Characterization of Anaerobic Heterotrophic Bacteria - NCBI

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1976, p. 83-90Copyright 0 1976 American Society for Microbiology

Vol. 31, No. 1Printed in U.S.A.

Characterization of Anaerobic Heterotrophic BacteriaIsolated from Freshwater Lake Sediments1

JOHN J. MOLONGOSKI* AND MICHAEL J. KLUG

W. K. Kellogg Biological Station, Michigan State University, Hickory Corners, Michigan 49060

Received for publication 25 July 1975

Strict anaerobic culture techniques were used to quantitatively and qualita-tively evaluate the anaerobic heterotrophic bacteria present at the sediment-water interface of hypereutrophic Wintergreen Lake (Augusta, Mich.). Anaero-bic plate counts remained constant from March through December, 1973, rang-ing from 2.4 x 106 to 5.7 x 106 organisms/g (dry weight) of sediment. The iso-latable bacteria represented a small percentage of the total microbial commun-ity, which was shown by direct microscopic counts to be 2.0 x 1011 orga-nisms/g (dry weight) of sediment during June and July. Bacteria of the genusClostridium dominated the isolates obtained, accounting for 71.8% of the 960isolates examined. A single species, Clostridium bifermentens, comprised 47.7%of the total. Additional bacterial groups and the percentage in which they wereisolated included: Streptococcus sp. (10.8%), unidentified curved rods (9.5%),gram-positive nonsporing rods (5.6%), and motile gram-negative rods (1.9%).Temperature growth studies demonstrated the ability of all the isolates to growat in situ sediment temperatures. Gas-liquid radiochromatography was used todetermine the soluble metabolic end products produced from [U-14C]glucose anda U-14C-labeled amino acid mixture by representative sedimentary clostridialisolates and by natural sediment microbial communities. At in situ temperaturesthe natural sediment microflora produced soluble fermentative end productscharacteristic of those elaborated by the clostridial isolates tested. These resultsare considered strong presumptive evidence that clostridia are actively metabo-lizing in the sediments of Wintergreen Lake.

Development of methodology for the strictanaerobic cultivation of microorganisms (1, 12)has facilitated the isolation of obligate anaero-bic bacteria from a variety of anoxic habitats(11, 12, 15). Although these techniques arestandard procedure in investigations concerningmethanogenic bacteria in lake sediments (6,19), they have yet to be widely applied to theisolation of the nonmethanogenic heterotrophicbacteria present in these sediments. Applica-tion of the roll tube technique (12) to theenumeration of nonmethanogenic bacteria insewage sludge has increased viable counts 10 to100 times over those obtained using aerobictechniques or less stringent anaerobic methods,which fail to maintain anoxic conditions duringthe entire sampling procedure (15). These re-sults suggest the presence of substantial num-bers of obligately anaerobic as well as faculta-tively anaerobic bacteria in this habitat.

Like sewage sludge, the initial substrates insediments are complex organic compounds not

I Kellogg Biological Station publication no. 290 andMichigan Agricultural Experiment Station journal articleno. 7312.

utilizable by methane bacteria. The diversity oforganic inputs to lake sediments suggests thepresence of a corresponding, physiologicallydiverse, heterotrophic microbial community ac-tive in the anaerobic dissimilation of thesesubstrates (7).The present study represents an attempt to

quantify and characterize the predominantgroups of anaerobic heterotrophic bacteriapresent at the sediment-water interface of aeutrophic lake, using strict anaerobic culturetechniques (1, 12). As a knowledge of the num-bers and types of bacteria isolated from a par-ticular habitat provides little information con-cerning the metabolic activities of these orga-nisms in the natural environment, gas-liquidradiochromatography (13) was used to assessthe soluble fermentative end products producedfrom [U-14C]organic substrates by natural sedi-ment microbial communities and by isolatesderived from these sediments. (Portions of thisinvestigation were presented at the 74th An-nual Meeting of the American Society for Mi-crobiology, 12-17 May 1974, Chicago, Ill., andat the 75th Annual Meeting of the American

83

84 MOLONGOSKI AND KLUG

Society for Microbiology, 27 April-2 May 1975,New York, N.Y.)

