GASTROINTESTINAL MICROFLORA OF ANTARCTIC BIRDS'

JOHN McNEILL SIEBURTH

Animal Pathology Section, Virginia Polytechnic Institute, Blacksburg, Virginia

Received for publication September 10, 1958

Most of the information on the gastrointestinalmicroflora of antarctic birds was obtained duringthe great decade of South-Polar explorationfollowing the turn of the century. The reportof Levin (1899) that bacteria could not be foundin the intestines of arctic birds at Spitzbergenprompted similar studies by the naturalists andship's physicians on the Antarctic expeditions.Ekel6f (1908) on the 1901-1903 Swedish SouthPolar Expedition cultured the intestinal contentsof antarctic birds at Snow Hill Island withnegative findings except for two skua gulls.Gazert (1912) on the 1902-1903 German SouthPolar Expedition to Kerguelen and HeardIslands cultured stomach and intestinal contentsof birds and incubated both aerobically andanaerobically. No bacteria were obtained fromthe petrels and only one tern and one Ad6liepenguin yielded bacterial growth. Harvey Pirie(1912), during the 1902-1904 Scottish NationalAntarctic Expedition to Laurie Island and theWeddell Sea, made agar and gelatin stab culturesof stomach and intestinal material and obtainedbacterial growth from 12 of the 21 birds cultured.Dr. Charcot, who led the 1903-1905 FrenchAntarctic Expedition to the western side ofPalmer Peninsula, examined intestinal materialfrom gulls, petrels, and penguins and foundmany types, although in smaller numbers thanin temperate regions. Tsiklinsky (1908) identifieda few species from Charcot's cultures. McLean(1919) on the 1911-1914 Australasian AntarcticExpedition reported the absence of bacteria in6 of 14 antarctic birds at Commonwealth Bay.These investigations between 1901 and 1914mainly reported by species those birds that

1 This report is a part of a study requested bythe Antarctic Committee of the Society of Ameri-can Bacteriologists and by the American Instituteof Biological Sciences. Support was made availableby the Arctic Institute of North America (sub-contract ONR-191), the Hydrographic Serviceof the Argentine Navv, Virginia Polytechnic In-stitute, and the U.S. Office of Naval Research.

contained bacteria and those that were "bacterio-logically sterile." Bunt (1955) during the 1951-1952 Australian Antarctic Expedition to Mac-quarie Island repeated the classical qualitativeapproach of the earlier workers and reportedthat 2 of the 18 birds cultured were "bacterio-logically sterile."During the 1957-1958 Argentine Antarctic

Expedition through the South American Quad-rant of Antarctica, bacteriological studies wereconducted in a temporary bacteriology laboratoryaboard the icebreaker ARA General San MUartin(Sieburth, 1958). Birds were collected duringthe numerous opportunities to go ashore bylanding craft and helicopter. The primarypurpose of this study was to examine quantita-tively gastrointestinal material in an attemptto explain the reports of "bacteriologicallysterile" antarctic birds. Data are presentedwhich show the estimated bacterial population,predominant cultivable bacterial types, and theability of gastrointestinal material to enhanceor inhibit bacterial growth. Antibacterial dietaryfactors are described which apparently modifythe gastrointestinal flora and may explain theearlier reports of "bacteriologically sterile"antarctic birds.

MATERIALS AND METHODS

Instruments, special materials, and preweigheddyes and reagents in dry form were the onlymaterials taken to Argentina. All glassware andbulk chemicals were graciously supplied by theHydrographic Service of the Argentine Navy.With the cooperation of Dr. R. Copes, the ship'sphysician, a laboratory was set up in a smallpharmacy, that had a small working surface,refrigerator, sink, balance, and microscope. Ahot air oven, incubator, and autoclave wereavailable in the adjacent infirmary. All suppliesand materials had to be secure at all times dueto the 45 degree 8 second lateral roll cycle ofthe icebreaker in open sea, which was extremely

521

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SIEBURTH

hard on glassware and the maintenance ofbacterial cultures.The flying birds were shot and the penguins

were sacrificed by cervical dislocation. Thebirds were opened as aseptically as possible, theentire gastrointestinal tract was removed andsamples were aseptically taken from either theproventriculus or ventriculus, central intestine,and terminal intestine. One g samples of theingesta were triturated with Eugon broth (BBL)to yield 10 per cent suspensions for bacterialenumeration and Gram stained smears formicroscopic examination for the presence ofbacteria and predominant bacterial forms. Sincepoured petri plates readily became contaminatedwith mold, all work was done in tubed media.Serial decimal dilutions of the gastrointestinalsuspensions were made in broth and 0. 1-mlamounts of each dilution were smeared on twolong Eugon agar slants (total aerobe and anaerobecounts) and one long Tergitol-7 agar slant(Difco) for the coliform count. One ml of eachdilution was used to inoculate a tube of SFbroth (Difco) for enteric streptococci. Anaerobicconditions were obtained by the simple buteffective method of Parker (1955). All cultureswere incubated at 38 C for 24 hr since longerincubation did not appreciably alter the countsand incubation and storage space were limited.Representative colonies from the highest dilu-tions having growth were isolated in order tostudy the predominant bacterial types. Thebacterial isolates that remained viable at theconclusion of the voyage were studied accordingto the Manual of Microbiological Methods(Pelezar and Conn, 1957) and the cultures otherthan Enterobacteriaceae were identified accordingto Borgey's Manual of Determinative Bacteriology(Breed et al., 1957). The isolates of Enterobacteri-aceae were kindly identified by Dr. W. H. Ewing,Communicable Disease Center, Chamblee,Georgia.

