Seasonal nitrogen fixation in the sediment of an Amazonian lake impacted by

4 Pages·1996·3.4 MB·

by  Enrich-Prast, A

AMAZONIANA XIY (ll2): 157 - 163 Kiel. Dezcmber 1996
Seasonal nitrogen fixation in the sediment of an Amazonian lake
impacted by bauxite tailings (Batata Lake-Pará)
by
A. Enrich-Prast & F.A. Esteves
M.Sc. A. Enrich-Prast & Prof. Dr. F.A. Esteves, Lab. de Limnologia, Dep. de Ecologia,
Universidade Federal do Rio de Janeiro, Cidade Universitária, Ilha do Fundào, Rio de
Janeiro, RJ, Caixa postal 68020, 21941-540, Brazil.
(Accepted fbr publication: September, I 996).
Abstract
Batata Lake is an Amazonian clear water lake that undergoes large seasonal fluctuatìons in water level.
For a period of l0 years (1979-1989), the northerrr end oftlie lake received a total of50,000 nlrdrof
bauxite tailings. As a consequence spelling aproximately 30 % olits sedinrerìts are covered by tailings. The
plincipal goal ofthis lesearch was to estimate rates ofnitrogen lìxation in tlre inrpacted and non-inrpacted
sedimerrt in the different hydroperìods that occur in this ecosystenr (drawdown, drying, fìlling and
flooding). Nitrogen fìxation was estimated using the acetylene reduction nrethod. The highest rates of
nitrogen fixation were observed to occur during the drying period and appear to be dìrectl¡' relaled to au
increase in prin.rary production by phytoplankton. Decreased rates ofnitrogen tìxatìon occulred duling the
hydroperiods of filling, flooding and drawdown with the greatest reductions occuring in the inrpacted area
ofthe lake. In the impacted area ofthe lake, bauxite tailings have reduced primary production in [l]e water
column, decreased labile authoctonous carbon availability to heterotrophic bacteria in the sedirnents, and
decreased nitrogen fixing activity olorganisrns preserlt the sedinrents.
Keywords: Bauxite tailings, Amazonian lake, nitrogen cycle, nitrogen fixation, sediment.
Resumo
O lago Batata é ur.n lago de águas claras, submetido a flutuaçào sazonal do nível d'água. Dulante | 0
anos (1979-1989), a regiào norte do lago recebeu unì total de 50.000 nri d' de rejeito de bauxita. Conro
consequência, cerca de 30%odaârea do lago fbi impactado por lejeito de bauxita. O principal objetivo
desta pesquisa foi estimar as taxas de fixaçào biológica de nitrogênio no sedinrento da área iurpactada e
natural durante vários períodos do ciclo hidrológico (vazante, seca, enchente e chcia). A fixaçào biológica
de nitrogênio foi estimada usando o método da redução de acetino. As rnaiores taxas de fixaçào de
nitrogenio foram observadas durante o período de seca e parecem estar diretanrente Ielacionadas ao
aurnento da produção prirnária fitoplanctônica. Redução das taxas de fixaçào biológicas ocorrerarn durante
o período de enchente, cheia e vazante, com grande redução sendo obselvada na área impactada do lago.
ISSN 0065-6755/19961157/ O MPI für Limnologie, AG Tropenökologie, Plön; INPA, Manaus
ls6 t57
Na área inrpactada, o rejeito de bauxita reduz a produçào primária fìtoplancttìnica rra coluna d'água, tl.ìoasure thc diazotrophic activily ol hetcrotrophic organisrls in the scdimenls. Alier 2.0 h ol'incLrbation.
dintintrindo a disponibilidade de carbono lábil autóctone para bactérias heterotrofàs do sedimento, a qual 3.5 r¡l ol'gâs wele withdrawn f'on'n the headspace ol'each bottle wìth a gas-trght syr-ingc, translercd to 7
prol.ìlove a reduçào da atividade dos organisnros tìxadores de nittogênìo presentes no sedirnento. nrl Vacuntainels, and analysed within 2 weeks afier santpling.
Iìthylene concentrations were detcrntincd in a Varian 3400 gas liquid chromatoglaplr using a llanrc
ionization detector (200"C), a Porapak N column (60 C¡, and an injeclor ternperature ol'l0() 'C'. TIrc
standat'd ratio ol 3:l lol ethylene reduced to nitrogen reduced was used to calculete absolr¡tc ¡ates ol'
Introduction nitrogen lìxed by the sedinrents (HARDY op. cir.¡.
