RIPARIAN WETLANDS: HYDROLOGY
MEETS BIOGEOCHEMISTRY
In e ac ions be ween hyd ology and
biogeochemis y wi hin ipa ian we lands
Po en ial implica ions o in e nal biogeochemical
p ocess dis ibu ions and solu e expo s
Disse a ion zu E langung des G ades Dok o de Na u wissenscha en (D . e . Na .)
an de Fakul ä Biologie/Chemie/Geowissenscha en de Uni e si ä Bay eu h
Vo geleg on
S en F ei
Geb. am 21. Juni 1979 in Augsbu g
Die o liegende Disse a ion wu de im Zei aum on Ap il 2008 bis Ok obe 2012 un e de
Be euung on D . Jan H. Fleckens ein am Leh s uhl ü Hyd ologie (P o . D . S e an
Pei e ) de Uni e si ä Bay eu h ange e ig .
Die A bei en im Rahmen de Disse a ion wu den du ch die Deu sche
Fo schungsgemeinscha (DFG) ge ö de im Rahmen des P ojek es Fl 631/6-2, einem
Teilp ojek inne halb de DFG Fo sche g uppe FOR 562.
Volls ändige Abd uck de on de Fakul ä ü Biologie, Chemie und Geowissenscha en de
Uni e si ä Bay eu h genehmig en Disse a ion zu E langung des akademischen G ades
eines Dok o s de Na u wissenscha en (D . e . na .).
Disse a ion einge eich am: 05.10.2012
Zulassung du ch die P ü ungskommission: 17.10.2012
Wissenscha liches Kolloquium: 28.03.2013
Am ie ende Dekan: P o . D . Bea e Lohne
P ü ungsausschuss:
D . Jan H. Fleckens ein (E s gu ach e )
P o . D . S e an Pei e (Zwei gu ach e )
P o . D . Michael Hauhs (Vo si z)
P o . D . Be nd Huwe
P o . D . Egbe Ma zne
RIPARIAN WETLANDS: HYDROLOGY MEETS
BIOGEOCHEMISTRY
In e ac ions be ween hyd ology and biogeochemis y wi hin ipa ian
we lands
Po en ial implica ions o in e nal biogeochemical p ocess dis ibu ions and
solu e expo s
Vo lu e nahe Feuch gebie e: Hyd ologie i Biogeochemie
In e ak ionen zwischen Hyd ologie und Biogeochemie in o lu e nahen
Feuch gebie en
Po en ielle Auswi kungen ü die in e ne biogeochemische P ozess e eilung
und au den Expo gelös e S o e
Ex ended Summa y
Acknowledgemen s
I would like o hank Jan H. Fleckens ein o he supe ision and help ul ad ice du ing all
phases o his wo k.
I would like o hank S e an Pei e o he oppo uni y, o wo k a he depa men o
Hyd ology and Klaus-Holge Kno o his suppo and his excellen con ibu ions o his wo k
and he occasional climbing sessions.
I would like o hank all membe s o he Hyd ology depa men and all he assiduous s uden
assis an s o hei help. Wi hou hei suppo , his wo k would no ha e been possible:
Sand a We b, Ch is ophe Shope, S enja Ba sch, Ch is iane Clemens, Ch is ian Es op, S e an
S ohmeie , Jü gen Leonbache , Ch is iane Neuman, Johannes Opi z and Sebas ian Wü ze .
I would like o hank Rob McLa en, Young-Jin Pa k, And ea B ook ield and Ed Sudicky a
he Uni e si y o Wa e loo, Canada o hei in aluable help wi h he ins and ou s o he
nume ical code Hyd oGeoSphe e. Fu he mo e, I would like o hank Daniel Pa ing on
School o Ci il, En i onmen al and Mining Enginee ing (Uni e si y o Adelaide) o he
use ul and p oduc i e coope a ion.
I would like o hank he help ul coo dina o s and echnicians o he Resea ch G oup FOR
562.