In order to detect the presence of bacterialinhibiting and enhancing substances in the non-sterile gastrointestinal material, a procedurewas developed that utilized Millipore filtermembranes, pads, and plastic dishes. The brothsuspensions (approximately 1.6 ml) of thegastrointestinal material were used to soak thepad which was covered by a sterile hydrosolassay membrane that provided a barrier fromthe contaminated material and a sterile surface

upon which to streak the test bacteria. The testorganisms included a typical strain of Escherichiacoli, two delayed lactose fermenting strains ofE. coli, Micrococcus ureae, and a diphtheroid-like gram-positive rod which were all isolatedfrom skua gulls. The growth of the test organismson the broth control was compared with that onthe test material. This procedure was latermodified to test the antibacterial activity ofblood serum.

RESULTS

The birds were obtained between December5, 1957, and February 5, 1958, at the followinglocations: Ushuaia, Tierra de Fuego Island (1);Half Moon Island, South Shetland Islands (3, 4,5, 14); Almirante Brown Base (15) and HopeBay (8, 11, 12, 20) on Palmer Peninsula; LaurieIsland, South Orkney Islands (21, 28); SouthernThule Island (23) and Zavodovski Island (26,27) in the South Sandwich Islands; and GeneralBelgrano Base (24) and Halley Bay (25) on theWeddell Sea shelf ice.During the early part of this study the in-

testinal contents from the anterior, central, andposterior segments were pooled. When greatindividual differences were observed betweenbirds, each bird was studied in greater detail,separate observations being made on the pro-ventriculus, ventriculus, and central and terminalareas of the intestine. Rather than study asmany species as possible, a few species werestudied according to their feeding habits. Asexamples of scavengers and predators the skuagulls, sheathbills, and one giant fulmar werestudied. The penguins were studied as examplesof birds feeding on marine animals. The birdswere identified according to Murphy (1936).

Gastrointestinal flora of scavenging and predatorybirds. The results on eight flying birds withscavenging and predatory feeding habits aregiven in table 1. The cultivable bacterial popula-tion is expressed as the log of the number oforganisms per g of wet material. The effect ofthe gastrointestinal material on the growth ofthe test organisms is expressed as relative inhibi-tion (-) or growth enhancement (+). Thepredominant bacterial types found in eachgastrointestinal segment are shown. All fourskua gulls had an aerobic, facultatively anaerobicflora. Rapid lactose fermenting coliforms werepresent in all four birds although delayed lactosestrains were more numerous. The enteric group

522 [VOL. 77

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GASTROINTESTINAL MICROFLORA OF ANTARCTIC BIRDS

TABLE 1

Estimated bacterial population, bacterial activity, and predominant bacterial types in the gastrointestinalmaterial of antarctic flying birds

GastrointestinalSegment

Intestine

IntestineIntestineVentriculus

Central intestineTerminal intestine

IntestineIntestine

Proventriculus

Central intestineTerminal intestine

ProventriculusCentral intestineTerminal intestine

Estimated Bacterial Population

Aerobes

Total

3.5 X 105

2.0 X 1065.5 X 1032.5 X 106

1.5 X 1071.1 X 107

6.8 X 165<102

3.1 X 104

4.8 X 106

1.1 X 107

<102

<1022.0 X 1c2

Coliform

3.0 X 103

6.4 X 1052.5 X 1031.0 X 106

1.0 X 1071.1 X 107

7.3 X 103<102

2.4.X 104

3.3 X 1069.5 X 106

<102

<102

En-tero-cocci

< 101

<101<101<101

<101<101

<101

< 101<101

<101

<101

<101

< lG-<101<101

Anaerobes

3.5 X 105

5.0 X 1053.5 X 1036.1 X 106

1.2 X 1074.8 X 107

2.1 X 1068.0 X 102

1.5 X 105

8.2 X 1064.0 X 107

<102<102

8.0 X 104

Effect of Gastro-intestinalMaterial

on Growth of

0

0..

-9

-9

-9

-3

-3-9

-9

-9

-9

0

-3-6

+6

0

-9

-9

+9+9

-6

+60

+6

0

-9

Predominant BacterialIsolates

Achromobacter eury-dice, Micrococcusureae. BethesdaGroup, Diphtheroid-like.

E. coli (atypical)E. coli (atypical)E. coli (atypical and

typical)E. coli (typical)E. coli (typical) and

Streptococci sp.

E. coli (atypical)Anaerobic, gram-posi-

tive rodsE. coli (atypical),Diphtheroid-like

E. coli (typical)Aerobac'er-Hafnia sp.,Di phtheroid-like

-9 None+6 None+6 "Anaerobic"

cocci

strepto-

of streptococci were not detected in these or

any of the other birds studied although otherstreptococcal species were observed. The isolatesfrom bird no. 1 which was scavenging in theharbor at Ushuaia, were the most diversifiedof any of the birds studied. The simple flora ofskua gulls 4 and 8 from the more isolated basesconsisted of delayed lactose fermenting strains(atypical) of E. coli with typical strains in smallernumbers. The gastrointestinal material of bird26, which was preying on penguin chicks on

uninhabited Zavadovski Island, was quiteinhibitory for the gram-positive test organismsand gram-positive organisms were only foundin the terminal portion of the intestine. Theventriculus contents were inhibitory for typicalE. coli but not for atypical strains which were

the most numerous. The intestinal materialwas inhibitory for the atypical E. coli types

which were not detected, but not for the typicaltypes which were the most numerous.

