Lacustrine wetland in an agricultural catchment: nitrogen - Hal

13 Pages·2015·0.96 MB·English

Hydrol. EarthSyst. Sci.,12,539–550,2008
Hydrology and
www.hydrol-earth-syst-sci.net/12/539/2008/
Earth System
©Author(s)2008. Thisworkisdistributedunder
theCreativeCommonsAttribution3.0License. Sciences
Lacustrine wetland in an agricultural catchment: nitrogen removal
and related biogeochemical processes
R.Balestrini,C.Arese,andC.Delconte
WaterResearchInstitute(IRSA-CNR),viadellaMornera25,Brugherio20047,Milano,Italy
Received: 18September2007–PublishedinHydrol. EarthSyst. Sci.Discuss.: 28September2007
Revised: 18January2008–Accepted: 10February2008–Published: 10March2008
Abstract. The role of specific catchment areas, such as the the effect of human activities on the environment. Unlike
soil-riverorlakeinterfaces,inremovingorbufferingtheflux pointsourcepollution,diffusepollutioncannoteasilybecon-
ofNfromterrestrialtoaquaticecosystemsisgloballyrecog- trolled,areductioncanonlybeachievedbyappropriateland
nizedbuttheextremevariabilityofmicrobiologicalandhy- managementtechniques. Itiswellknownthatmodernagri-
drologicalprocessesmakeitdifficulttopredicttheextentto culturalpracticessignificantlycontributetocatchmentnitro-
whichdifferentwetlandsfunctionasbuffersystems. Inthis gen losses. In Italy for example, agriculture accounts for
paper we evaluate the degree to which biogeochemical pro- 68%ofnitrogen(N)inputtothePorivercatchmentandthis
cessesinalacustrinewetlandareresponsibleforthenitrate percentage increases to 83% if we consider the N input to
removalfromgroundwatersfeedingCandiaLake(Northern groundwater (Autorita` di Bacino del fiume Po, 2003). In
−
Italy). Atransectof18piezometerswasinstalledperpendic- the Po Valley, nitrate (NO ) contamination of ground wa-
3
ulartotheshoreline,inasub-unitformedby80mofpoplar ters considerably constrains the readily available amount of
plantation,closetoacropfieldand30mofreedswamp.The drinkingwater.
−
chemical analysis revealed a drastic NO -N ground water
3 Overthelastfewdecades,muchinteresthasbeenfocused
depletionfromthecropfieldtothelake,withconcentrations
on specific catchment areas such as riparian zones which
decreasing from 15–18mg N/l to the detection limit within
are able to reduce or buffer the flux of N from terrestrial to
the reeds. Patterns of Cl−, SO2−, O , NO−-N, HCO− and
4 2 2 3 aquatic ecosystems (Peterjohn and Correll, 1984; Haycock
DOCsuggestthatthemetabolicactivityofbacterialcommu-
et al., 1993; Sabater et al., 2003). The significant role of
nities,basedonthedifferentialuseofelectrondonorsandac-
wetland can be seen in the European Water Framework Di-
ceptorsinredoxreactionsisthekeyfunctionofthissystem.
rective’s (EU, 2000/60), recent integrated approach to river
−
The significant inverse relationship found between NO -N
3 basinmanagementinEurope. Inthisframework,riparianar-
−
and HCO is a valuable indicator of the denitrification ac-
3 easareconsideredasasignificantelementofthehydrologi-
tivity. Thepluviometricregime,thetemperature,theorganic
calnetworkinwhichprotectionandrestorationarerequired
carbon availability and the hydrogeomorphic properties are
toachieve“goodwaterstatus”forsurfaceandgroundwaters
the main environmental factors affecting the N transforma-
(WetlandsHorizontalGuidance,2003).
tionsinthestudiedlacustrineecosystem.
Riparian areas may attenuate nitrogen inputs to aquatic
ecosystems through plant uptake, microbial denitrification,
soil storage and dilution (Hill, 1990; O’Neill and Gordon,
1 Introduction 1994;HaycockandPinay,1993;Groffmanetal.,1996;De-
vito et al., 2000). More attention has focused on denitrifi-
Knowledge of the processes controlling the extent and dy- cation as this process can lead to a real removal of N by
namics of nutrient fluxes from diffuse sources in hetero- the transformation of dissolved N into gaseous forms, till
geneous catchments is crucial in predicting and controlling favourable conditions for denitrifying bacteria persist. The
relative importance of these processes seems to be strictly
linkedtothehydrologicalpropertiesofsoil-waterinterfaces
Correspondenceto: R.Balestrini (MitschandGosselink,2000;Pinayetal.,2000;Burtetal.,
([email protected]) 2002).
PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion.
540 R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment
Table1.MedianconcentrationofmajorphysicalandchemicalvariablesinsamplescollectedbetweenDecember2003andJuly2004.