I would like o hank all people p o iding ad ice and help ul commen s and some imes he
necessa y dis ac ion. Pa icula ly, I wan o hank Ma ianne Ruidisch, Ma in Reiche ,
Sabine Thüns and T ang
Tôi cám ơn gia đình Việ nam của ôi Chú Tụng à Cô Phương đã cho ôi mộ mái ấm gia đình
nồng hậu ở Bay eu h à nhấ là Cô Phương,người ấ hường xuyên quan âm chăm sóc ôi, ấ
cả họ đều đã đóng góp mộ phần quan ọng cho sự hành công ong công iệc của ôi.
I would like o hank my amily and my pa en s Ing id and We ne o he suppo du ing all
phases o my s udies.
I would like o hank Hugo o he help, he pa ience and o he good ime.
[1]
TABLE OF CONTENTS
Table o Con en s
Table o Con en s .................................................................................................................................... 1
Lis o Figu es ......................................................................................................................................... 3
Lis o Tables .......................................................................................................................................... 4
Summa y ................................................................................................................................................. 5
Zusammen assung ................................................................................................................................... 7
1In oduc ion ..................................................................................................................................... 9
1.1In e ac ions be ween hyd ology and biogeochemis y - an in e disciplina y challenge ......... 9
1.2Ripa ian We lands: Complex hyd ology mee s complex biogeochemis y .......................... 11
2Resea ch Objec i es and Hypo heses ........................................................................................... 15
3Ma e ials and Me hods .................................................................................................................. 17
3.1S udy Si e .............................................................................................................................. 17
3.2Hyd ological Modeling ......................................................................................................... 19
3.2.1Vi ual We land Modeling (S udy 1, 2 and 3) ............................................................... 20
3.2.2Ca chmen Scale Modeling (S udies 4 + 5) ................................................................... 23
3.3Biogeochemical Modeling (S udies 2 + 3) ........................................................................... 27
3.3.1Coupling Hyd ology and Biogeochemis y .................................................................. 27
3.3.2Implemen ed Reac ion and Bounda y Condi ions ........................................................ 28
4Resul s and Discussion ................................................................................................................. 33
4.1E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno gene a ion in a
i ual ipa ian we land (S udy 1) ..................................................................................................... 33
4.2Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in we land sys ems –
a i ual modeling expe imen . (S udy 2) ......................................................................................... 36
4.3Rep esen ing e ec s o mic o- opog aphy on uno gene a ion and sub-su ace low
pa e ns by using supe icial ill s o age heigh a ia ions (S udy 3). ............................................... 39
4.4Concen a ions and luxes o dissol ed o ganic ca bon in uno om a o es ed ca chmen :
insigh s om high equency measu emen s (S udy 4) ..................................................................... 42
4.5In e p e ing low gene a ion mechanisms om in eg a ed su ace wa e -g oundwa e low
models o a ipa ian we land and ca chmen (S udy 5). .................................................................... 44
5Conclusions and Ou look .............................................................................................................. 47

[2]
TABLE OF CONTENTS
6Re e ences ..................................................................................................................................... 49
7Appendix ....................................................................................................................................... 59
8Con ibu ions o he included manusc ip s .................................................................................... 63
S udy 1:E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno gene a ion in a
i ual ipa ian we land ......................................................................................................................... 65
S udy 2:Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in we land sys ems –
a i ual modeling expe imen . ............................................................................................................. 97
S udy 3:Rep esen ing e ec s o mic o- opog aphy on uno gene a ion and sub-su ace low
pa e ns by using supe icial ill s o age heigh a ia ions .................................................................. 151
S udy 4:Concen a ions and luxes o dissol ed o ganic ca bon in uno om a o es ed ca chmen :
insigh s om high equency measu emen s ...................................................................................... 179
S udy 5:In e p e ing low gene a ion mechanisms om in eg a ed su ace wa e -g oundwa e low
models o a ipa ian we land and ca chmen ....................................................................................... 207
E klä ung ............................................................................................................................................ 249
[3]
LIST OF FIGURES
Lis o Figu es