Sheathbill no. 3 was scavenging on seal carcasses

near the sledge dogs at Half Moon Island. Theflora of this bird was also aerobic and facultativelyanaerobic. Although rapid lactose fermentingstrains of E. coli were present, atypical strainsconstituted the predominant flora. Bird 11,which was obtained in the Adelie penguinrookery at Hope Bay and was presumably feedingin the intertidal zone, was devoid of a detectableaerobic flora. However, a small anaerobic floraof gram-positive rods was present. The thirdsheathbill, no. 28, had been feeding on a mixtureof birds eggs, marine forms, and refuse from theLaurie Island weather station. The flora was

aerobic and facultatively anaerobic. Althoughall segments were inhibitory for gram-positiveorganisms, diphtheroid-like organisms were

Avian Species

Skua gull

Sheathbill

Giant fulmar

6-o

1

48

26

311

28

27

5231959]

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

524 SIEBURTH

present in both the proventriculus and terminalintestine. The contents of the proventriculusand central intestine were noninhibitory forE. coli which was the predominant organismin these segments. The material in the terminalintestine was very inhibitory to both typicaland atypical E. coli and an Aerobacter-Hafniaspecies was the predominant gram-negativeorganism present.The only giant fulmar studied was bird 27,

which like skua no. 26, was obtained on Zavo-dovski Island and had been feeding on penguinchicks. The proventriculus contents were veryinhibitory for all the test bacteria and no cultiva-ble bacteria were observed. Although the con-

[VOL. 77

tents of the central intestine were stimulatoryfor E. coli, this material also failed to yielddetectable bacterial growth. The contents inall three segments were very inhibitory forgram-positive bacteria. However, Gram stainedsmears indicated that an apparently homogeneouspopulation of large streptococci was present inall segments. This organism was cultivated poorlyaerobically and better anaerobically only inmaterial from the terminal intestine.

Gastrointestinal flora of penguins. The resultson eight penguins feeding on marine forms aregiven in table 2. All the penguins except theemperor penguin chick no. 25 were feeding onthe lobster krill Euphausia superba and were

TABLE 2Estimated bacterial population, bacterial activity, and predominant bacterial types in the gastrointestinal

material of antarctic penguins

GastrointestinalSegment

Intestine

Intestine

ProventriculusVentriculusCentral intestine

Terminal intestine

VentriculusCentral intestineTerminal intestineVentriculusCentral intestineTerminal intestine

ProventriculusVentriculusCentral intestine

Terminal intestineProventriculusVentriculusCentral intestine

Terminal intestine

VentriculusCentral intestine

Terminal intestine

Estimated Bacterial Population

Aerobes

Total

<102

<102

2.0 X 10'1.2 X 103

1.6 X 105

<1024.3 X 1041.3 X 106

<1024.9 X 1652.4 X 107

3.5 X 1037.0 X 1031.3 X 104

2.6 X 1042.2 X 1061.0 X 1033.5 X 103

4.0 X 10'

1.2 X 1041.8 X 105

9.2 X 107

Coliforms

<102

<102

<10'<102

<102

<102<102t<102<1C2<102t<102

<102<102<102

<102<102<102<102

<102

1.0 X 1032.4 X 104

2.1 X.1C7

En-tero-cocci

<10'

<10'

< 101< 10'

<10'

<101< 10'

< 10'<10'<10'< 1lC

<101<10'<10'

<10'<10'<10'<10'

<10'

< 10'< 10'

<10'

Anaerobes

2.5 X 104

1.2 X 103

4.8.X 1051.3 X 106

2.2 X 106

2.0 X 1021.0 X 1052.1 X 106

<1023.1 X 1053.7 X 1G7

3.6 X 1031.9 X 1045.5 X 104

5.5 X 1C47.5 X 1061.3 X 1071.5 X 1C6

2.6 X 106

1.4 X 1048.0 X 104

1.2 X 108

Effect of Gastro-intestinal Material

on growth of

En

0

C. ux

cd *-

-90

+7

0

+3-1.5+6

0

-9+7

0

+9 +6+9 +6+9 +6

+3+4+2-2

-9

-6-6

-9

+3+6

0+6

+6

-90

1-9

+6-3*+6

0

+6+6+6

+6+6

00

+3

00

-6

Predominant BacterialIsolates

Anaerobic, gram-posi-tive rods

Anaerobic, gram-posi-tive rods

Fastidious gram-nega-tive rods

E. coli (atypical)

NoneProvidence Group

Fastidious gram-nega-tive rods, Streptococ-cu8 equinus

Fastidious gram-nega-tive rods, S. equinus

Providence group, E.coli (typical)

* Inhibitory to own aerobic flora.f No lactose fermenting coliforms detected, lactose nonfermenting organisms, equal or higher than total aerobic count.

