—5 0510 15 20U

10

20

30

40

10020030000

10

20

30

40

TEMPERATURE; CONDUCTIVITY

AMMONIUM; NITRATE

02040600204060OXYGEN

NITRITE; NITROUS OXIDE

Figure 1. A. Temperature (degrees Celsius), conductivity (siemens per meter), dissolved oxygen (milligramsper liter) in the east lobe of Lake Bonney. B. Ammonium (micromolar), nitrate (micromolar), nitrite (micro-molar) and nitrous oxide (micromolar-nitrogen) in the east lobe of Lake Bonney. Measurements made inthe austral summers between 1989 and 1991. The permanent icecap was approximately 4 m thick whenmeasurements were made. All depths are from the surface of the water within the sampling hole (that is,piezometric level).

Terrestrial biology

Water column transformations of nitrogen in Lake Bonney, aperennially ice-covered antarctic lake

JOHN C. PRISCU, Department of Biological Sciences, Montana State University, Bozeman, Montana 59717BESS B. WARD, Marine Sciences, University of California, Santa Cruz, California 95064

MALCOLM T. DOWNES, National Institute of Water and Atmospheric Research Ltd., Christchurch, New Zealand

permanently ice-covered Lake Bonney presents a specialsituation where turbulence and upper trophic levels are

virtually nonexistent. During our past studies on the photobi-ology of both the east and west lobes of this lake (Priscu et al.1990), we made preliminary vertical profiles of nutrients,nitrous oxide, and bacterial number and activity. These datareveal several unique features. First, greater than 250 micro-molar nitrate and 30 micromolar nitrite exist in a region of theeast lobe which is apparently devoid of oxygen (figure 1A, B).This is the first case, to our knowledge, of the existence of alarge nitrate pool in apparently anoxic water. Second, nitrous

oxide levels in the east lobe just below the chemocline reachalmost 40 micromolar nitrogen (figure 1B). This value exceedsair saturation by over 500,000 percent and is the highestrecorded in a natural aquatic system. Lack of oxidized nitro-gen (for example, nitrate, nitrite, nitrous oxide) in the anoxiczone below the chemocline of the west lobe (figure 2A, B)implies intense denitrification, which conforms to conven-tional biogeochemical concepts for nitrogen within this lobe.

A tentative calculation of nitrous oxide diffusion acrossthe 20.5-meter (m) plane of the east lobe (using a moleculardiffusion coefficient of 10-6 square centimeter per second)

yields an upward flux of 1.2micrograms nitrous oxide-nitrogen per square meter perday which, given the surfacearea of the lake and assuming aconstant annual nitrous oxidegradient, equates to about 1.3kilograms (kg) nitrous oxide-nitrogen per year. Our prelimi-nary measurements of nitrousoxide within the air above thelake ice on the east lobe re-vealed that they were about 40percent greater than the globalaverage for nitrous oxide orthat measured over nearbyMcMurdo Sound (Rasmussenand Khalil 1986; Priscu et al.1990) indicating that nitrousoxide escapes from the lakedespite a permanent ice cap.Although this is a small out-ward flux relative to the globalannual production of nitrousoxide (about 2x10 7 kg nitrousoxide-nitrogen per year;Thiemens and Trogler 1991), itmay have regional importance.

These unusual results(from the east lobe), in concert

ANTARCTIC JOURNAL - REVIEW 1993

237

TEMPERATURE; CONDUCTIVITY

—5 0510 15 20:ii

00 TEMPI COND U OXY

0

10

20

20

30

30

A1

40 40

with previous reports showing that nitrogen can limit phyto-plankton growth during certain periods (Sharp and Priscu1990), have led us to propose the following hypotheses:• Regeneration of nitrate and nitrite by nitrifying bacteria

and ammonium by heterotrophs supports primary pro-duction in the upper trophogenic zone whereas ammoni-um diffusing upward across the chemocline supports themajority of production in the lower trophogenic zone, par-ticularly in the west lobe.