MATERIALS AND METHODSStudy site. All investigations were conducted on

Wintergreen Lake, a small hardwater basin locatedwithin the W. K. Kellogg Bird Sanctuary, Augusta,Mich. The lake is shallow, with a maximum depthof 6.3 m and a mean depth of 3.54 m. Annualmean primary productivity values identify Winter-green Lake as hypereutrophic, the latter designa-tion based on compression of the trophogenic zone toa point where light rather than available nutrientslimit productivity (22). The pelagic zone of Winter-green Lake is characterized by significant autoch-thonous organic input, the annual succession ofphytoplankton being punctuated in the summer bydense blooms of protein-rich blue-green algae. Thehypolimnion of the lake is anaerobic for nearly 7months of the year, with anoxic conditions extend-ing to 3 m during summer stratification. The de-velopment of extensive anoxia bordering on thephotic zone insures that the majority of mineraliza-tion of particulate organic matter in this basin issedimentary and anaerobic. All sediment samplessecured during the investigation were taken at adepth of 6 m or greater.

Bacteriological sampling procedures. Sedimentsamples for bacteriological analyses were securedwith a Plexiglas gravity corer (8.5 by 100 cm), whichpreserved the sediment-water interface. Sampleswere collected at approximately monthly intervalsfrom March through December, 1973. Cores wereimmediately plugged with air-tight rubber stoppersfor transport to the laboratory, where a subsamplewas taken using a sterile glass subsampler (3 by 23cm). The subsample was continuously gassed with amixture of 90% N2-10% H2 or 85% N2-5% C02-10% H2before being plugged with rubber stoppers, clampedin a press, and transferred to an anaerobic glovebox. The glove box atmosphere was initially 90%N2-10% H2, but was subsequently altered to 85%N2-5% C02-10% H2. No significant differences wereobserved in the types or numbers of bacteria iso-lated under the two gas atmospheres. The recoveryof organisms under a gas atmosphere lacking H2was not evaluated.A 1:10 dilution (vol/vol) of duplicate subsamples

of surface sediment (0- to 2-cm layer) was made inprereduced (PR) dilution fluid. The composition ofthe diluent varied with the gas phase used. Under aN2-H2 atmosphere, the diluent contained, in gramsper liter: (NH4)2S04, 0.5; MgSO4-7H20, 0.05; NaCl,1.0; K2HPO4, 6.5; KH2PO4, 3.5. Under N2-C02-H2,the dilution fluid contained, in grams per liter:K2HPO4, 6.5; KH2PO4, 3.5; Na2CO3, 2.5. In addi-tion, each diluent contained 0.05% cysteine-hydro-chloride-water and 0.2% of a 0.1% resazurin solu-tion. The final pH of each solution was 6.9.

The initial 1:10 dilutions were either stirred on amagnetic stirrer for 5 min or mixed in a Waringblender for 1 min. Appropriate additional 1:10 and1:1 dilutions were made in anaerobic dilution fluidand inoculum from six dilutions was each spread on

APPL. ENVIRON. MICROBIOL.

duplicate PR agar plates containing resazurin as anoxidation reduction indicator. Pour versus spreadplates were not evaluated. Media inoculated in-cluded 1% peptone-yeast extract-glucose (PYG),0.5% PYG, brain heart infusion (Difco, Detroit,Mich.), sediment extract supplemented with Tryp-ticase, yeast extract, and mineral salts, and Medium10 of Caldwell and Bryant (5) modified for use undera predominantly N2 atmosphere. The complete com-position of each medium is listed in Table 1.

Inoculated plates were incubated anaerobically at15 C for 14 days. The plates were placed in Anaero-bic Jars (BBL) containing a palladium catalyst(palladium-coated alumina pellets; Engelhardt Indus-tries, East Newark, N.J.). The jars were sealed andremoved from the glove box to a 15 C incubator.After 14 days, the jars were returned to the glovebox and opened, and the plates were examined.Quantitative counts were made on 1% PR PYG (de-termined in preliminary experiments to yield thehighest number of microorganisms of the four mediaused). Colonies selected for additional characteriza-tion and identification were obtained at bimonthlyrather than monthly sampling intervals and werechosen during periods of lake stratification, as wellas during spring and fall overturn.

Isolation of sedimentary anaerobes and presump-tive identification of strains. Plates of each mediumwere carefully examined for colony diversity. Fiftyto 60 colonies were randomly picked from eachmedium and transferred to 1% PR PYG plates.Although colonies were selected at random, aneffort was made to insure that representatives of alldistinct colony types present on the plates wereincluded in the colonies transferred. The PYGplates were incubated at room temperature (23 to26 C) within the glove box. Resulting colonies werestreaked onto 1% PR PYG plates, and isolatedcolonies were transferred to PR PYG slants. Isolateswere stored at 4 C until further characterizationscould be carried out.