T pH HCO−3 NH+4-N NO−3-N NO−2-N ON SO24− Cl− O2 DOC
◦C meql−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1
P1 11.7 6.17 0.60 0.018 14.76 0.69 16.9 12.6 2.35 2.04
P2 11.7 6.61 1.97 0.015 10.29 0.13 0.31 20.8 14.0 1.90 2.11
P3 10.5 7.00 3.20 0.012 13.06 0.13 0.81 36.2 16.5 1.68 2.93
P4 10.9 6.90 3.72 0.012 8.06 0.11 0.35 45.4 17.8 1.31 3.89
P5 11.3 7.02 5.03 0.012 4.85 0.15 0.51 56.7 17.5 0.85 4.35
P6 10.8 7.00 4.49 0.010 5.76 0.16 0.46 52.5 16.2 0.62 4.51
P7 11.7 7.15 5.04 0.027 3.69 0.10 0.35 51.5 15.2 1.39 3.82
P8 14.2 7.41 7.00 0.015 0.08 0.30 92.1 14.9 2.39 4.34
P9s 13.5 6.86 4.16 0.063 0.33 0.50 77.2 13.7 1.55 5.03
P9d 13.9 7.17 7.71 0.031 0.00 0.24 34.0 9.5 1.17 3.89
R1s 12.2 6.22 2.37 0.356 0.03 2.47 197.7 12.6 0.69 15.01
R1m 11.6 6.89 9.24 0.047 0.05 1.27 31.3 4.4 2.82 5.90
R1d 12.2 6.85 10.19 0.029 0.00 0.71 0.7 2.8 1.55 5.10
R2s 12.0 6.48 2.71 0.138 0.05 1.07 276.4 10.5 0.50 8.29
R2m 12.9 6.90 8.75 0.041 0.05 0.86 6.7 4.6 3.25 6.63
R2d 11.4 6.93 9.92 0.016 0.00 0.51 0.7 1.5 1.16 6.22
R3s 11.8 6.60 4.41 0.532 0.05 0.96 84.2 9.6 0.78 7.90
R3d 11.5 6.75 7.68 0.198 0.05 1.08 0.7 8.3 0.91 6.46
lake 17.0 7.95 1.08 0.088 0.18 0.64 6.2 4.8 10.60 7.04
Italy,withthescarcityofongoingstudiesonnaturalwetlands
(Balestrinietal.,2004;Borinetal.,2004),thereisaneedto
observetypologiescharacteristicofthenationalterritory.
The general aim of the present paper is to evaluate the
N buffering capacities of a riparian area adjoining a small
lake (Lake Candia) which receives a large nutrient input
from agricultural land. We examined spatial and temporal
variation in water chemistry and hydrology to identify the
mainmechanisms(biological,physical,chemical)leadingto
−
groundwaterNO depletionatsoil-lakeinterfaces. Patterns
3
of nitrate, dissolved oxygen, dissolved organic carbon, al-
kalinity and sulphate concentrations were used to delineate
1 km
redox conditions and define the occurrence and location of
denitrificationzones. Theeffectofsomeenvironmentalfac-
tors,suchasorganiccarbonavailability,temperatureandhy-
Fig.1.MaplocationofCandiaLakecatchment.
dromorphology,ontheNdynamicwasinvestigated.
Despitetheincreasingnumberofpapersoverrecentyears
2 Methods
on the role of riparian areas in N removal, the extreme
variability of microbiological and hydrological processes 2.1 Studyarea
makeitdifficulttopredicttheextenttowhichdifferentwet-
landsfunctionasbuffersystems. Insomeareaslessthan5m Lake Candia (45◦ 19’ 25” N, 7◦ 54’ 43” E) is a small
is enough to remove all the NO− entering the riparian ar- (1.49 km2), inter-morainic shallow lake, located in the re-
3
eas(Balestrinietal.,2007;Blicher-MathiesenandHoffman, gion of Piedmont in North-West Italy (Fig. 1). The shore-
1999), while in others, little or no N removal is observed lineisalmostentirelyoccupiedbyabeltofemergentmacro-
(Devitoetal.,1989). Little research has been conducted in phytesinwhichPhragmitesaustralisdominates. Theannual
largeuplandaquifersreceivinghighNinputsandlittleinfor- floating leaved water chestnut (Trapa natans) covers about
mationisavailableonthevariabilityofthedepthofdenitrify- 20ha of the littoral zone (Galanti and Romo, 1997). Lake
ingactivityinriparianzones(Hilletal.2000).Inaddition,in Candia does not have any inflows but is mainly fed from
Hydrol. EarthSyst. Sci.,12,539–550,2008 www.hydrol-earth-syst-sci.net/12/539/2008/
R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment 541
Table2.MeanconcentrationofmajorchemicalvariablesinsamplescollectedbetweenMarchandNovember2003.
pH HCO− NH+-N NO−-N ON SO2− Cl− DOC
3 4 3 4
meql−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1 mgl−1
P2 7.07 0.56 24.5 10.4
P3 7.32 4.86 0.029 2.67 33.3 13.8 2.21
P4 7.34 3.45 0.065 0.34 1.17 27.2 11.6 3.38
P5 7.42 4.70 0.201 0.04 0.77 26.8 13.0 6.39
P6 7.62 4.38 0.100 0.10 1.77 26.3 12.6 4.95
P8 7.52 4.79 0.145 0.06 1.34 32.3 11.3 6.10
P9s 7.38 5.26 0.224 0.06 2.97 5.7 9.7 7.65
R1s 6.77 3.74 1.212 0.02 3.56 174.2 11.2 30.9
lake 8.07 1.11 0.217 0.17 0.95 4.8 4.5 6.70
groundwater and rainfall. An unconfined aquifer intercepts
cropfield poplar grove
the lake (at 226m a.s.l.) from the higher elevation of the
230
moraine (330m a.s.l.) on the West towards the North East.
reed swamp
Acathlniedms1etac6hto-eeynaeodnafnrdtupheaeeelpriaaeorvrdeeaa(rq1ahug9aie8sf6etber–me2oe0pcn0ce2arda)setiwsuocrnareasibla1leny1dd.1faep◦serCedhcsauinpmtdhitiea9dt1-litao2eknmmed.pmuer,Triarnhetgee- Elevation m a.s.l. 222222468 P1glacial tilPlm2aximum watPer3 levelP4 P5 P6PP78 P9sd R1msd R2smdminimRu3dsmla wkeater level
spectively. About 51.1% of the Lake Candia basin is occu-
222
pied by agricultural areas, 45.5% by forests and grassland 0 20 40 60 80 100 120

and only 3.4% by urban areas (Ciampittiello et al., 2005).