• Nitrous oxide is produced in the chemocline of both lobesby nitrification and nitrifier denitrification (of nitrite). Thisprocess is most pronounced in the east lobe.

• The nitrite maxima are the result of nitrification in excessof denitrification.

• Nitrous oxide and nitrite levels in the chemocline of thewest lobe are considerably lower than in the east lobebecause they act as terminal electron acceptors for denitri-flers in the former, that is, denitrification is not a sink fornitrous oxide or nitrite in the east lobe.

• The production of nitrous oxide in the east lobe signifi-cantly lowers upward diffusion of available "new" nitrogenfor phytoplankton primary production.

• Heterotrophic denitrification does not occur in the eastlobe whereas high rates of denitrification exist in the deepwaters (below 25 m) of the west lobe. The latter acts as amajor loss of nitrogen from the west lobe of the lake.

• Lake Bonney is an atmospheric source of nitrous oxide.These hypotheses have formed the basis of our current 3-

year study on Lake Bonney. The available background data forthe two lobes of Lake Bonney, each of which has distinctivecharacteristics with respect to nitrogen dynamics, indicate

that Lake Bonney provides anideal situation to test thesehypotheses. The purpose ofthis report is to present someof our preliminary results(notably those for nitrousoxide) and discuss them brieflywithin the context of hypothe-ses derived from them. Thisarticle also serves as an intro-duction to the reports of Ward,Cockcroft, and Priscu (Antarc-tic Journal, in this issue), Wool-ston and Priscu (AntarcticJournal, in this issue), andSmith and Priscu (AntarcticJournal, in this issue) whichaddress various aspects of theproposed hypotheses.

BBartlett et al. (Antarctic

Journal, in this issue) describea number of techniques that

1 had to be developed (to over-TE; NITROUS OXIDEcome the high salt content

common to antarctic lakes) touse nitrogen- 15 successfully as

an isotope for the measurement of nitrogen transformationswithin Lake Bonney. Finally, Wing and Priscu (Antarctic Jour-nal, in this issue) present results from a microbial communitywe discovered within the permanent icecap of Lake Bonney.We plan to propose further research to determine the biogeo-chemical importance of this ice community to the lakeecosystem.

Our study incorporates direct-rate measurements of allmajor nitrogen transformations together with immunochemi-cal and molecular probes for microorganisms responsible forspecific transformations. These techniques, together withmeasurements of various nitrogen species within the lake overtime, will provide a better understanding of biological nitrogentransformations in high-latitude aquatic systems and willcomplement recent studies on nitrogen transformations inarctic systems (Alexander, Whalen, and Klingensmith 1989).

This work was supported in part by National ScienceFoundation grant OPP 91-17907 to John C. Priscu.

References

Alexander, V., S.C. Whalen, and K.M. Klingensmith. 1989. Nitrogencycling in arctic lakes and ponds. In W.F. Vincent and J.C. Ellis-Evans (Eds.), High latitude limnology. Dondrecht, The Nether-lands: Kluwer Academic Publishers.

Bartlett, R., J.C. Priscu, and C. Woolston. 1993. Influence of high salin-ity levels on 15-nitrogen extraction efficiency in Lake Bonney,Antarctica. Antarctic Journal of the U.S., 28(5).

Priscu, J.C., M.T. Downes, L.R. Priscu, A.C. Palmisano, and C.W. Sulli-van. 1990. Dynamics of ammonium oxidizer activity and nitrousoxide (N20) within and beneath antarctic sea ice. Marine EcologyProgress Series, 62(1), 37-46.

0204050 0OXYGEN

NITRI

Figure 2. A and B. As for figure 1 except for the west lobe of Lake Bonney.