The isolates were characterized by the methodsdescribed in the Virginia Polytechnic Institute(VPI) Anaerobe Laboratory Manual (10). All char-acterization tests were performed using PR media,except for the determination of oxygen sensitivitywhen incubation was aerobic. Characterizationtests were performed at 25 C except for temperaturegrowth studies. One percent PY was used as thebasal medium in all fermentation tests, with thevarious test substrates being added at a concentra-tion of 1% (wt/vol).

Analysis of soluble and gaseous fermentationend products. Volatile fatty acids formed in glucose-containing media were analyzed as described byHoldeman and Moore (10). Lactic acid was deter-mined by conversion to its propyl ester (boron tri-fluoride-propanol esterification reagent; AppliedScience Laboratories, Inc., State College, Pa.).Both volatile and nonvolatile acids were analyzedon a Packard model 409 gas chromatograph equippedwith a flame ionization detector. Samples wereseparated on a coiled stainless-steel column (2 m by0.3 cm outer diameter) packed with 10% SP-1200-1%H3PO4 on Chromosorb W (AW-DMCS, 80/100 mesh;

ANAEROBIC BACTERIA FROM LAKE SEDIMENTS

TABLE 1. Composition of media used for isolation of sediment anaerobesa

% in mediumb

Component Sdmn x1% PYG 0.5% PYG BHI Sediment ex- Modified M10

Peptone 1.0 0.5Yeast extract 1.0 0.5 0.4 0.1Glucose 1.0 0.5 0.1Cellobiose 0.1Soluble starch 0.1Trypticase 0.2 0.4Hemin 0.0002Volatile fatty acids (vol/vol) 0.31Salts solution (vol/vol)

jC 90 90 90 902d 2.0

CaCl2 2H20 0.06 0.06 0.06 0.06Brain heart infusion 3.7Sediment extract (vol/vol) 30eResazurin 0.0001 0.0001 0.0001 0.0001 0.0001Cysteine-hydrochloride-water 0.05 0.05 0.05 0.05 0.05Agar 1.5 1.5 1.5 1.5 1.5

a Final pH, 6.8 to 7.0.b Final percent as weight/volume or as indicated. BHI, Brain heart infusion.cComposition, in grams per liter: K2HPO4, 0.56; KH2PO4, 0.33; Na2CO3, 0.22; MgSO4-7H20, 0.11;

(NH4)2S04, 0.22; NaCl, 0.53; FeCl3, 0.005.d Composition, in grams per liter: (NH4)2S04, 1.0; MgSO4-7H20, 0.1; NaCl, 2.0; K2HPO4, 7.0.e Sterilized separately and added aseptically after autoclaving. Extract consisted of clarified supernatant

obtained after autoclaving 1 kg of sediment with 1 liter of distilled water.

Supelco, Inc., Bellefonte, Pa.). Helium was used ascarrier gas at a flow rate of 25 ml/min. Chromato-graphic operating conditions were: inlet tempera-ture, 200 C; oven temperature, 120 C; detector tem-perature, 180 C.

H2 and CO2 produced by individual isolates grownin 1% PYG broth under 100% N2 were separated on acoiled stainless-steel column (2 m by 0.3 cm outerdiameter). The packing material was silica gel(60/80 mesh; Applied Science Laboratories, Inc.). N2was used as carrier gas (25 ml/min). Operatingconditions were: inlet temperature, 130 C; oventemperature, 50 C; detector temperature, 60 C; fila-ment current, 150 mA.

Metabolism of labeled organic compounds bynatural sediment microbial communities. A 1:10dilution of the initial 2 cm of sediment was madeusing PR filter-sterilized (0.22-ALm pore size mem-brane filter; Millipore Corp., Bedford, Mass.) in-terstitial water as diluent. Ten milliliters of dilutedsediment was added to replicate flasks containingeither [U-_4C]glucose (50 ,ug, 40 ,Ci; Amersham/Searle, Arlington Heights, Ill.) or a U-_4C-labeledamino acid mixture (50 ,tg, 38 tCi; New EnglandNuclear Corp., Boston, Mass.), respectively. Un-labeled glucose and amino acid mixture (SigmaChemical Co., St. Louis, Mo.) were added to therespective flasks to give final substrate concentra-tions of 120 Mg/ml of sediment slurry. Control flaskswere killed by addition of 1 ml of 2% HgCl2. Theflasks were incubated anaerobically at 10 C for 72h.