Avian Species

Adelie pen-guin

Gentoo pen-guin

6z-(:

5

12

20

14

15

Ringedguin

pen- 21

23

25Emperor pen-guin

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

1ASTROINTE'STINTAL MICROFLORA OF ANTARCTIC BIRDS

devoidl of typical strains of E. coli. As in theflying birds, no streptococci of the enteric groupwere detected. Adelie penguins 5 and 12 werefree of a detectable aerobic flora. Both of thesepenguins had a small but detectable anaerobicflora which consisted of pleomorphic gram-positive rods that grew poorly on subculture andwere soon lost. Adelie penguin no. 20 had a florain all segments which grew more luxuriantlyunder anaerobic conditions. These fastidiouslactose nonfermenting gram-negative rods soondied out on subculture. The material in theproventriculus was very inhibitory for gram-positive bacteria and that in the ventriculuswas very inhibitory for its own aerobic flora.These observations might explain the low aerobicpopulation in the central intestine and theabsence of gram-positive bacteria. D)espite thegrowth enhancing properties of the contentsof the central intestine for the test bacteria, thebacterial population did not increase one logplace throughout the gastrointestinal tract.The gentoo penguins (birds 14 and 15) came

from different areas, but shared a number ofcharacteristics. They were heavily infested withcestodes and the epithelial lining of the proventric-ulus and ventriculus was discolored green bythe phytoplankton in the stomach of the lobsterkrill upon which the birds were feeding. Althoughthe ventriculus contents had a negligible flora,Gram stained preparations revealed an abundantgram-negative eoccobacillary flora which failedto grow upon repeated cultivation. Rapid lactosefermenting coliforms were not detected, however,the Tergitol-7 slants contained a populationequal to or greater than that on the Eugon agarmedium. In bird 14, the flora consisted of alactose nonfermenting E. coli. The predominantflora of bird 15 belonged in the Providence Group.The ventriculus contents of this bird, whichinhibited four atypical E. coli strains, were freeof a cultural flora while the contents of thecentral intestine which were growth enhancingyielded a bacterial flora.The ringed penguins (birds 21 and 23) also

came from different areas and shared a numberof characteristics. There was both an aerobicand a facultatively anaerobic flora in each ofthe four segments studied. Neither of the birdscontained detectable numbers of coliforms. Thepredominant flora of both birds consisted of a

fastidious lactose nonfermenting gram-negative

rod and Streptococcus equinus. These penguinswere feeding on lobster krill that apparentlyhad not been feeding recently on phytoplankton.The gastrointestinal material was growth en-hancing rather than inhibitory.The emperor penguin chick, no. 25, was the

only penguin studied in detail which was notfeeding on lobster krill. The proventriculuscontained a fluid with black particulate matter,approximately 200 g of pebbles, and 10 pairs ofcephalopod beaks. The aerobic and anaerobicpopulations were similar and a cultivable florawas present in all segments despite the inhibitoryactivity of the proventriculus contents whichwere highly acidic (pH 1.35). All segments werequite inhibitory for the gram-positive testorganisms. The predominant organisms belongedto the providence group. Typical strains of E.coli were also numerous and were the onlyrapid lactose fermenting strains found in anyof the penguins studied. Emperor penguins cango for long periods without eating (at least 40days). In order to determine if a bacterial florapersists in such a starved bird, an adult emperorpenguin (no. 24) which had been held captiveat the General Belgrano Base without food for15 days was briefly studied. Oral and rectalswabs indicated an abundant flora at both endsof the gastrointestinal tract. Diphtheroid-likeand klebsiella-like organisms were isolated fromthe rectal cultures.

Biochemical activities of the predominantenterobacterial types. Of the 16 birds studied indetail, four had an obligately anaerobic flora.Three birds had a flora of fastidious gram-nega-tive rods with S. equinus being present in lowernumbers in two of these birds. The remainingnine birds had an aerobic, facultatively anaerobicflora in which members of the family Entero-bacteriaceae were the predominant types. Thebiochemical characteristics of these entero-bacterial types are given in table 3.

All nine birds had delayed or negative lactosefermenting types instead of, or in addition to,the rapid lactose fermenting types. Althoughrapid lactose fermenting coliforms were presentin six of the eight flying birds, typical strains ofE. coli only predominated in skua gull 26 andsheathbill 28. Only one of the eight penguins(no. 25) had typical strains of E. coli.The atypical E. coli strains showed different

carbohydrate fermentation patterns. Organisms

lt35'3] 525

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

TABLE 3Biochemical activities of the enterobacterial types that constituted the predominant aerobic intestinal

microflora of 10 antarctic birds

EnterobacterialType

Escherichia coli(typical)

Escherichia coli(atypical)

Providence group

Bethesda groupAerobacter-Hafnia

speciesKlebsiella-like

O AntigenGroup

OXI, 88

88

849811

8

an()0

x

AG

A

AgAGAGAg

AGAg

AG

0v]u

AG

A

aa

0

u

AG

AG AG

v

0

AG

A

AGAGAg

AGAg

AG

00

cd

AG

AgAGAG

AGAg

AG

+

+

++++

.0

+=>(1S5

+

++++

0

v v

to -d

4(L.-

0.