Tenrperature and oxygen concentrations olthc water-sedir¡ent interfäce werc lneasulecl in thc I'ieìcl
The aquatic and terrestrial ecosystems located in the central pan of the Amazon using a Model DO-l lP TAO Oxir¡eter. Total organic carbon (by the dichlomate oxiclation rrcthod)
concentratiorls of'the sedir¡ent wcre perlbrmed accolding to CìOLTERMAN et al. (197u).
region are avowedly poor in nutrients (SIOLI 1984). Yet, despite the scarcity of nu-
trients. Amazonian aquatic ecosystems are very productive due to high mltrient recy-
cling rates (JLTNK et al. 1989).
Results and discussion
Nitrogen is a critical nutrient in aquatic ecosystems because it often limits primary
productivity (HUTCHINSON 1957; WETZEL 1972; HORNE & GOLDMANN 1994). The large fluctuation in water level that occur in Amazonian ecosystelns is a basic
This is certainly alsci true of Amazonian ecosystems. MELACK & FISCHER (1988)
detenninant of ecological structure and function in the Amazon basin. This phenomenon
showed that lack of nitrogen was severely limiting to algal growth in Amazonian
is discussed by JTINK et al. (1989) and is commonly reffered to as flood pulse. In this
floodplain, directly decreasing primary productivity of the system they studied. In this
paper we show that variations in water level of 1.5 to 9.5 m over our collection stations
paper, rates of nitrogen fixation are estimated for sediments covered by bauxite tailings
directly influence nitrogen f,rxation in the sediments of the non-impacted area of Batata
and narurally sedimented areas of Batata Lake as large seasonal fluctuations occur in
Lake. An ANOVA analyses showed that the nitrogen fixation rates during drying period
water level.
was significantly greater (cr = 0.05) than other periods of the hidrocycle (Fig. 2). During
drying periods, observed nitrogen fixation rates are approximately 7 tirnes greater for
ì ì)
non-impacted sediments (0.40 nrnol g D.W.h compared to impacred sedirnents (0.06 I
Study area nmol g-r D.W.h-r). The non-impacted area present significantly higher values (ANOVA.
o = 0.05) than the impacted area, during the flooding, drawdown and drying. Nitrogen
Batata I-ake (Pará-Brazil), is a clear water lake on the rvest bank of the Trombetas ftxation rates in impacted sediments does not vary significantly (ANOVA. cr = 0.0-5)
River (1'30'S e 56"20'W; Fig. l). For ten years this lake received,50,000 m3d-r of
during flooding, drawdown or drying periods, varying only from 0.034 to 0.05 nmol g'r
bauxite tailings containing ore residues and 7 to 9 Yo clays particles ofless than 50 pm D.W.h I during these periods (Fig. 2). During filling, nitrogen fixation is ahnost not
in size. This mining effluent covered aproximately 30 % of the total lake sediments at
detectable in the sediments of both impacted and non-impacted areas of the lake.
the northem end of the lake. Batata Lake is also subject to large fluctuations in hydro- HUZSAR (1996), who studied the phytoplanktonic populations of Batata Lake.
period. During the course of this research, the lake fluctuated approximately 8.5 meters
reported that phytoplancton communities increase during the drying period. It is during
in dept exhibiting periods of filling (rising water level); flooding (highest water level);
this porlion of the hydroperiod that Batata Lake is not connected with the igapo f-orest
drawsown (water level decreasing), and drying (lowest water level). (flooded forest). As a consequence, sedimentation of the organic carbon from the
phytoplancton communities in the water column is most likely promoted as lakc water
levels and currents decrease during the drying period (FERRAO-FILHO & ESTEVES
Material and methods
1994). However, this organic carbon, whose source is the phyoplancton. is mainly
labile caibon, i.e. carbon that can be readily assimilated by heterotrophic bactcria
Sedir¡ent sanrples were collected at two stations in Lake Batata: a) in the northern area impacted by
including nitrogen frxers (WETZEL et al. 1972; KNOWLES 1982). During the drying
bauxite tailings, and b) in the southern non-impacted or r'ìatural area olthe lake. Samples were collected
during fìlling (7.5 meters depth - March), llooding (9.5 rneters depth - June), drawdown (6.5 meters - hydroperiod, we observed an increase in phytoplankton occuring in the non-inpactcd
September) and drying (1.0 nreter depth - December) in 1994 wìth a hand held coring device. At eaclr sediments, a decrease in organic carbon to these same sediment (6.0 o/o), and a peak ill
station, l0 corcs wcre collected with a corcr sarnplcr as proposed by AMBÜHL & BÜHRER (1975). TIre nitrogen fìxation suggesting that heterotrophs are rapidly utilizing this carbon as it
top three cnt ofeach core were calefully removed, pooled, throughly homogenized, and placed in 100 ml becomes available to them (Table I and Fig. 2). In contrast, we observed that organic
glass bottles capped with serum stoppers. Bottles were kept cool and transported to the laboratory where carbon concentrations (1.9 %) were greatest in the sediments during the same pcriod
they were evacuated at 75 ml nrin r pure nitrogen for 4 min to maintaining anaerobic conditions. suggesting that the altered sediments do not support a high population of organic carbon
Rates of nitrogen fixation in the sediment samples were estimulated using the acetylene reduction utilizing heterotrophs (Table l).