The reactions were subsequently terminated by

addition of 0.2 ml of 6 N H2SO4 to each flask. Solu-ble metabolites were determined as described above,except that volatile fatty acids were analyzed bythermal conductivity detection (filament current,200 mA). Organic components of the chromato-graphic effluent were combusted by passage througha heated copper oxide tube, and the resulting [14C]-CO2 was collected by bubbling the effluent gas intoa vial containing 14 ml of methanol-ethanolamine(1-2.5) solution. Six milliliters of scintillation fluid(15 g of 2,5-diphenyloxazole, 1 g ofp-bis-(O-methyl-styrl)-benzene, toluene to make 1 liter) was added,and radioactivity was determined by liquid scintilla-tion counting (Beckman model LS-150 liquid scintil-lation spectrometer).

RESULTSQuantitation of anaerobic bacteria at the

sediment-water interface. Figure 1 depicts via-ble plate counts of anaerobic heterotrophicbacteria at the sediment-water interface ofWintergreen Lake from March through Decem-ber, 1973. The anaerobic heterotrophic com-munity showed little seasonal variation. Themean viable count obtained on 1% PR PYGduring this period was 4.24 x 106 organisms/g(dry weight) of sediment. The viable countsreached a maximum in June, corresponding tothe onset of significant oxygen depletion in thehypolimnion. Although fall overturn restoredoxic water to the hypolimnion by late October,

VOL. 31, 1976 85

86 MOLONGOSKI AND KLUG

the viable counts did not decline significantly,remaining at peak summer levels through No-vember.

Periodic direct microscopic counts were per-

formed on the sediment-water interface of Win-tergreen Lake during June and July. Micro-organisms were dispersed from the sedimentparticles by the method of Bohlool and Schmidt(3). Samples were stained with acridine orangeand examined by epifluorescence microscopy(9). A mean direct microscopic count of 2.03 x

1011 organisms/g (dry weight) of sediment was

obtained. Chaining rods and filaments, presentprimarily in discrete microcolonies, dominatedthe morphotypes observed in the samples ex-

amined. Lesser numbers of cocci and vibroid-

6

o 4

1 2

F M A M J J A S 0 N D

1973

FIG. 1. Viable plate counts of anaerobic hetero-trophic bacteria at the sediment-water interface ofWintergreen Lake from March through December,1973. Duplicate dilution series were spread on 1%PR PYG plates. Each point represents the meanvalue obtained from 24 plates.

APPL. ENVIRON. MICROBIOL.

shaped cells were generally evident as well.Characterization of anaerobic sedimentary

isolates. Table 2 summarizes the major groupsof facultative and strictly anaerobic bacteriaisolated from the sediment-water interface ofWintergreen Lake. Fermentation and addi-tional physiological characteristics of eachgroup are found in Table 3. Group I, tenta-tively identified as a Streptococcus sp., was

composed of gram-positive, facultatively anaer-

obic cocci, 1 to 1.5 ,um in diameter. The cellswere either spherical or dumbbell shaped andoccurred singly, in pairs, or in short chains. NoH2S or indole was produced, and fermentationend products in PYG included major amountsof lactic and acetic acids with trace amounts ofpropionic, butyric, and isobutyric acids. Strainsof group I fermented all sugars tested, producedneither H2 nor CO2 from PYG, and grew at 5 C.These isolates comprised 10.8% of the totalorganisms examined.Group II, comprising 9.5% of the total iso-

lates, consisted of unidentified, gram-negative,curved rods measuring 0.5 by 3 ,um. These

organisms were obligately anaerobic, producedH2S, and were indole negative. Fermentationend products included major quantities of aceticand propionic acids with lesser amounts ofbutyric, isobutyric, and isovaleric acids. Theseorganisms fermented only maltose among thesugars tested and grew at 10 and 15 C, but notat 5 C.By far the largest group of isolates obtained,

TABLE 2. Presumptive identification features and distribution of anaerobic bacteria in Wintergreen Lakesedimentsa

Morphological groupIsolated Gram re- Motility Oxygen Fermentation prod- Tentative identifica-

strains ( action tolerance ucts tion

I Gram-positive cocci, 1 10.8 + - F LApbi-b Streptococcusto 1.5,m

II Small curved rods, 0.5 9.5 - + A APbi-bi-v Unknownby 3 ,um

III Sporeforming rods 71.8 Clostridiuma 47.7 + + A Apbi-bi-vi-C C. bifermentensb 17.8 + + A Apbi-bvi-V C. sporogenesc 6.3 + + A ApBi-B C. butyricum