Distance from the crop m
Itreceivesalargeinputofnutrientsanditsecologicalstatus
iscurrentlyclassifiedas“poor”(DECRETOLEGISLATIVO
Fig.2. Verticalcrosssectionoftheexperimentalfield. Themaxi-
N.152/1999). mumandminimumrecordedwaterlevelsareshown. Codesrepre-
Fieldactivitieswereconductedonthenorth-westernshore sentpiezometernestlocations.•=indicatepositionofthepiezome-
ofthelakeinasub-unitformedby80mofpoplar(Populus tersscreen. ●
nigraL.)plantation,intheupperpartclosetoacropfieldand
30mofreedswampmainlyformedbyPhragmitesaustralis.
removed. Lakesampleswerecollectedclosetotheshoreline
Farmersdeclaredthatthesyntheticfertilizersregularlyap-
at a depth of 20–30cm at the same time as the ground wa-
pliedtothecropfieldconsistedofNPKcompoundandam-
ter samples. Water temperature and dissolved oxygen (O2)
moniumnitrateforatotalamountof200kgNha−1.
concentration were measured in the field, during the sam-
ple collection, from May to July 2004. The scarce rainfall
2.2 Samplingandanalysis
duringspringandsummer2003notablyconstrainedthewa-
Atransectofpiezometerswasinstalledperpendiculartothe ter availability in the soil and it was not possible to collect
shoreline, extending from the crop field to the lake. A total waterinanumberofthepiezometers. Inaddition,thecom-
of 18 piezometers were positioned at a 5 to 15m distance plete installation of piezometers in the reed swamp ended
from one another and at a depth ranging from 1.5 to 3m. in November 2003. For these reasons the chemical data of
In the reed swamp and in the transition zone between the March–November 2003 (10 samplings) is shown separately
twosub-unitsthepiezometerspositionedatthesamedistance (Table 2). For a better analysis of the spatial variations we
weresitedingroupsoftwoorthree,atdifferentdepths: 54– comparedonlytheresultsavailableforallthepiezometersof
80cm,surface(s),143–150cm,middle(m),and196–219cm the studied transect corresponding to 11 sampling sessions
deep (d) (Fig. 2). The piezometers consisted of PE pipe fromDecember2003toJuly2004.
(4cmid)with10cmslottedendsinstalledinholesdrilledby Soil profiles were obtained from cores collected mainly
handauger. Thewatertableelevationwasmeasuredatleast during piezometer installations. Soil samples were an-
once a month, from March 2003 to July 2004 using a suit- alyzed for organic carbon content using the Walkley-
able sounding probe. Ground water samples were collected Black method (Gaudette et al., 1974). Hydraulic con-
with the same frequency of hydrological measurements, si- ductivity was measured in the poplar sub-unit by pump-
multaneouslyinthepoplargroveandinthereedswampaf- ing tests and calculated using the Bower and Rice equation
ter at least one volume of the piezometer content had been (BouwerandJackson,1976).
www.hydrol-earth-syst-sci.net/12/539/2008/ Hydrol. EarthSyst. Sci.,12,539–550,2008
542 R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment
ternal quality controls at different concentrations, were 3%
250 rainfall 228,0 forNH+-N,2%NO−-N,4%forTN,3%forSO2−, 6%for
P2 4 3 4
200 P6 227,5 Cl−. Detectionlimitswere5µgl−1forNH+4-N,0.02mgl−1
rainfall mm 110500 R1 222222667,,,050elevation m asl ffoorrTNNO.−3-N, and Cl−, 0.05mgl−1 for SO24− and 0.1mgl−1
50 225,5 3 Results
0 225,0
Jan-03Feb-03Mar-03Apr-03May-03Jun-03Jul-0A3ug-03Sep-03Oct-03Nov-03Dec-03Jan-0F4eb-04Mar-04Apr-04May-04Jun-04Jul-04Aug-04 3.1 Soilfeatures
Fig.3.Monthlyvariationofthewatertableinselectedpiezometers Inthesoilprofile,thesiltlayersoflacustrineoriginpredom-
andamountofmonthlyrainfall. inate, alternatedwithsandanddeepgravel-sandylayers. In
everyprofileweobservedredoximorphicfeaturesasgreyin
colour with red spots and deep grey layers. Organic matter

230 P1 P2 P3 P4 P5 P6 P7 P8 P9 R1 R2 R3 of a very dark colour was often detected within the hydric

layers. Intheupperpartofthetransect,thesoilswereclas-
229
sifiedasTypicUdorthens,whilethoseclosetothelakewere
m a.s.l. 228 AquicUdorthens. The organiccarbonabundance wasmea-
on 227 sured in three soil cores collected at P3, P6 and P8. Within
Elevati 226 thefirst30–40cmthepercentagesrangedbetween1.8to2.5;
225 upto70–80cmindepththevaluestendedtodecreaseto0.4–

224 0.8. Atagreaterdepth(to150–170cm)thelevelsincreased
0 20 40 60 80 100 to 7%. The hydraulic conductivity values gradually dimin-
Distance from the crop m ished from 380cmd−1 in the upper part of the transect, to
ground 16-dic-03 29-apr-04 27-lug-04 2cmd−1 at the end of the poplar plantation. The measured
piezometers 24-feb-04 24-giu-04
values fall into the range reported for fine sands rich in silt
Fig.4. Fluctuationsofwatertableprofile,averagedbyrow,during (Fetter,2001).
thestudyperiod. Codesaboverepresentpiezometernestlocations.
•=indicatepositionofthepiezometersscreen. 3.2 Hydrology
A relevant seasonal cycle of the water table fluctuations is
Precipitation data at the meteorological station located at
evidentfromthemeasurementsconductedfromMarch2003
2kmawayfromtheexperimentalareawereused.