nrethod (HARDY et al. 1968). Each pooled sediment sample was analysed in triplicate, accompanied by Deposition of greater amounts of carbon to the non-impacted sediments as well as
a control sample. Acethylerle was added to the headspace to an initial concentration of l5 Vo (v:v) except decreased rates of nitrogen fixation are found during hlling, fìooding and drawdown
lor controls that did not have acetylene added to them. Samples were incubated in the dark at 26 "C to periods suggesting low utilization of iabile carbon by microorganisms during thcse
158 159
portions of the hydroperiod. Extensive areas of forest adjacent and upstreams of Batata HUZSAR, V.L.M. (1996): Phytoplankron of'an Anrazonian Iake inrpactcd by bauxite tailing (Lago lìatata,
Lake arca ir.rundated during periods of fìlling and flood. During thcse periods considera- Pará, Brazil). - Proceedings ol XXVI SIL Congress: in press.
ble allochthonous carbon erìters Lake Batata as refractory carbon, i.e. carbon in the form IIUTCIIINSON, L.G. (1957): A Treatise on Linrnology. - Vol. I, Part l, John Wilcy & Sons: 540 p¡r.
ol cellulose. lignin and pectin. that is not available for rapid utilization by bacterial .lUNK, W.J., IIAILEY, P.U. & R.Li. SPARKS (1989): Thc fìood pulse concept in liver-lìrxrtl¡rlain sy-
hetelotlophs. This carbon is observed to increase in non-impacted sediments at high stems. - In: D.P. DODCìtrì (ed.). Proc. Int. Lalge River Symp. Can. Publ. I'ish. Âquat. Sci: ll0-127.
KNOWLUS, tl. (1982): l)enitrifìcation. - Mrclobiological Iìeviews, (March): 43-70.
lates during frlling (6.7 %), flooding (9.4 %). and drawdown (7.2 %).
MELACK,.l.M. & T.R. FIStIUR (1988): Denitrification and nitrogcn fìxatìon in arr Arlazon lìooclpÌain
The tlood pulse affecting Batata Lake certainly changes the niirogen fixing activity lake. - Verh. Internat. Verein. Lirnnol. 23:2232-2236.
of the sediments. The low values for nitrogen fixation observed during filling, flooding ROLAND, F-., tiSTBVtsS, F.A. & F'.A.R. UARUOSA (1996): The inf-luence ol'the bauxite tailing on the
and drau'dowu for both nor.r-impacted and impacted sediments are probably related to light regirne and its consequence olr the phytoplankton pr'ìnrary production in an Anrazonian llood-
rapid chauges in water volume during these periods which effect changes in the phy- plain lake. - Proceedings ol XXVI SIL Congress: in press.
toplancton populations (FERRAO-FILHO & ESTEVES 1994). This decrease in rate of SlOLl, H. (1984): Introduction history olthe discovery of'the Ar.lrazon and of'r'esealch of'Anlazonia watcrs
nitrogen fixatior.r is rnost notable for the impacted areas of the lake. Bauxite tailings in and landscapes. - In: H. SIOLI (ed.) The Anrazon. - Linrnology and landscape ecolog¡, ol'a n.righty
suspension in the water coiuurn have at least two deleterious effects on this ecosystem. tropical river and ìts basin, The llague. Junk Publ.: l-13.
