditions where the resazurin was pink is
shownby the stippled area.
A number of approaches to measuring these digestion prod-ucts
have been used, but only recently have these evolvedtowards a
routine test method. We began our work byevaluating the Owen et al.
(12) and closely related ASTMprovisional (4) methods since these
seemed to be the mostpromising of the existing methods.We found
that a 10% sludge inoculum will generally yield
10 to 40 ml of gas, depending on the retention time andpercent
organic matter of the sludge. An addition of 50 pg ofcarbon per ml
will theoretically yield 10.5 ml of gas, whereasthe actual gas
yield (corrected for solubility) may range from5.5 to 8.5 ml
depending on the stoichiometry of mineraliza-tion. Concentrations
of 50 pg of C per ml are notneeded for reliability and could lead
to more cases of toxicityby the test chemical. Although a
preincubation of the sludgeto reduce background gas production
would allow for use ofa less dilute inoculum or of concentrations
of test compoundof
-
856 SHELTON AND TIEDJE
TABLE 5. Percent theoretical methane production from nine
substrates by sludges from nine municipalities"% Theoretical
methane production
Sludge Ethanol Polyethylene p-Cresol Phthalic m-Cresol
Di-n-butyl- 2-Octanol m-Chlorobenzoic Propionanilideglycol 20,000
acid phthalate acid
Adrian 62 61 91 88 82 24 0 85 0Jackson 78 43 88 80 103 49 22 0
0Ann Arbor 86 IDb 101 132 85 91 58 0 36St.Johns 78 78 79 113 0 37 0
0 33Ionia 71 51 80 73 0 36 0 0 0Holt 33 53 88 86 77 ID 0 85 23Mason
94 82 94 96 91 0 87 0 0Chelsea 54 38 62 60 ID 0 0 0 0Portland 52 98
77 96 0 0 0 0 0
"Fresh sludge (10%) from primary digesters incubated for 8
weeks, after which methane was measured by gas chromatography.b ID,
Insufficient data; generally due to leaky bottles or only one
bottle showing evidence for degradation.
retention time and 0.89% organic matter. We recommendthat serum
bottles be incubated for a minimum of 8 Weeks.Lag times for
p-cresol and phthalic acid in fresh sludgevaried from 2 to 4 weeks;
thus, shorter incubation times arenot advisable.Anaerobic methods
are always subject to error from 02
contamination; however, the high O2-consuming capacity ofmost
sludges, the use of new, thick butyl rubber stoppers,and the use of
standard anaerobic methods should preventany serious errors. If 02
intrusion does occur it will reducethe gas pressure somewhat (due
to respiratory activity). Wedid not find it necessary to add
reductant (sulfide) to theanaerobic medium because of the high 02
scavenging capaci-ty of the sludge. Therefore, use of a reductant
is optional. Ifused, it should be in concentrations of
-
ASSESSING ANAEROBIC BIODEGRADABILITY 857
We have found gas production at -75% of theoretical
(aftercorrection for gas solubilities) to be indicative of
completemineralization.Summary of recommended protocol for
anaerobic biodegra-
dation test. (i) Use primary anaerobic sludge with 15- to 30-day
retention time and total organic solids of approximately1.0 to
2.0%. Sludges can be stored for up to 4 weeks at 4C intightly
capped containers; however, fresh sludge should beused whenever
possible. Since sludges vary in their selectedpopulations, it may
be useful to use more than one sludgewhen working with more
persistent compounds.
(ii) We suggest use ofRAMM since it has been thoroughlyevaluated
under the test conditions and minimizes precipita-tion. If another
anaerobic mineral medium is used, a 4 mMphosphate buffer and 1.2 g
of NaHCO3 per liter with a 10%C02-90% N2 headspace is recommended.
Sulfide can beadded to ensure reducing conditions in concentrations
not toexceed 1 mM, but it is not necessary.
(iii) Incubate a 10% homogeneous sludge solution with 50,ug of C
per ml of test chemical in 160-ml serum bottles withnew butyl
rubber stoppers and aluminum crimp seals. Eachchemical should be
tested in triplicate. A standard deviationof