IV Gram-positive rodsa 4.4 + - A AB Eubacteriumb 1.2 + - F ND Unknown

V Motile gram-negative 1.9 - + F ND Unknownrods, 0.5 by 2 to 3

a Based on a total of 960 isolates. Reactions given are those for a majority of the strains within a group.Symbols and abbreviations: A, Strictly anaerobic; F, facultative; +, positive reaction; -, negative reaction;fermentation products-A, a, acetic; P, p, propionic; B, b, butyric; i-B, i-b, isobutyric; V, v, valeric; i-V, i-v,isovaleric; C, c, caproic; i-C, i-c, isocaproic; L, 1, lactic. Uppercase letters refer to major end products;lowercase letters refer to products produced in lesser amounts as described in the VPI Anaerobe LaboratoryManual (10). ND, Not determined.

ANAEROBIC BACTERIA FROM LAKE SEDIMENTS

TABLE 3. Fermentation and additionalphysiological characteristics of selected sedimentary

isolatesa

Group no.b

Characteristic IIII I

a b cc

Glucose a n w w aCellobiose a n n n aGalactose a n n n aLactose a n n n aMaltose a a w w aSucrose a n n n aUrease production ND ND _ _Indole production - - + - _H2S production - + + + +Catalase - _ _ _Nitrate reduction - - _ _ _H2 production - ND + + +CO2 production - ND + + +Growth at:d

SC + - - + -lOc + + + + +15C + + + + +

a Abbreviations: a, Acid reaction, terminal pH 5.5or less; w, weak reaction, terminal pH 5.5 to 6.0; n,no fermentation, terminal pH greater than 6.0; ND,not determined.

b Fermentative and additional physiological testswere performed on the first three groups of isolatesonly. Reactions given are those for a majority of thestrains within a group.

c Identification ofC. butyricum is tentative due tothe variability of response of these isolates to thebiochemical tests performed.

d Tubes demonstrating no visible growth after in-cubation for 30 days were scored negative.

representing 71.8% of the total, group III con-sisted of obligately anaerobic, sporeformingrods of the genus Clostridium. This group couldbe further subdivided into three subgroupsidentified on the basis of fermentation endproducts and additional biochemical reactionsas Clostridium bifermentens, C. sporogenes,and C. butyricum. C. bifermentens representedthe most predominant species obtained, com-prising 47.7% of the total isolates.The predominantly proteolytic nature of

these clostridial isolates was substantiated bytheir inability to ferment a variety of sugars,the sole exception being C. butyricum, whichfermented all of the sugars tested. All of theclostridial isolates produced hydrogen in PYGbroth and grew at environmental temperatures(10 C). C. sporogenes was able to grow at 5 Cas well.

Groups IV and V represented a minor por-tion of the isolates, together comprising only7.5% of the total. Group IV consisted of gram-

positive, nonsporing rods and included two sub-groups, the first an obligately anaerobic rod,which produced acetic and butyric acids asmajor fermentation end products and whichmay be related to a species of Eubacterium,and the second an unidentified, facultativelyanaerobic rod. Group V consisted of unidenti-fied, motile, gram-negative rods comprisingless than 2% of the total isolates.

Table 4 illustrates the hydrolytic capabilitiesof each group of isolates towards a number ofcomplex organic substrates. No cellulolytic ac-tivity was demonstrable, as shown by the in-ability of all of the isolates to hydrolyze eitherball-milled cellulose or carboxymethyl-cellu-lose. Moreover, only the presumptive Eubacter-ium, the group V isolates, and one subgroup ofclostridia, C. butyricum, fermented starch.Only two groups, the group V isolates and thepresumptive Eubacterium, exhibited lipase ac-tivity as demonstrated by the ability to hy-drolyze tributyrin.The majority of the isolates demonstrated

proteolytic capabilities. The proteolytic natureof the clostridia was evident from the ability ofC. bifermentens and C. sporogenes to hydrolyzeboth casein and gelatin. Only C. butyricumfailed to attack these substrates. The remain-ing strictly anaerobic isolates, the group IIcurved rods and the group IV Eubacterium,also demonstrated proteolytic capabilities,whereas among the facultative isolates onlymotile gram-negative rods of group V were ableto liquify gelatin.