+

IL.2

=

- clv

+

+

- + + + --+ +

+ + + +

A = Acid, a = delayed acid; G = Gas; g = small amount of gas; + = positive reaction; and - =

hydrates were observed for 14 days.

from skua gull 26 and sheathbill 28 failed toproduce gas or attack mannitol. The strain fromsheathbill 3 attacked only dextrose and mannitolrapidly. The strains from skua gulls 4 and 8 weretypical except for delayed action on lactose.The strains from gentoo penguin 14 failed toattack lactose in two to three weeks and pro-duced hydrogen sulfide.The noncoli enterobacterial isolates fell into

four other groups. The gentoo penguin 15 andemperor penguin 25 which were obtained fromwidely separated areas had a predominant floraof Providence Group organisms. Skua gull 1living close to human habitation contained anorganism of the Bethesda group. Sheathbill 28,in addition to typical and atypical strains ofE. coli also contained an Aerobacter-Hafniaspecies. A klebsiella-like organism was obtainedfrom emperor penguin 24.

Antibacterial food materials. The results on thegastrointestinal microflora indicated that thediet of both the predatory birds and the penguinscontained antibacterial materials which wereapparently influencing the gastrointestinal micro-flora. In an attempt to determine if the anti-bacterial activity of blood serum could accountfor the inhibitory activity of gastric materialin the predatory birds, 12 sera were tested fortheir gram-negative antibacterial activity. Serumdilutions of 1:20, 1:40, 1:80, and 1: 160 inEugon broth were used to soak the Milliporefilter pads. One typical and two atypical strains

-00

P.40.

0n.

Source of Cultures

Skua gull (26),Sheathbill (28)

Skua gull (26),Sheathbill (28)

Sheathbill (3)Skua gull (4, 8)Gentoo penguin (14)Gentoo penguin (15)Emperor penguin (25)Skua gull (1)Sheathbill (28)

Emperor penguin (24)

negative reaction. All carbo-

of E. coli were streaked on hydrosol assayMillipore membranes placed on the nutrient pads.Growth was compared with that on the brothcontrols and the results are given in figure 1.All 12 sera were inhibitory at a 1:20 dilution,10 at 1:40, 7 at 1:80, and 5 inhibited at leastas high as a 1:160 dilution. Although the gram-positive test organisms were not used in thistrial, other tests indicated that penguin serumwas active against gram-positive bacteria.The proventriculus and ventriculus contents

of the penguins feeding on lobster krill had avariable activity against the test organisms.When the lining of these organs was discoloredwith phytoplankton, presumably from the gastriccontents of the krill, this activity was markedand appeared to modify the gastrointestinalflora. In an attempt to verify this observation,a number of live adult krill (Euphausia superba),about 2 in in length, were obtained at HalfMoon Island. The krill were dissected and thedifferent segments were stored in crushed icefor eight days until the San Martin returned.The effect of 10 per cent suspensions of the krillsegments on bacterial growth is shown in table4. Only the stomach contents of the krill wereactive. This material which consisted of phyto-plankton and was a bright green in color com-pletely inhibited the gram-positive test bacteria.During the trip into the Weddell Sea, a numberof immature krill about % in long were caught inthe plankton net. Since they also had green

526 SIEBURTH [VOL. 77

+

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GASTItOINTESTINAL MICROFLORA OF ANTARCTIC BIRDS

Figure 1. The inhibitory effect of blood serumfrom antarctic birds on gram-negative bacteria.

TABLE 4Effect of adult lobster krill (Euphausia superba) on

bacterial growth

Growth of Test Organismst

Krill Segment* Micro- Diphth- Esc.erichia colicoccus eroid- Eshrci oi (typi-ureae like (atypical) cal)

Broth control.. 2+ + 3+ 3+ 3+Eye......... 2+ + 5+ 5+ 5+Stomach-......_ 4+ 4+ 5+Muscle......... 2+ + 4+ 4+ 5+

* Ten per cent suspension of segments stored at0 C for 8 days.

t Relative growth compared to broth control:- = no growth; values greater than those of con-trol indicate growth enhancement.

stomach areas they were squashed between aMillipore filter pad and a Millipore membraneand the test organisms were streaked toward thekrill areas. Three of the five immature formstested also inhibited gram-positive bacteria. Itis interesting to note that in the absence ofinhibitory material from the phytoplankton,the krill material usually enhanced bacterialgrowth and pigmentation.The antibacterial activity of the lobster krill

appeared to be in their phytoplankton ladenstomachs. In order to determine if sea waterphytoplankton exerted a differential antibacterialactivity, sea water and plankton net samples were

obtained at the hydrographic stations made bythe San Martin in the Weddell Sea. The phyto-plankton in six water samples was concentratedon hydrosal assay Millipore membranes whichwere rinsed with sterile physiological saline or

autoclaved, filtered sea water, placed algae sidedown on Eugon broth soaked pads, and streakedon the upper sterile surface with the test bacteria.The phytoplankton concentrates from I to 500ml of three of the six samples tested were inhibi-tory. On three occasions the phytoplankton was

so dense that the plankton net was covered by a

bright green jelly-like material which appearedto consist almost exclusively of Halosphaeraalgae. Dry weight determinations were between4 and 6 per cent dry matter. Serial 10-folddilutions of the fresh material were made inEugon broth and 1-ml aliquots were concentratedon the Millipore membranes and treated in thesame manner as the concentrates from the watersamples. The weakest preparation obtainedinhibited the gram-positive test organisms at0.01 per cent and the gram-negative at 1.0 percent wet weight concentrations. The results on

the most active plankton net sample are givenin figure 2. As in the other samples, the gram-

positive bacteria were more susceptible than thegram-negative test organisms. This materialwas active at a concentration of 0.0001 per centor 100 parts per million on a wet weight basis.On a dry weight basis this would be 5 ppm whichindicates that an extremely potent antibacterialsubstance is present in some phytoplanktonsamples. As the algae became more concentrated,

0

0

4c

m

co

I-

0

0

U.U.

z Z

z a

0.00001 0.0001 0.001 0.01 0.1 1.0

PER CENT PHYTOPLANKTON (WET VOLUME)

Figure 2. The effect of antarctic sea waterphytoplankton on bacterial growth.