Figu e 1: T adi ional hyd ologic and biogeochemical pe spec i e on anspo and eac ion ............. 10
Figu e 2: Concep ual model o he Lehs enbach ca chmen ................................................................ 18
Figu e 3: Pic u e o he Schlöppne b unnen II ield si e. ..................................................................... 18
Figu e 4: Geome y o he i ual we land segmen : a) plana e e ence model showing he main
d ainage di ec ion and channel loca ion; b) smoo hed ealiza ion o he we lands hummocky
mic o- opog aphy; c) c oss sec ion (Y=5m) o he mic o- opog aphy model. ............................. 21
Figu e 5: Fini e elemen g id o he Lehs enbach ca chmen model.. .................................................. 23
Figu e 6: Obse ed and simula ed discha ge alues (es ima ed a he ca chmen ou le ) o he
calib a ion and alida ion pe iods o he ca chmen scale model.. ............................................... 26
Figu e 7: Concep o he applied s eam ube app oach o ep esen a ion o biogeochemis y along
isola ed subsu ace low pa hs (dashed line). ............................................................................... 28
Figu e 8: Typical oxygen dep h p o ile obse ed o a we land si e o he Lehs enbach ca chmen . .. 31
Figu e 9: Six consecu i e snapsho s o he e ol ing su ace low ne wo ks du ing he la ges low
e en o he yea (day 217 o day 218).. ....................................................................................... 33
Figu e 10: a) Rela ionship be ween discha ge and g oundwa e le el o wo peak low e en s,
obse ed o a small ca chmen loca ed in B i ish Colombia, Canada (modi ied a e Fi zge ald e
al. (2003)).b) Simula ed ela ionship be ween g oundwa e le el and channel discha ge o he
mic o- opog aphy model. .............................................................................................................. 35
Figu e 11: Resul s o he biogeochemical simula ions shown o he sul a e educ ion p ocess o he
mic o- opog aphy scena io wi h he mean leng h 0.5m. ............................................................... 38
Figu e 12: Snap sho s aken a he end o a s eady ain all simula ion showing he ully de eloped
su ace low ne wo ks (yellow) which a e gene a ed in he mic o- opog aphy model as well as in
he models wi h ill s o age heigh a ia ions (p- s-low and p- s-high) bu no o he plana
e e ence case. ............................................................................................................................... 40
Figu e 13: Typical non-linea and hys e e ic ela ionships be ween obse ed DOC concen a ions in
uno and discha ge ..................................................................................................................... 43
Figu e 14: Calcula ed s eam and o e land low gene a ion, es ima ed by applying he “hyd aulic
mixing-cell” me hodology o he Lehs enbach ca chmen model.. ............................................... 45
Figu e A1: Soil e en ion unc ions used o ep esen a iably sa u a ed low in he we land soils and
he egoli hic aqui e o he ca chmen scale nume ical model and he i ual we land model... . 59
Figu e A2: Sa u a ed hyd aulic conduc i i ies Ksa assigned o he en sub-laye s SL1-SL10 o he
we land a eas o he ca chmen scale model..... ........................................................................... 60
[4]
LIST OF TABLES
Lis o Tables
Table 1: C i ical concen a ions which a e con olling he sequen ial ini ializa ion o he edox
sequence... ..................................................................................................................................... 29
Table A1: O e iew o he pa ame e iza ion o he ca chmen scale model o ep esen
su ace/subsu ace low and in e ac ions o he h ee di e en zones (we lands, upslope a eas
and s eam a eas).... ...................................................................................................................... 61
[5]
SUMMARY
Summa y
In e ac ions be ween hyd ology and biogeochemis y a a ious spa io- empo al scales a e impo an
con ol mechanisms wi hin e es ial and aqua ic ecosys ems and exis among di e en compa men s
and ansi ion in e aces. Unde s anding he undamen al mechanis ic couplings be ween hyd ological
and biogeochemical p ocesses and how hese couplings eed back in o ecosys em se ices and
unc ions is an in e disciplina y challenge ha mus be add essed especially in he con ex o humanly
media ed clima e change. Ripa ian we lands, as a ansi ion zone be ween e es ial and aqua ic
ecosys ems, occupy la ge ac ions o e es ial ecosys ems and p o ide impo an ecohyd ological
se ices. Due o hei anoxic en i onmen s, ipa ian we lands a e able o s o e signi ican amoun s o
ca bon as pea and ac as an e ec i e nu ien sink e.g. o sul u , phospho ous and ni ogen. Ripa ian
we lands a e cha ac e ized by highly dynamical in e ac ions be ween hyd ologically con olled
anspo mechanisms and biogeochemically con olled subs a e a ailabili y, which go e ns nu ien
cycling as well as he sink and sou ce unc ions o we lands. Gene ally, hese in e ac ions and hei
po en ial implica ions on ecosys em unc ions a e only poo ly unde s ood. The ep esen a ion o he
igh couplings be ween hyd ology and biogeochemis y in mechanis ic models is a e y challenging
ask because hey ha e e ealed a complexi y which is o en beyond he capabili ies o cu en
models. The objec i e o his hesis is o in es iga e in e ac ions be ween hyd ology and
biogeochemis y in ipa ian we lands and o unde s and hei po en ial implica ions o in e nal
biogeochemical p ocess dis ibu ions and solu e mobiliza ion. Addi ionally, one majo ocus o he
hesis is he a emp o ep esen such undamen al couplings in a p ocess-based,
hyd ological/biogeochemical modeling app oach. To his end, his hesis uses a combina ion o ield
and i ual expe imen s, as well as ca chmen -scale nume ical modeling, pe o med o he
Lehs enbach ca chmen , which was exempla ily chosen as main s udy si e.