● to July 2004. 2003 was an atypical year characterized by
Theanalysiswasperformedonfilteredsamples(0.45µm), a very dry spring and summer, minimum water table levels
except for measurements of pH, electrical conductivity and were measured from the March–November period (Fig. 3).
totalnitrogen(TN),forwhichunfilteredsampleswereused. IntenseprecipitationeventsoccurredattheendofNovember
The alkalinity was measured by two-endpoint potentiomet- and beginning of December and consequently the water ta-
ric titration with HCl. N-NH4 analyzed by molecular ab- bleabruptlyincreasedinallthepiezometers, particularlyin
sorption spectrometry (Perkin Elmer UV-VISLambda2) us- thoselocatedonthehillslope(Figs.3and4). Themaximum
ing the indophenol-blue method. NO3-N, SO4 and Cl were values, occurredduringwinter2004, followedbyminimum
determined by ion chromatography using a Dionex 2000i levelsinsummer2004whenmanypiezometerswereempty.
equipped with AS4A column. TN was measured using Thebreadthofthecycleisgreaterupslopeanddiminishesto-
molecularabsorptionspectrometry,followingpersulphatedi- wardsthelake(Fig.4).AtP1thevariationswereatleast3m;
gestioninanautoclaveat120◦C.Theorganicnitrogen(ON) theyreducedto1.8–1.3mwithinthepoplarplantation,from
wasestimatedbycomparingthedifferencebetweenTNand P2mtoP6,andtheyfurtherdeclinedto0.7–0.5minthereed
+ − −
inorganic N (NH -N + NO -N + NO -N). Dissolved or- swamp. Thelargerangemeasuredonthehillslopesuggests
4 3 2
ganic carbon (DOC) was assayed by high temperature cat- apatternofwinterrechargecontrolledbyconditionsoccur-
alyticoxidationusingaShimadzuTOC-5000Aanalyzer. ringupslope. Incontrast,themoreconstrainedresponseob-
Thequalityofchemicalanalysiswascheckedbyincluding served in the rest of the transect, particularly near the lake,
methodblanks,repeatedmeasurementsofinternalandcerti- suggests that both the hill slope and the lake exert control
fied reference samples and by regular inter-laboratory tests overtheriparianwatertable. Inmid-summer(27July)when
and international intercomparisons (Mosello et al., 2002). the slope discharge fell to under measurable levels, proba-
The repeatabilities, based on repeated measurements of in- bly because of high evaporation losses and lack of rainfall,
Hydrol. EarthSyst. Sci.,12,539–550,2008 www.hydrol-earth-syst-sci.net/12/539/2008/
R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment 543
20 10
Median 25%-75% Min-Max Median 25%-75% Min-Max
18 9
16 8
14 7
12 6
N-NOmg/l3 108 HCOmeq/l3 45
6 3
4 2
2 1
0 0
-2 -1
P1 P2 P3 P4 P5 P6 P7 P8 P9s P9d P1 P2 P3 P4 P5 P6 P7 P8 P9s P9d
24 9
Median 25%-75% Min-Max
21 8
18 7
6
15
Cl mg/l12 DOC mg/l45
9
3
6 2
3 Median 25%-75% Min-Max 1
0 P1 P2 P3 P4 P5 P6 P7 P8 P9s P9d 0 P1 P2 P3 P4 P5 P6 P7 P8 P9s P9d
− − −
Fig.5.BoxplotshowingthedistributionofNO -N,HCO ,Cl andDOCinthepoplarsub-unit.
3 3
theinflowfromthelakealonesustainedthewatertableina depth. Conversely, Cl− SO2− and NH+-N linearly dimin-
4 4
limitedportionoftheriparianzone,fromP7toR3(Fig.4). ished with depth (R−0.933, R−0.625, R−0.556, respec-
tively, p<0,00001). Compared to the poplar sub-unit, the
−
3.3 Groundwaterchemistry reedsubsurfacewaterwasreallypoorinNO -N,withcon-
3
centrationsoftenbeingundetectable(<0.02mgl−1),andrel-
We found strong variations in the chemistry of subsurface ativelyrichinNH+-Nrangingbetween0.010–0.532mgl−1.
4
waterwithinthetwosub-unitssamplingregions(poplarand Likewiseammonia,ONandDOCwereshowninhighercon-
reeds)ofthelakeCandiariparianzone(Table1and2). centrations in the reeds than in the poplar sub-unit. In par-
BoxplotsofFig.5summarizetheconcentrationsofsome ticular at R1s (at lower depth), close to the boundary be-
variables of interest measured in the poplar sub-unit as a tween poplars and reeds, the median concentrations of ON
function of the distance from the crop field. A drastic andDOCreached2.47mgl−1and15mgl−1,respectively.
NO−-N depletion, from 14.7mgl−1 to 0.08mgl−1, is evi- The O2 measures, indicated low levels at both sub-unit
3
dentfromtheupslopetotheboundarybetweenpoplarsand sampling regions (Fig. 7), with median values ranging be-
reeds. Opposite behaviour was observed in HCO− concen- tween0.5–3.3mgl−1. Maximumvalues(around10mgl−1)
3
trations, which gradually increased from the crop field to- were measured at upslope (P1 and P3) while the minimum
wards the lake. Within a length of 60m, the median alka- values(about0.5mgl−1)weremeasuredinthereedswamp
linity ranged from 0.60meql−1, at P1, to 7meql−1, at P8. atthelowerdepth.
The Cl− concentrations exhibited a bell-shaped distribution
with a mean value of 12.6mg/l at site P1, increasing val-
4 Discussion
uestoamedianof17.8mg/latP4,anddecreasingvaluesto
13.7and8.5mgl−1atP9sandP9d,respectively. DOClevels
4.1 Denitrificationindicators
moderatelyincreasedfromamedianof2.1mgl−1,upslope,
to5.3mgl−1downslope. SlightNH+-NandONconcentra- TheresultsobtainedfromourstudyintheCandiaLakebasin
4
tions(0.018–0.027mgl−1 and0.3–0.5mgl−1,respectively) showthatnarrowregionsofterrestrial-aquaticinterfacesare
weremeasuredwithinthepoplars. biogeochemicallyhighlyvariableenvironments.