Metabolism of labeled organic compoundsby natural sediment microbial communities.Table 5 illustrates the labeled volatile fattyacids produced by the natural sediment micro-bial community when incubated anaerobicallyat 10 C for 72 h in the presence of [U-14C]glucoseor a U-14C-labeled amino acid mixture, re-spectively. Although the specific activity ofeach individual labeled acid is unknown, it isnonetheless evident that 14C was incorporatedinto an array of soluble metabolites by thenatural sediment microflora. Of particular in-terest is the appearance of label in thoseregions corresponding to isovaleric, valeric,isocaproic, and caproic acids. These metabolitesare highly characteristic of clostridial fermenta-tions, particularly those carried out by C.bifermentens and C. sporogenes (10, 18).

DISCUSSIONThe viable plate counts of anaerobic hetero-

trophic bacteria reported in this study are sim-ilar to those reported in the sediments of otherlakes of similar trophic level as WintergreenLake. Boylen and Brock (4) reported anaerobic

VOL. 31, 1976 87

TABLE 4. Hydrolytic characteristics of anaerobic isolates"

Carboxy-Casein diges- Gelatin Starch hy- methylcellu- Cellulose TributyrinGroup tion liquefaction drolysis lose diges- digestion hydrolysis

tion

I Facultative cocci - - - - - -II Anaerobic curved rods ± + - - - -III Clostridium

a + +b + + _ _ _ _cb _ + _ _

IV Gram-positive rodsAnaerobic - + + - - +Facultative -

V Gram-negative faculta- - + + - - +tive rods

a Symbols: +, Positive reaction; -, negative reaction. Reactions given are those for a majority of thestrains within a group.

b Identification of C. butyricum is tentative due to the variability of response of these isolates to thebiochemical tests performed.

TABLE 5. Metabolism of [U-'4Cjglucose andU-'4C-labeled amino acid mixture by natural

sediment microbial communities"

Substrate (counts/min)

Volatile fatty acid U-'4C-[U-'4C]glucose labeled amino

acids

Acetic 1,473 653Propionic 1,230 380Isobutyric 1,168 412Butyric 869 384Isovaleric 1,106 464Valeric 2,121 818Isocaproic 1,977 1,111Caproic 1,620 740

a Incubated anaerobically with shaking at 10 Cfor 72 h.

heterotrophic microbial communities of 4.4 x105 to 2.8 x 106 organisms/g (wet weight) ofsediment in Lake Wingra, Wisconsin, duringthe winter. Surface sediment heterotrophicbacterial densities in the lower Great Lakeshave consistently ranged between 106 and 107organisms/g (dry weight) of sediment, regard-less of the medium used or the time of sampling(2, 7, 21). Similarly, no significant differenceswere noted in the quantitative or qualitativerecovery of microorganisms on the four mediaused in the present investigation. The largediscrepancy observed between the viable anddirect microscopic counts of the bacterial com-munity of Wintergreen Lake sediments reflectsthe inadequacy of most media used to enu-merate sediment microorganisms, as well asthe difflculties associated with the enumerationof organisms adhering to a particulate surface,

such as sediment or soil. The high percentageof obligate anaerobes isolated in the presentstudy by strict anaerobic culture techniquessuggests that aerobic or less stringent anaero-bic culture techniques may not be adequate toenumerate substantial portions of the sedimentmicroflora. Additional bacteriological investi-gations of natural sediments and anoxic waterswhich use strenuous anaerobic methodologyare required to determine the relative numbersand metabolic activities of obligate as well asfacultative anaerobes in these habitats.The most striking feature of the bacterial

strains isolated from Wintergreen Lake sedi-ments is the predominance of clostridia amongthe organisms examined. The genus Clostridiumaccounts for 71.8% of the total isolates. Greaterthan 50% of the clostridial strains, in turn,were identified as C. bifermentens and C. sporo-genes (47.7 and 17.8%, respectively). Althoughtoxicity tests were not performed, C. sporo-genes was nevertheless identifiable due to itscharacteristic "Medusa-head" colonial morph-ology. A saccharolytic isolate tentatively iden-tified as C. butyricum represented a minorportion of the clostridial isolates (6.3%). Thisorganism could not be identified with certaintydue to the small number of isolates obtainedand the variability of response of these isolatesto the biochemical tests performed.The largely proteolytic nature of the clos-

tridial isolates obtained is likely due to thepredominance of nitrogenous organic sub-strates in Wintergreen Lake sediments. Asindicated, this proteinaceous material is de-rived primarily from extensive blooms of blue-green algae which dominate the phytoplanktonof the lake. The presence of a large ammonify-