1959] 527

15. GENTOO 20. ADELIE 21. RINGED+1 PENGUIN PENGUIN PENGUIN

4

-i 04W 11:160

mC_D z t40SERUM1: 20 DILUTIONS

> 18. GENTOO 22. ADELIE 23. RINGED

l'- PENGUIN PENGUIN PENGUINI 0

WZ -I ' 1 1 l '_I_

z

-2

. SKUA 19. SKUA 25. EMPEROR+

GULL GULL PENGUINU.

o 0

-

+*I 26. SKUA 27. GIANT 28. SHEATHGULL FULMAR BILL

w l0

-2

w -3

-0-- GRAM P01TIVE BACTERIA -,--0-- GRAM NEGATIVE BACTERIA P

.~~~~ ~ ~ ~ ~ ~ ~ ~ ~~~~~1

0% /1

pYa 0.

q

I-r-

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SIEBURTH

the antibacterial activity was lost and thematerial even became growth enhancing.

DISCUSSION

One of the main bacteriological questions posedby previous investigations in the Antarctic wasthe existence of "bacteriologically sterile" birds.Five of the six expeditions reported, by species,birds from which bacteria could not be cultivated.McLean (1919) proposed that the food of thesebirds was sterile. Bunt (1955) suggested thatyeasts or protozoans might be causing completeinhibition of bacterial growth by direct competi-tion for nutrients or the production of anti-biotic substances. McBee (1957, unpublisheddata) cultured the intestinal contents of threebirds, found growth to be plentiful in agar shaketubes and absent from surface cultures, andconcluded that the earlier reports on the sterilityof the intestinal contents of antarctic birds werebased upon inadequate observations. Of thesixteen birds studied in detail in this study,four did not yield a detectable aerobic flora. Inthe scope of earlier work in which anaerobicprocedures were crude or were not used, thesebirds might have been regarded as being"bacteriologically sterile." However, an anaerobicintestinal flora was present in all birds studied.The proventriculus and ventriculus contents ofsome birds were free of a cultivable flora, eitheraerobic or anaerobic. Gram stained smears fromthese materials indicated an abundant bacterialflora. The presence of bacterial forms in foodmaterial from which neither aerobic nor anaerobicbacteria could be cultivated suggested thepresence of strong antibacterial substances inthe diet of these birds. A procedure was developedto estimate the bacterial activity of nonsterilesubstances in order to determine if antibacterialactivity of penguin tissues in the giant fulmar'sproventriculus contents, in which bacteria werepresent but did not grow, indicated that tissuefluids were exerting a bacteriostatic or bacteri-cidal activity. The antibacterial activity ofnormal serum is well known and has been re-cently reviewed by Skarnes and Watson (1957).Assays on the antibacterial activity of bloodserum from twelve antarctic birds indicated thatall were inhibitory at a 1:20 dilution and fivewere inhibitory at a 1:160 dilution. Assumingthat these birds, like chickens (Sturkie, 1954),

have approximately 8 per cent blood volumeand that serum constitutes 50 per cent of theblood volume, then a 4 per cent suspension or1:25 dilution of serum would occur in a maceratedbird. Therefore, there appeared to be a sufficientconcentration of inhibitory substances in serumto explain the antibacterial activity of the in-gesta of predatory birds. The antibacterialactivity of the penguin tissues in the giant fulmarappeared to not only inhibit its gram-positiveanaerobic flora, but to eliminate any possiblegram-negative aerobic flora. A differential anti-bacterial activity of normal food materials maybe responsible for the apparently aerobe-freebirds. This hypothesis is strengthened by theobservation that all four birds free of detectableaerobes had an anaerobic flora that consisted ofone morphological type.

Antibacterial activity in the gastric contentsof some of the penguins feeding on lobster krilldrew attention to the green discoloration of theproventriculus and ventriculus lining of thesebirds. The antibacterial activity and greenishmaterial in the lobster krill (Euphausian shrimpor Euphausia superba) was found to be in theirphytoplankton laden stomachs. Samples of seawater phytoplankton also exhibited markedantibacterial activity. Sea water phytoplank-ton, freshly caught krill, and the gastric contentsfrom penguins possessed a similar differentialantibacterial activity. Gram-positive bacteriawere apparently more susceptible than gram-negative bacteria. This differential antibacterialactivity against gram-positive bacteria mayexplain the predominant gram-negative micro-flora in antarctic birds which possess an aerobicflora.The antibacterial property of sea water phy-

toplankton may affect the bacterial populationand ecology in surface sea water. A heat-labile,filterable, bactericidal factor of sea water forsewage organisms was observed by ZoBell (1936).Studies on the viability and dispersal of coliformbacteria in the sea (Ketcham et al., 1949; Vaccaroet al., 1950) also indicated that a bactericidalfactor was present in sea water which was heatlabile, increased during the aging of sea waterand underwent a seasonal variation. Harvey(1955) states "there is some natural brake uponthe growth of planktonic bacteria in the opensea" and "it is an outstanding question why