−
Within the reed swamp the chemistry of subsurface wa- The spatial trend of NO -N concentrations observed at
3
ters also varied strongly mainly with the depth of soils or thepoplarsub-unit,suggeststhataportionofriparianregion
sediments (Fig. 6). In particular, HCO− values increased functionedlargelyasanitratesink.TheNO−-Ngradientwas
3 3
linearly at a depth of (R 0.897, p<0.00001) by a factor verysteep, andcouldrangefrommorethan6mgl−1 tothe
of 5 (from about 2 to 10meql−1) between 50 and 220cm detectionlimitoverdistancesoflessthan5m.
www.hydrol-earth-syst-sci.net/12/539/2008/ Hydrol. EarthSyst. Sci.,12,539–550,2008
544 R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment

12 400
Median 25%-75% Min-Max
350

10
300
8 250
CO meq/l3 6 SO mg/l4125000
H
4 100
50
2
0
0 Median 25%-75% Min-Max -50
R1s R1m R1d R2s R2m R2d R3s R3d R1s R1m R1d R2s R2m R2d R3s R3d
14 0,9
Median 25%-75% Min-Max Median 25%-75% Min-Max
12 0,8
0,7
10 0,6
Cl mg/l 68 N-NH mg/l4 000,,,345
4 0,2
0,1
2
0,0
0 -0,1
R1s R1m R1d R2s R2m R2d R3s R3d R1s R1m R1d R2s R2m R2d R3s R3d
Fig.6.BoxplotshowingthedistributionofHCO−,SO2−,Cl−andNH+-Ninthereedsub-unit.
3 4 4
Both biological and physical processes can produce the thors (Robertson and Kuenen, 1984) isolated an organism
− −
NO -N depletion observed in our study area. Changes in (Thiosphaera pantotropha) that uses both O and NO as
3 2 3
NO− toCl− molarratiosareoftenusedtoinfertheremoval terminalelectronacceptoratconcentrationof90%ofairsat-
3
of nitrate by biological processes as chloride is a conserva- uration corresponding to a concentration of 6.9mg/l. Most
tive ion. Large declines in chloride concentrations along a likelytheO2 thresholdthatmaycausethecessationofdeni-
hydrologic flow path can indicate the dilution of a particu- trificationvariesamongorganisms. Recent“infield”studies
larwatersource(e.g.agriculturalrunoff)byasecondwater indicate that denitrification can occur even with detectable
sourcesuchasdeepgroundwaterorrainfall. Suchapattern O2concentrations. Thismaybeexplainableconsideringthat
canindicatethatassociateddeclinesinnitratelevelsaredue thisbacterialprocessoccurinmicrosites,oftendefined“hot
to the mixing of different water sources rather than actual spots”, where O2 has been completely removed (Mariotti,
biological removal. However, in the current paper median 1986;Parkin,1987).Forexample,hotspotsofdenitrification
chloride concentrations are relatively constant in the upper may occur along root channels where moisture and organic
transect P1–P3 as are median nitrate values. In contrast, in matter content are high (McClain et al., 2003), or in small
the central portion P4–P7 median nitrate concentrations de- patches of organic matter in the C horizon of riparian soils
clinerapidlywhereas,chlorideconcentrationsshowsmallin- (Jacintheetal.,1998). Hilletal.(2000)alsoreportedtheoc-
creases. These patterns do not indicate dilution, but instead currenceofdenitrificationatdepth,attheinterfacesbetween
suggestnitrateremovalisoccurring. sandsandpeatorburiedchanneldepositsinasouthernOn-
tarioriparianzone.Manystudiesonriparianareasreporteda
Thepatternsobservedforsomechemicalspeciessuggests
−
sharpdeclineofNO ingroundwateratO2concentrationof
thatdenitrificationplaysapredominantroleintheobserved 3
− <2–3mg/L(Ceyetal.,1999;VidonandHill,2004)andthis
NO -Nattenuation. Firstly,thelowO levelsinthesubsur-
3 2 canbeinterpretedasaconditionfordenitrificationtooccur.
facewaterindicateconditionsgenerallysuitablefordenitri-
fying bacteria (facultative anaerobes) which, in small zones Wefoundasignificantinverselinearrelationshipbetween
− −
wheretheO iscompletelyremoved,couldshifttoananaer- NO -N and HCO concentrations (R 0.845p<0.0001)
2 3 3
obic metabolism. Hypoxic conditions (O <2mgl−1) were (Fig. 8) that indirectly indicates the occurrence of denitrifi-
2
−
often observed at the locations where a NO -N consump- cation. In fact this microbial reduction implies an alkalin-
3
tion was evident. Some studies examining the relationship ityincreaseof1eqpermoleofnitratereduced(Stummand
− −
between O and denitrification rate, in laboratory experi- Morgan, 1981). The relationship between HCO and NO
2 3 3
ments, reported a sharp inhibition effect of O on the mi- is not generally analysed in experimental surveys aimed at
2
crobial process at restricted range of concentration with av- studying N dynamic in riparian areas. A link between the
−
erage of 10µmol/l O or less (Tiedje, 1988). Other Au- lossofO andNO andtheproductionoflargeamountsof
2 2 3
Hydrol. EarthSyst. Sci.,12,539–550,2008 www.hydrol-earth-syst-sci.net/12/539/2008/
R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment 545
12 10
a) Median 25%-75% Min-Max
10 8 y = -4.873x + 6.318
R2 = 0.715
8 eq/l 6
m
mg/l 6 -O 3
O 2 HC 4
4
2
2
0
0.00 0.40 0.80 1.20
0
P1 P2 P3 P4 P5 P6 P7 P8 P9s P9d NO--N meq/l
3
12 − −
Median 25%-75% Min-Max Fig.8. LinearregressionbetweenNO3-NandHCO3 concentra-
b) tioninpoplarsub-unitgroundwatersamples. Soliddotsrepresent
10
anintenserainyevent(December2003)notincludedintheregres-
sion.