APPL. ENVIRON. MICROBIOL.88 MOLONGOSKI AND KLUG

ANAEROBIC BACTERIA FROM LAKE SEDIMENTS

ing bacterial population in eutrophic LakeErie sediments (7) similarly suggests theavailability of a ready source ofcomplex nitrog-enous substances in enriched lake sediments.Although clostridia are generally held to be

widely distributed in nature (20), very little isknown concerning their ability to actively growand metabolize in natural habitats. Studies oftheir occurrence in freshwater and marinesediments have for the most part been limitedto considerations of these habitats as reservoirsfor clostridial spores pathogenic to humans andother animals (14, 17). Matches and Liston (16)examined the sediments of Puget Sound for thepresence of clostridia. They reported anaerobicplate counts ranging from 0.73 x 104 to 23.5 x

104 cells/ml of sediment-water slurry. Approxi-mately 30% of these organisms were deter-mined to be clostridia, the three species isolatedin greatest numbers being C. perfringens, C.bifermentens, and C. novyi. These workerspostulated that the mesophilic clostridial popu-

lation of Puget Sound sediments is passivelyaccumulated from terrestrial sources and is notderived from active growth of a metabolizingresident population. They based their findingslargely on an inability of most clostridia togrow at in situ sediment temperatures (16).The three clostridial isolates obtained from

Wintergreen Lake were able to grow at 10 C.C. sporogenes also grew at 5 C, the lowest tem-perature tested. Since the temperature of thesediment-water interface of Wintergreen Lakeexceeded 10 C for nearly 6 months of 1973,reaching a maximum of 13.3 C, clostridia couldbe actively metabolizing in these sedimentsduring this period. Finne and Matches (8) haverecently reported the isolation of low-tempera-ture growing clostridia from marine sedimentsthat can grow at 5 C or less, well within thein situ temperatures of most marine and fresh-water sediments.The production of characteristic clostridial

fermentation end products from [U-14C]glucoseand a U-14C-labeled amino acid mixture by nat-ural sediment microbial communities providesadditional presumptive evidence that clostridia,and possibly C. bifermentens and C. sporogenesin particular, are active in the degradation oforganic substrates in Wintergreen Lake sedi-ments. C. bifermentens and C. sporogenes elabo-rate an array of characteristic volatile metabo-lites, including the distinctive valeric andcaproic series of fatty acids. With the excep-tion of a few species of anaerobic cocci(Peptostreptococcus sp.) and a small number ofBacteroides sp. and Eubacterium sp., the pro-duction of valeric and isovaleric acids as

fermentative end products is generally limited

to the genus Clostridium (10, 18). Furthermore,production of major quantities of caproic andisocaproic acids as fermentative metabolites isalmost exclusively restricted to the clostridia(10, 18).

These results support the hypothesis that aresident clostridial community is actively me-tabolizing in Wintergreen Lake sediments,particularly during periods in the summer andearly fall when sediment temperatures aremost favorable for growth of the clostridia. Theability of these organisms to form spores wouldenable them to withstand periods less favorablefor growth, such as during lake turnover andmixing or during periods of reduced sedimenttemperature. Clearly, the role of clostridialspecies in the processing of organic matter innatural environments must be more thoroughlyexamined. The diverse modes of anaerobicenergy-generating metabolism displayed bymembers of this genus, their demonstratedability to grow at environmental temperatures,and their ability to form dormant spores when-ever environmental conditions become unsuit-able for growth should afford the clostridia adefinite advantage in organically enrichedanoxic habitats such as lake sediments.

ACKNOWLEDGMENTSThis investigation was supported by financial assistance

provided by the Agricultural Experiment Station of Michi-gan State University, and by National Science Foundationgrants GB-36069X and BO-15665 (Coherent Areas Programfor Investigation of Freshwater Ecosystems).

LITERATURE CITED1. Aranki, A., S. A. Syed, E. B. Kenney, and R. Freter.

1969. Isolation of anaerobic bacteria from humangingiva and mouse cecum by means of a simplifiedglove box procedure. Appl. Microbiol. 17:568-576.