528 [voi,. 77

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GASTROINTESTINAI, MICROFLORA OF ANTARCTIC BIRDS

offshore sea water, which contains sufficientnutrient for the production of several millionbacteria per cubic centimetre . . . normally sup-ports a population of no more than 10-200 bac-teria per cm.3 as free living planktonic cells."The fresh water algae chlorella was found byPratt et al. (1944) to liberate an antibacterialsubstance chlorellin in poor yield which was foundby Spoehr et at. (1949) not to be in the freshcells but to result from the photooxidation ofunsaturated fatty acids. Steemann Nielsen (1955)reported one trial in which a dilute chlorellaculture with its accompanying bacterial popula-tion consumed less oxygen in light than in darkbottles. On the basis of data collected by anotherinvestigator which indicated a variable rate ofoxygen consumption in- dark bottles taken in theSargasso Sea, Steemann Nielsen postulated theexistence of antibiotic production by planktonicalgae.

During the assay of sea water and phytoplank-ton samples in the Weddell Sea, water samplescontaining algae remained clear when left exposedto the air, while aliquots of the same samplewhich had the algae removed by Milliporemembrane filtration became very turbid andformed a large sediment of bacterial cells uponstanding. The ability of fresh algae to inhibit bothgram-positive and gram-negative organisms invery small quantities (five parts per million dryweight) indicates that a strong antibiotic-likesubstance which is water soluble can be liberatedby certain species of algae. As in the chlorellaexperiments (Pratt, 1948; Steemann Nielsen,1955) a higher antibacterial activity was observedin lower than in higher concentrations of algae.This substance was heat labile and filterable andmay be the bactericidal substance observed byZoBell (1936), Ketcham et al. (1949), and Va-carro et al. (1950). The marked activity of freshpreparations indicated that this material wasdifferent from chlorellin. Since all antarctic birds,animals, and fishes live either directly or in-directly on phytoplankton, this material mayplay a large role in the bacterial ecology of antarc-tic fauna. This phenomenon may not be restrictedto the dense phytoplankton areas of the antarc-tic.The predominant gastrointestinal microflora

of domesticated birds consists of lactobacilli,coliforms, and the enteric group of streptococci,

in that order, (Sieburth et al., 1951, 1952, 1954)and appears to reflect the microflora of the feed.Lactobacilli and the enteric group of streptococciwere not detected in the antarctic birds studied.Slow lactose fermenting strains of E. coti, otherslow or lactose nonfermenting enterobacteria,streptococci, other than the enteric group, anddiphtheroid-like organisms were the predominantbacterial types found in the birds with an aerobicflora. The fastidious diphtheroid-like and gram-negative rods and the anaerobic organisms diedout on subculture and were not studied. Proteusmirabilis, P. rettgeri, P. vulgaris, and Alcaligenesspecies were found in low numbers in some birds.Of the five cultures from antarctic birds broughtback by Dr. Charcot, Mlle. Tsiklinsky (1908)identified Bacillus pyocyaneus obtained from apenguin and Bacillus megaterium which wasfound in both a sea gull and a cormorant. Bunt(1955) observed E. coli, Bacillus species, non-sporing rods, and micrococcus. Since most de-scriptions of bacteria from antarctic birds aremorphological and incomplete it is almost im-possible to compare bacterial types observed bythe various workers.

ACKNOWLEDGMENTS

The author would like to sincerely thank themany individuals that helped make this studypossible. Among these were Drs. Hiden T. Cox,Sidney Galler, W. B. Bell, and Adm. L. 0. Col-bert who made the necessary last minute ar-rangements for support. Particular acknowledg-ment is due: Capitan de Fragata Luis R. A.Capurro, the captain of the General San Martinwho took interest in this project and made everyfacility available; personnel in the HydrographicService and Naval Antarctic Group of the Ar-gentine Navy who provided the Laboratorymaterials and logistic support essential for thisstudy; and Dr. W. H. Ewing, and Mrs. NancyP. Oakey who worked on the cultures broughtback for further study.

SUMMARY

During the 1957-1958 Argentine AntarcticExpedition through the South American quadrantof Antarctica, microbiological studies wereconducted on antarctic birds in an improvisedbacteriological laboratory aboard the icebreakerARA General San Martin.

193591 529

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SIEBURTH

Twenty eight birds were studied, sixteen indetail. "Bacteriologically sterile" birds were notobserved; in the four birds in which an aerobicflora was not detected, obligate anaerobes werefound. In the remaining twelve birds which hadan aerobic flora, slow or negative lactose fer-menting gram-negative rods were the predomi-nant bacterial types.

Lactobacilli and the enteric group of strepto-cocci were not detected but other species ofstreptococci and diphtheroid-like organisms werepresent in good numbers in some birds. Typicalrapid-lactose fermenting strains of Escherichiacoli were not observed in the penguins (Pygoscelisspecies) feeding on lobster krill and delayedlactose fermenting strains were more prevalentin the flying birds.The strong antibacterial property of blood

serum in some of the antarctic birds appeared tobe the factor responsible for modifying the gas-trointestinal flora of a predatory giant fulmar.Algae in sea water phytoplankton had a marked

differential antibacterial activity. This observa-tion not only explains the growth inhibitingfactor in the gastrointestinal material of penguinsfeeding on the algae laden lobster krill but mayalso explain in part the low planktonic bacterialpopulation in sea water and the predominance ofgram-negative organisms.