8
mg/l 6 10
O 2
surface
4 8
middle/deep
6
2
eq/l y = -1.983x + 10.48
m
0 R1s R1m R1d R2s R2m R2d R3s R3d 2-4 4 R2 = 0.8106
O
S
2
Fig.7.BoxplotshowingthedistributionofO2inthepoplar(a)and
inthereedsub-unit(b). 0
−
HCO isreportedforaglacialoutwashaquiferinMinnesota
3 -2
by Puckett and Cowdery (2002). In a study we carried out 0 3 6 9 12
in a river wetland (Balestrini et al., 2004 and 2007) the de-
HCO- meq/l
3
clining nitrate found in groundwater along a transect, from
acropfieldtothestream, wasnotrelatedtoalkalinity. The Fig.9. RelationshipbetweenSO2−vs. HCO−concentrationfrom
alkalinity seems to be a sensitive indicator of the extent of 4 3
groundwatersamplescollectedatreedsub-unit. Theequationand
denitrification but only in specific types of wetlands. Our thecoefficientofdeterminationreferstothelinearregressionfound
data set and literature are not broad enough to identify the forthesurfacesamples.
environmental features, presumably geological and hydro-
−
logical, which favour HCO accumulation in groundwater −
3 Withinthereedswamp,theverylowinputofNO -Nfrom
asaconsequenceofredoxreactions. 3
both the adjacent terrestrial zone and the water of the lake,
Anadditionalsignofdenitrificationactivitywasthepres- probablyseverelyconstrainedthedenitrification. Inthissys-
ence of detectable NO−2-N concentrations at the P2–P7 tem,whichwasverypoorinoxygen,SO24− couldbecomea
piezometers of the poplar sub-unit (Table 1). The dynamic suitableelectronacceptorfortheoxidationoftheorganicC.
−
ofNO2 insoilandwatercanbeveryrapidasitcanbepro- As sulphate reduction also produces alkalinity (Stumm and
duced or consumed by a number of biological or chemical Morgan,1981)thehighsignificantinverserelationfoundbe-
reactions. Somestudiesreportthatnitrificationanddenitrifi- tween SO2− and HCO− (R 0.901, p<0.00001; Fig. 9) sup-
− 4 3
cationarethemainprocessescontributingtoNO2 accumu- ported our hypothesis. In addition, the release of sulphide
+
lationinsoil(Burnsetal.,1996). InourstudyareaNH -N was frequently detected during the sampling, by the typical
4
concentrationingroundwaterwasverylowand,conversely, smellofrotteneggs. Thepatterndescribedfollowsthewell-
−
therewasagreatinputofNO -N,consequentlywecanex- studiedthermodynamicsequencereportedforlakeormarine
3
−
pectthatdenitrificationwasthemajorNO sourceprocess. sediments (Stumm and Morgan, 1981; Kuivila and Murray,
2
www.hydrol-earth-syst-sci.net/12/539/2008/ Hydrol. EarthSyst. Sci.,12,539–550,2008
546 R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment
230 230
a) Spring-summer 2003 a) Spring-summer 2003
m a.s.l. 228 0.6 2.6 0.3 0.1 0.06 0.2 m a.s.l. 228 2.2 3.4 5.0 7.7 31 6.6
Altitude 222246 0.04 0.06 0.02 Altitude 222246 6.4 6.0
222 222
0 20 40 60 80 100 120 0 20 40 60 80 100 120
230
230 b) Winter 2004
b) Winter 2004
Altitude m a.s.l. 222222468 13.6 6.8 14.0 7.8 6.15.11.60.0702..070 00..022 0.00.300 0.2 Altitude m a.s.l. 222222468 2.5 3.4 3.5 5.1 4.8 5.43.73.584..01 7.859..626.576..7496..20 6.9
0.03
222
222 0 20 40 60 80 100 120
0 20 40 60 80 100 120
230
c) Spring 2004
230
c) Spring 2004
Altitude m a.s.l. 222222468 15.0 110. 13.2 8.5 5.45.47.50.100..032 00.0.13 00.0.066 0.2 Altitude m a.s.l. 222222468 1.7 2.4 4.3 3.2 3.33.33.14.342..89 155.06.26.82.375.2.9 8
0.02
222
0 20 40 60 80 100 120
222
0 20 40 60 80 100 120 230
d) Summer 2004
230
d) Summer 2004 s.l. 228 1.6
a.
de m a.s.l. 222268 15.7 10.7 8.5 0.9 0.00 0.05 0.02 0.03 0.1 ● Altitude m 222246 2.1 2.8 3.96.77.55.3 44.2.536 7.241 71.22 8.39.7
Altitu 224 0.060.0●40.10.00 0.020.03 222 5.5
0 20 40 60 80 100 120
222 Distance from the cropfield (m)
0 20 40 60 80 100 120
Distance from the cropfield (m) ≥ 10 mg O2 l-1 <1 ≤ 2 mg O2 l-1
0.1- 0.4 mg N-NO l-1 >2 < 6 mg O2 l-1 ≤ 1 mg O2 l-1
2
ground water glacial till ground water glacial till
Fig. 10. MFeiga.n 10n. iMtreaante nitaranted anndi tnriitrtiete (( )) cocnocnenctreantiotrnas t(imons (mg N l−1) Fig.11.MeanDOCconcentrations(mgl−1)alongthestudiedtran-
alongthestudiedbetrr a20n0s3e, cb)t Ddeucerminbegr:-M(aarc)h M200a4r,c ch) –September2003,(b) sectanddissolvedoxygenrangesduring: (a)March–August2003,
December–March 2004, (c) April–May 2004 and (d) June–July (b)December–March2004,(c)April–May2004and(d)June–July
2004. 2004.
1984, Dillon et al., 1997), but is not often used to inter- facttheoccurrenceofmethanogenesis,expectedonthermo-
pret the chemical transformation in riparian zones (Hedin dynamicbasis,doesnotaffectalkalinity. Ontheotherhand
et al., 1998; Pukett and Cowdery, 2002; Hill et al., 2000). we have to exclude the effect of dilution processes caused
AsshowninFig.8atthehigherdepth(150–220cm)where by the inflow of lake water (see Tables 1 and 2). We could
all the sulphate was consumed, the alkalinity continued to hypothesise the presence of older waters at depth, isolated
increase,fromabout6to11meql−1 andthechloridegradu- bysurfacewaters,characterisedbylowerCl−andSO2−and
4
−
allydecreased.Boththesefindingsaredifficulttoexplain.In higherHCO concentrations.