Tailings cloud in the water column directly reducing primary phytoplanktonic produc- V,/ETZÊ.L, R.G. (1972): Limnology. - W.[J. Saunders Conrpany: 743 pp.
tion (ROLAND et al. 1996) by decreasing photosynthetically active radiation available WETZF.L, R.G., RICH, P.H., MILLER, M.C. & H.L. ALLEN (i972): Mctabolisnr of dissolved and
to photosynthetic organisms and thus the availability of labile carbon for heterotrophs particulate detrital carbon in a temperate hard-water lake. - ln: Detr'ìtus and its role in aquatic ecosy-
sterns. Proceedings ol an IBF-UNESCO Symposiurn: 185-243.
such as nitrogen f,rxers. Secondly. the bauxite taiiings covering the sediments modify
abiotic conditions in the sedirnents by limiting nitrogen transfer to the sediments. Total
organic carbon in the sediments covered with tailings ranges fron about l7 to 32o/o of
the organic carbon present in natural sedinents in the lake (Table l). Oxygen concen-
trations at tl.re sediment-water interface were always less for the non-impacted sedinents
tiran for the impacted sediments suggesting tliat the natural sediments could support Table l: Organic calbon concentrations of tlie sediment and teÍìrperature and oxygen concentration ot'thc
larger populations of oxygen utilizing organisms than in the impacted areas. sedirrent-water interlàce fì'onl the natural arrd irnpacted areas of Batata Lake duling the yc'at'ol'
1994.
Acknowledgments
Carbon T'C Ox¡'gen
Hydroperiods % D.W mg/l o/" sat.
We are very grateful to'Dr. Fábio Roland and Dr. Reinaldo Luiz Bozelli for their
kind assistance during our held activities, to our colleagues of the Laboratory of
Limnology of the Ecology Department of the Universidade Federal do Rio de Janeiro
Filling (March)
for their criticisms of the manuscript, and Dr. David Biesboer of the University of
lmpacted area 1.8 29.4 3.5 46.0
Minnesota, U.S.A., for the criticisms of the manuscript and English support. We also Natural area 6.7 29.7 0.8s l 1.0
thank Mineraçao Rio do Norte Company for the given frnancial support. Flooding (July)
hnpacted area 1.6 26.8 4.0 49
Natural area 9.4 26.4 3.0 ¿+.5
References Drawdown (September)
Impacted area t.3 30.9 t.26 t7
AMBUIIL, H. & II. BUIIRER (1975): Zur Technik der Entnahnre ungestörter Grossproben von See- Natural area '1 .l 29.1 0.26 J
sedinren(en: ein verbessertes llohrlot. - Schweiz. Z. Hy<lrol.37: 175-186. Drying (DecembeL)
FERRÀo-FILIlo, A.S. & F-.4. ESTEVES (1994): Nutritive value ancl sedimenrarion rares of particulate Impacted area 1.9 28.0 8.l5 105
n.ìatter in the course oltwo flood pulses in an Ar¡azonian Várzea lake. - Arch. Hydlobiol. 130: 325- Natural area 6.0 2 8.8 7.68 102
337.
GOLTERMAN, H.L., CLYMO, R.S. & M.A.M. OHNSTAD (1978): Methods tbrphysical and chemical
analysis of lreshwater. - Blackwell Scientilic Publ: 214 pp.
I-lARDY, R.W.F., HOLSTEN, R.D. & E.K. JACKSON (1968): The acetylene-ethylene assay for nitrogen
fixation: Laboratory and fìeld evaluation. - Plant Physiol.43: I 185-1207.
ilORNE, A.J. & C.R. GOLDMAN (1994): Linrnology. - McGraw-Hill International Editions, Second
Edition: 576 pp.
r60 161
l
0.s5
0.5
10
0.45
ElNatural Area
0' 0.4
8
l-llmpacted
Area
0.35
*
0.3 6 Øk \ùy'ater level
IJ ()
8" Þo 0.25 () -
o.2 4
70' 60' 50'
0.15
0.1 2
0.05
o 3rt 0 0
\ MLauksesurá filling flooding drawdown drying
lmpacted Region
Natural Region Fig.2:
Values lorm the nitrogen fixation rates in the sediment fi'om the natural and impacted regions of Batata
Trombetas River
Batata Lake in 1994.
Lake
G
Fig. l:
Location of the study area and the sampling stations.
162
163