AMMONIUM; NITRATE

100200300

o AMMONIUM I NITRATEA NITROUS OXIDE A NITRITE

AA"I

ANTARCTIC JOURNAL - REVIEW 1993238

C,

• W. LOBEo E. LOBE

Priscu, J.C., T.R. Sharp, M.P. Lizotte, and P.J. Neale. 1990. Photoadap-tation by phytoplankton in permanently ice-covered antarcticlakes: Response to a non-turbulent environment. Antarctic Jour-nal of the U.S., 25(5), 221-222.

Rasmussen, R.A., and M.A.K. Khalil. 1986. Atmospheric trace gases:Trends and distributions over the last decade. Science, 232,1623-1624.

Sharp, T.R., and J.C. Priscu. 1990. Ambient nutrient levels and theeffects of nutrient enrichment on primary productivity in LakeBonney. Antarctic Journal of the U.S., 25(5), 226-227.

Smith, J.J., and J.C. Priscu. 1993. Microbial respiration potential inLake Bonney using a novel tetrazolium -reduction method. Antarc-tic Journal of the U.S., 28(5).

Thiemens, M.H., and W.C. Trogler. 1991. Nylon production: Anunknown source of atmospheric nitrous oxide. Science, 251,932-934.

Ward, B.B., Cockcroft, A.R., and J.C. Priscu. 1993. Nitrifying and deni-trifying bacteria in Lake Bonney. Antarctic Journal of the U.S.,28(5).

Wing, K.T., and J.C. Priscu. 1993. Microalgal communities in the per-manent icecap of Lake Bonney, Antarctica: Relationships betweenchlorophyll a, gravel, and nutrients. Antarctic Journal of the U.S.,28(5).

Woolston, C., and J.C. Priscu. 1993. Phytoplankton utilization ofammonium and nitrate in Lake Bonney: A preliminary assess-ment. Antarctic Journal of the U.S., 28(5).

Nitrifying and denitrifying bacteria in Lake BonneyBESS B. WARD and ALIx R. COCKCROFT, Marine Sciences Department, University of California, Santa Cruz,

California 95064JOHN C. PRISCU, Department of Biology, Montana State University, Bozeman, Montana 59717

The chemical profiles and initial microbiological data col-lected previously in Lake Bonney (Priscu, Ward, and

Downes, Antarctic Journal, in this issue) identified severalinteresting features in the nitrogen cycle of the lake, raisingunusual questions. First,nitrate (NO3) is present in high 5concentrations in the east lobeNUMBER (X1O

of the lake, indicating that048although the water column isU

apparently anoxic below thechemocline, denitrificationdoes not occur. By contrast,nitrate is depleted in the anox-ic bottom waters of the west1

lobe, indicating that denitrifi-cation does occur there. Sec-ond, nitrous oxide (N 2 0) ispresent at extremely high 1ev- H 20els just below the chemoclinein the east lobe yet much lowerconcentrations are present inthe west lobe. N20 accumula-tion at low (but nonzero) oxy-gen levels is an indication ofnitrification (Cohen and Gor-don 1978; Goreau et al. 1980).Finally, bacterial numbers inthe east lobe show a discrete40maximum near 25 meters (m),05below the oxic/anoxic inter-ACTIVITY (xloface, a feature not evident inthe west lobe of the lake (figure1A and B). The distribution ofbacterial activity in the east

lobe measured by uptake of tritiated-thymidine (a measure ofheterotrophic bacterial activity) shows a somewhat typicaldepth profile, in that it decreases with increasing depth and isalmost undetectable at the chemocline. Thus, the peak in bac-

NUMBER (XlOb)SPECIFIC ACTIVITY (X1010)

1204812 0.00.51.0

10 05103 -3

ACTIVITY (X10)

Figure 1. Depth distribution of bacterial abundance (cells per milliliter), bacterial activity (nanomoles ofthymidine per liter per day) and specific activity (nanomoles of thymidine per cell per day). Data were col-lected 24-26 November 1992. Depths are referenced from the piezometric water level in the samplinghole. A. East lobe. B. West lobe. C. East and west lobes.

ANTARCTIC JOURNAL - REVIEW 1993239