2. Bell, J. B., and B. J. Dutka. 1972. Microbiologicalexamination of Lake Ontario sediments. I. Distribu-tion of aerobic and anaerobic heterotrophs in severalLake Ontario sediments. p. 1-8. In D. V. Andersonand J. S. Seddon (ed.), Proc. 15th Conf. Great LakesRes., vol. 15. Int. Assoc. Great Lakes Res., AnnArbor, Mich.

3. Bohlool, B. B., and E. L. Schmidt. 1973. A fluorescentantibody technique for determination of growth ratesof bacteria in soil. Bull. Ecol. Res. Commun.(Stockholm) 17:336-338.

4. Boylen, C. W., and T. D. Brock. 1973. Bacterial de-composition processes in Lake Wingra sedimentsduring winter. Limnol. Oceanogr. 18:628-634.

5. Caldwell, D. R., and M. P. Bryant. 1966. Mediumwithout rumen fluid for nonselective enumerationand isolation of rumen bacteria. Appl. Microbiol.14:794-801.

6. Cappenberg, T. E. 1974. Interrelations between sulfate-reducing and methane-producing bacteria in bottomdeposits of a freshwater lake. I. Field observations.Antonie Van Leeuwenhoek J. Microbiol. Serol.40:285-295.

7. Dutka, B. J., J. B. Bell, and D. L. S. Liu. 1974. Micro-biological examination of offshore Lake Erie sedi-ments. J. Fish. Res. Board Can. 31:299-308.

8. Finne, G., and J. R. Matches. 1974. Low-temperature-

89VOL. 31, 1976

90 MOLONGOSKI AND KLUG

growing clostridia from marine sediments. Can. J.Microbiol. 20:1639-1645.

9. Francisco, D. E., R. A. Mah, and A. C. Rabin. 1973.Acridine orange-epifluorescence technique for count-ing bacteria in natural waters. Trans. Am. Microsc.Soc. 92:416421.

10. Holdeman, L. V., and W. E. C. Moore (ed.). 1972.Anaerobe laboratory manual. Virginia PolytechnicInstitute and State University, Blacksburg.

11. Holdeman, L. V., and W. E. C. Moore. 1972. Roll-tubetechniques for anaerobic bacteria. Am. J. Clin. Nutr.25:1314-1317.

12. Hungate, R. E. 1950. The anerobic mesophilic cellu-lolytic bacteria. Bacteriol. Rev. 14:149.

13. Karmen, A. 1969. Combined gas-liquid-chromatog-raphy and radioassay of 14C and 3H-labeled com-pounds, p. 465-482. In J. M. Lowenstein (ed.),Methods in enzymology, vol. 14. Academic Press Inc.New York.

14. Laycock, R. A., and A. A. Longard. 1972. Clostridiumbotulinum in sediments from the Canadian Atlanticseaboard. J. Fish. Res. Board Can. 29:443-446.

15. Mah, R. A., and C. Sussman. 1967. Microbiology ofanaerobic sludge fermentation. I. Enumeration of thenonmethanogenic anaerobic bacteria. Appl. Micro-

APPL. ENVIRON. MICROBIOL.

biol. 16:358-361.16. Matches, J. R., and J. Liston. 1974. Mesophilic clos-

tridia in Puget Sound. Can. J. Microbiol. 20:1-7.17. Matches, J. R., J. Liston, and D. Curran. 1974. Clos-

tridium perfringens in the environment. Appl. Micro-biol. 28:655-660.

18. Moore, W. E. C., E. P. Cato, and L. V. Holdeman.1966. Fermentation patterns of some Clostridiumspecies. Int. J. Syst. Bacteriol. 16:383-415.

19. Nelson, D. R., and J. G. Zeikus. 1974. Rapid method forthe radioisotopic analysis of gaseous endproducts ofanaerobic metabolism. Appl. Microbiol. 28:258-261.

20. Slepecky, R. A. 1972. Ecology of bacterial sporeformers,p. 297-313. In H. 0. Halvorson, R. Hanson, andL. L. Campbell (ed.), Spores V. American Society forMicrobiology, Washington, D.C.

21. Vanderpost, J. M., and B. J. Dutka. 1971. Bacterio-~logical study of Kingston basin sediments, p. 137-156In D. V. Anderson and J. S. Seddon (ed.), Proc.14th Conf. Great Lakes Res., vol. 14. Int. Assoc.Great Lakes Res., Ann Arbor, Mich.

22. Wetzel, R. G. 1966. Variations in productivity of Gooseand hypereutrophic Sylvan Lakes, Indiana. Invest.Indiana Lakes Streams 7:147-184.