REFERENCES

BREED, R. S., MURRAY, E. G. D., AND SMITH, N. R.1957 Bergey's manual of determinative bac-teriology, 7th ed. The Williams & WilkinsCo., Baltimore.

BUNT, J. S. 1955 A note on the faecal flora ofsome Antarctic birds and mammals at Mac-quarie Island. Proc. Linnean Soc. N. S.Wales, 80, 44-46.

EKELOF, E. 1908 Bakteriologische Studienwahrend der Schwedischen Suid-polar Expe-dition. Wiss. Ergeb. Schwed. Sudpol-Ex-ped., 1901-1903, Band IV, Lief 7, Stockholm.

GAZERT, H. 1912 Untersuchungen iber Meeres-bakterien und ihren Einfluss auf den Stoff-wechsel im Meere. Deutche Sudpolar Expe-dition, 1901-1903, Band VII, Heft III, pp.268-296. Georg Reimer, Berlin.

HARVEY, H. W. 1955 The chemistry and fertilityof sea waters. Cambridge University Press,Cambridge, England.

HARVEY PIRIE, J. E. 1912 Notes on Antarctic

bacteriology. Scottish National AntarcticExpedition. Report of the Scientific resultsof the S. Y. Scotia, Vol. 3 Botany, No. 10, pp.137-148. Edinburgh.

KETCHAM, B. H., CAREY, C. L., AND BRIGGS, M.1949 Preliminary studies on the viabilityand dispersal of coliform bacteria in the sea.Limnological aspects of water supply andwaste disposal, published by Am. Assoc.Advance. Sci., pp. 64-73.

LEVIN, M. 1899 Les microbes dans les regionsArctique. Anrn. inst. Pasteur, 13, 558-567.

McLEAN, A. L. 1919 Bacteriological and otherresearches. Australasian Antarctic Expedi-tion, 1911-1914. Scientific Reports, SeriesC, Vol. VII, Part 4, pp. 15-44, Sydney.

MURPHY, R. C. 1936 Oceanic birds of SouthAmerica. Two Vols. Macmillan Co. and Amer.Museum of Natural History, New York.

PARKER, C. A. 1955 Anaerobiosis with ironwool. Australian J. Exptl. Biol., 33, 33-38.

PELCZAR, M. J. AND CONN, H. J. 1957. Manualof microbiological methods. Society of Ameri-can Bacteriologists. McGraw-Hill Book Co.,New York.

PRATT, R. 1948 Studies on Chlorella vulgariis.XI. Relation between surface tension andaccumulation of chlorellin. Am. J. Botany,35, 634-637.

PRATT, R., DANIELS, T. C., EILER, J. J., GUNNI-SON, J. B., KUMLER, W. D., ONETO, J. F.,STRAIT, L. A., SPOEHR, H. A., HARDIN, G. J.,MILNER, H. W., SMITH, J. H. C., AND STRAIN,H. H. 1944 Chlorellin, an antibacterialsubstance from chlorella. Science, 99, 351-352.

SIEBURTH, J. McN. 1958 Antarctic microbi-ology. Am. Inst. Biol. Sci. Bull., 8(3), 10-12.

SIEBURTH, J. McN., GUTIERREZ, J., MCGINNIS,J., STERN, J., AND SCHNEIDER, B. H. 1951Effect of antibiotics on intestinal microfloraand on growth of turkeys and pigs. Proc.Soc. Exptl. Biol. Med., 76, 15-18.

SIEBURTH, J. MCN., MCGINNIS, J., AND SKINNER,C. E. 1952 The effect of terramycin on theantagonism of certain bacteria against speciesof Proteus. J. Bacteriol., 64, 163-169.

SIEBURTH, J. MCN., JEZESKI, J. J., HILL, E. G.,AND CARPENTER, L. E. 1954 Some micro-biological observations on the antibiotic-fedchick. Poultry Sci., 33, 753-762.

SKARNES, R. C. AND WATSON, D. W. 1957 Anti-microbial factors of normal tissue and fluids.Bacteriol. Revs., 21, 273-294.

SPOEHR, H. A., SMIITH, J. H. C., STRAIN, H. H.,MILNER, H. W., ANI) HARDIN, G. J. 1949

530 [VOL. 77

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

GASTROINTESTINAI, MICROFLORA OF ANTARCTIC BIRDS

Fatty acid antibacterials from plants. Car-negie Inst. Wash. Publ. No. 586, Washington,D. C.

STEEMANN NIELSEN, E. 1955 The productionof antibiotics by plankton algae and its effectsupon bacterial activities in the sea. Pap.Marine Biol. Oceanog. Deep Sea Research,3(Suppl.), 281-286.

STURKIE, P. D. 1954 Avian physiology. Com-stock Publishing Assoc., Ithaca, New York.

TSIKLINSKY, MLLE. 1908 Flore microbienne.Expedition Antarctique Francaise 1903-1905.Masson et Cie., Paris.

VACCATRO, R. F., BRIGGS, M. P., CAREY, C. L.,AND KETCHAM, B. H. 1950 Viability ofEscherichia coli in sea water. Am. J. PublicHealth, 40, 1257-1266.

ZoBELL, C. E. 1936 Bactericidal action of sea

water. Proc. Soc. Exptl. Biol. Med., 34, 113-116.

1959] 531

on April 23, 2020 by guest

http://jb.asm.org/

Dow

nloaded from