3
Hydrol. EarthSyst. Sci.,12,539–550,2008 www.hydrol-earth-syst-sci.net/12/539/2008/
R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment 547
Inthereedswamp,characterizedbystillwaterandabun-
10
dantplantbiomass,themainsourceofmineralizednitrogen
wasrepresentedbyammonia,butthemajornitrogenspecies
8 y = 4.82e-0.058x
isorganicnitrogenwithrelativelyhighconcentrations,rang-
ing from 0.35 to 4.3mgl−1. The restricted amount of ni- g/l 6 R2 = 0.540
tratereachingthereedstanddrivestosupposethatotherpro- m
cessesinsteadofDissimilativeNitrateReductiontoAmmo- C
niumcouldhaveacrucialroleintheNdynamicinthisunit. O 4
D
On the other hand the great availability of organic matter
deriving from plant decomposition suggest the importance 2
of ammonification. The possibility that the Phragmites belt
couldactasasourceofNH4 andONforthelakehasto be 0
evaluatedwithmorespecificinvestigations.
0 5 10 15 20
4.2 Environmentalfactorsaffectingnitrogendynamics NO --N mg/l
3
Theresultsgainedfromthehydrochemicalmonitoringover −
Fig. 12. Relationship of NO -N vs. DOC concentrations from
3
18monthsallowedustotracethesequenceofprocessesthat
groundwatersamplesatthepoplarsub-unit.
occurredinthelakeCandiawetlandoverthecourseofayear
andtoidentifythemainenvironmentalvariablesaffectingni-
trogencycling. Thesequenceofhydrologicalandbiochem- tionscharacterizedbyvalueslessthan2mgl−1(Fig.11c).In
ical changes are presented in Figs. 10 and 11 as a series of addition,wedetectedthepresenceofnitrite,anintermediate
schematicsnapshots. productofthedenitrificationreaction,inlocationscharacter-
The long period of drought during spring-summer 2003 ized by low oxygen, and an elevated availability of nitrate
causedtheprogressiveloweringofthewatertableinallthe (Figs.10–11c).
transectresultinginthecompletedisconnectionbetweenthe Lateron,overthedryandhotsummerseason(Fig.10d),
poplar and reed sub-units in September. In these condi- thewatertableleveldeclinedatalllocationsuptoP7,aswell
tions the spring fertilizer applications did not generate the asthenitrateconcentrations.Amarkedandabruptdeclineof
expected rise of NO− in the groundwater and the observed NO−-N was evident between P3 and P4 where the concen-
3 3
levelsremainedverylowduringallof2003(Fig.10a). trationdecreasedfrom8.5to0.9mgl−1 within10m. Thus,
The abundant precipitation which occurred in Decem- the nearly complete nitrate removal (NO−-N<0.1mgl−1)
3
ber 2003 promoted the prompt development of the shallow
occurred in a larger area covering 60% of the whole wet-
aquiferandtheconsequenthydrologicalconnectionofallthe
land, despite the constant input of nitrate from the P1 site.
piezometers (Fig. 10b). The interaction of the rising wa-
The O levels further diminished and, conversely, the DOC
− 2
ter table with the NO rich soil waters of the crop field,
3 availability increased mainly in the transition area between
−
caused a sharp increase of NO concentrations, simultane- poplar and reeds (Fig. 11d). At the end of July when the
3
ouslyfromP1toP6sites. AtP7(67mfromthecropfield) completedryingoutoftheshallowaquifercausedadiscon-
−
andP9s(75mfromthecropfield,at80cmdepth)theNO nection between upland and the locations closer to the reed
3
flush was detected a month later, during the January sam- swamp(P7–P9),theNO−sourceareastoppedsupplyingthe
− 3
pling. Conversely, the NO concentration did not rise at riparianzone.
3
thedeeperpiezometers,P8andP9d(157and216cmdepth, TheseasonalevolutionofNO−-Ninthesubsurfacewaters
3
respectively) and in the whole reed sub-unit. The elevated ofthepoplarsub-unitsuggeststheimportanceoftemperature
O2 concentrations measured during the February sampling in the biological processes responsible for N removal. The
(Fig.11b),withlevelsover10mg/lfromP1toP3,indicated increaseintemperatureof5–7◦Cobservedinthegroundwa-
the inputof waterrich inelectron donors. Inthe rest of the ter from February to the end of May could have stimulated
−
transect the NO3 concentrations gradually decreased along thedenitrification. Somestudieshavereportedasignificant
withtheoxygenvalues. direct effect of temperature on denitrification rates (Martin
During spring (Fig. 10c), the water table remained quite etal.,2000; Knowles,1982). Therenewalofthevegetation
−
high and the NO3 continued to flush from the upland (P1), activityinthespringandsummermonthsmayalsobeanim-
wherethelevelsweresteadyaround14–15mgNl−1,thereby portant factor in the decreasing nitrate losses from the soil
−
ensuringthesupplyofNO fordenitrificationintheriparian at that time of the year. Higher daily soil respiration rates
3
−
zone. TheNO patterndidnotdiffersignificantlyfrompre- and an abundant biomass of live roots in spring and early
3
viousmonthsbutconditionsappearedmoresuitableforden- summerwerereportedforapoplarstandinariparianareain
itrification. In fact the O concentrations were remarkably Central Iowa (Tufekcioglu et al., 1999). The rapid increase
2
lower compared to the winter period with most of the loca- ofDOCconcentrationobservedfromMaytoJuneatcertain
www.hydrol-earth-syst-sci.net/12/539/2008/ Hydrol. EarthSyst. Sci.,12,539–550,2008
548 R.Balestrinietal.: Lacustrinewetlandinanagriculturalcatchment
sites(P5–P9)mayderivefromgreaterorganicmatterinputs R3) the fluctuations of the water table were more restricted
fromplantstothesoil,thereforeprovidingbetterconditions andduringseveraleventsthegroundwaterreachedandrose
for nutrient sequestration within the riparian area. Some abovethesoilsurface. Recentresearchcarriedoutindiffer-
studies suggest that denitrification can be limited by a sup- ent riparian areas spread all over Europe demonstrated that
plyoforganiccarbon(Hedinetal.,1998;StarrandGillham, thefunctioningofriparianareasdependsontheexistenceof
1993; Bradley et al., 1992; Hill et al., 2000) as it provides topographicandsoilconditionsthatproduce,atleastseason-
energytothebacterialcommunityand,indirectly,promotes ally,ahighwatertable(Burtetal.,2002). Thesaturationof
the occurrence of anoxic conditions through O consump- theuppersoillayersrichinorganicCcreatesasuitablecon-
2
tion by heterotrophic bacteria. Hedin et al. (1998) reported dition for denitrification (Burt et al., 1999; Cosandey et al.,
that NO only accumulated in subsurface waters with DOC 2003)andconsequentlyenhancestheNbuffereffectofripar-
3
concentrations of less than 2mgl−1 in Smith Creeck near- ian zones. The development of soil saturation also depends
stream environments; while other Authors (Starr and Gill- on thehydraulic conductivity ofthe soil whichis important
ham,1993)suggestthatconcentrationsof4–5mgl−1arenot in optimizing the ground water residence time (Burt et al.,
enoughtosustaindenitrification. Therelationshipfoundbe- 1999 and 2002). The aquifer has to be permeable enough
−
tweenDOCandNO -NusingthedatarelativetoP1–P9sites to support interactions with the biotic components or allow
3
(Fig. 12), is similar to that reported by Devito et al. (2000) forthedevelopmentofanoxicconditions. Ontheotherhand
forasandaquiferinOntario(Canada)andsuggeststhatden- low hydraulic conductivity may give origin to a subsurface
−
itrification can be limited by a carbon supply of less than fluxofNO -Nthatistoosmalltobeeffective. Intheupper
3
3mgl−1. DOC concentrations always higher than 3mgl−1 partofthestudyareathepresenceofconductivesoillayers
havebeenmeasured,withinthepoplarsub-unit,startingfrom (138–328cmd−1) in depth permitted the transport of water
−
P5 to P9 (Fig. 11). The presence of black soil layers rich richinNO . Thediminishingofthehydraulicconductivity
3
in organic carbon deposits were also detected at a greater tovaluesrangingfrom67to2cmd−1,respectivelyfromP5
depth (to 150–170cm) at the same sites. On the contrary, toP8mostlikelypromotedthebacterialprocesses.
−
at P1 where the NO level did not decline even in summer,
3
theDOCconcentrationneverexceeded3mgl−1.
In addition to its role as an organic C supplier, the direct 5 Conclusions
roleofvegetationinincreasingtheNuptakerateduringthe
growing season has to be expected. In this regard, it is im-
Our study of Lake Candia clearly demonstrates the crucial
portant to note that the temperature increase, typical of the
roleofwetlandsintheabatementofNfluxfromagricultural
summer months, coincided with soil drying and a lowering
sourcestoaquaticecosystems.Themetabolicactivityofbac-
ofthewatertable. Somestudieshavedemonstratedthatthe
terialcommunities, basedonthedifferentialuseofelectron
nutrientplantuptakeiscontrolledbythewatertablefluctu-
donorsandacceptorsinredoxreactionstoderiveenergyfor
ations, asplantrootbiomasshasbeenshowntosharplyde-
growth, is the key function of this system. In a thermody-
cline with depth (Ehrenfeld et al., 1992; Tufekcioglu et al.,
namicperspectivewecanidentifythreedifferentfunctional
1999). Within the poplar sub-unit, the water table lowered
units within the study area: 1) the steep upslope zone ad-
by about 70cm reaching a depth of 1–3m during the sum- −
jacenttothecropfieldwhereNO -Nrichandoxicground-
3
mer.Wedidnotinvestigatethedistributionoftherootsinthe
watersupportedtheaerobicrespiration,2)aflatterzoneadja-
soil profile but a number of papers indicate that poplars are
centtothereedswampdominatedbydenitrificationthatcon-
generally characterized by a very deep root system. Tufek- −
sumed all NO , 3) the reed swamp rich in electron donors,
3
cioglu et al. (1999) detected the presence of coarse roots of
especiallyDOC,andpoorinbothoxygenandnitrate;there-
at least 180cm under poplars in a fine-loamy riparian soil;
dox condition favoured bacteria capable of deriving energy
Heilman et al. (1994) reported that roots of 4-year-old hy-
fromsulphate. Thebiochemicalresponsesoccurringinthese
brids of Populus trichocarpa x Populus deltoides extended
differentunitsarecoupledtothehydrologicalresponsesand
to depths beyond 3.2m and Gifford (1966) reported a simi-
thedifferentreactiontorainfallinputsovertheyear. Theto-
lar rooting depth, about 290 cm, for Populus tremuloides in
pographyassumesacriticalroleinregulatingthestrengthof
asandyloamsoil. Onthebaseofthesefindingsitisreason-
thehydrologicalconnectionswithintheriparianarea.
abletohypothesizethatpoplartreesplayaroleinlowering
This successful experimental approach, which comprises
thesubsurfacenitrateconcentrationsduringthegrowingsea-
bothbiogeochemicalandhydrologicalmonitoring,allowsus
sonatCandiaLakewetland.
togainknowledgeoftheroleplayedbyphysicalandbiolog-
Theresultsobtainedindicatethattheefficiencyofthewet- icalprocessesincontrollingtheNdynamicwithinthisspe-
land study in buffering the flux of N is also associated with cificportionoftheCandiaLakecatchment. Althoughthere-
−
thetopographicfeaturesofthearea,inthatsignificantNO - sultspresentedhereareobtainedfromasinglecasestudyina
3
Nremovaloccurredinaspecificportionofthetransectchar- specificclimaticenvironment,theymaycontributetodevel-
acterizedbythediminishingslope. Inthisarea(fromP5to opingmodelsonNcyclingprocessesincatchmentswhereN
Hydrol. EarthSyst. Sci.,12,539–550,2008 www.hydrol-earth-syst-sci.net/12/539/2008/