Resul s om he i ual expe imen s show e y complex small-scale hyd ological dynamics wi hin
he ipa ian a eas. He e, uno gene a ion p ocesses a e s ongly in luenced by he spa ial s uc u e o
he we land- ypical mic o- opog aphy (hummocks and hollows). Su ace low is episodically
gene a ed by a highly dynamical, h eshold-con olled p ocess whe e ex ended su ace low ne wo ks
d ain la ge ac ions o he we land's a ea. Du ing in ensi e ains o m e en s hese su ace low
ne wo ks, which con ibu e o s eam discha ge due o a ill and spill mechanism, domina e uno
gene a ion. These as low componen s a e cha ac e ized by e y low esidence imes (minu es o
hou s) and once hey a e ac i a ed, he su ace low ne wo ks a e able o apidly mobilize la ge
amoun s o solu es, like ni a e o dissol ed o ganic ca bon (DOC), ou o he we lands by bypassing
deepe anoxic laye s. The impo ance o as low componen s o he ca chmen -scale mobiliza ion o
DOC was u he con i med by ield in es iga ions and ca chmen -scale nume ical modeling. High
equency measu emen s o DOC in uno o he Lehs enbach ca chmen e ealed ha DOC expo is
[12]
INTRODUCTION
hese educ ion p ocesses occu sequen ially, known as he mic obially in luenced edox chain
(Zehnde , 1988). The loca ion o he edox-cline in we lands, as he de ined bounda y be ween he
educed and oxidized en i onmen , is igh ly coupled o he loca ion o he local wa e - able (Ci mo
and McDonnell, 1997). Rapid luc ua ions o he wa e able in esponse o onse o ain all a e a
commonly obse ed phenomenon in we land sys em (Ci mo and McDonnell, 1997; De i o and Hill,
1997; De i o and Hill, 1997). The apid esponse o he wa e - able o ain all is discussed in he
li e a u e as an e ec o a la ge capilla y inge in nea - su ace laye s o soil o pea , whe e small
amoun s o ain all o snowmel may esul in apid upwa d mo emen o he wa e - able (Gillham,
1984; Helio is and DeWi , 1987). Wa e le el manipula ion expe imen s in he ield (Kno e al.,
2009; Kno and Blodau, 2009) ha e demons a ed ha luc ua ions o he wa e - able a e di ec ly
linked o apid changes in he p edominan edox p ocesses (i.e. i on(III) educ ion, sul a e educ ion
and me hanogenesis), he loca ion o he edox-cline and he mine aliza ion o o ganic ma e ial.
A he landscape scale we lands a e commonly assumed o be e ec i e sinks o solu es like sul a e o
ni a e, because anae obic condi ions and la ge ca bon supplies enhance educ i e biogeochemical
ans o ma ions like deni i ica ion o sul a e educ ion (Johns on, 1991). Howe e , his pe spec i e
neglec s ha physically-con olled anspo and biogeochemical ans o ma ion p ocesses wi hin
we lands a e no s a ic. Hyd ology, biogeochemis y and hei in e ac ions a e dynamic p ocesses,
especially in we lands o ipa ian a eas, which a e equen ly a ec ed by apid luc ua ions in
hyd ological and me eo ological bounda y condi ions (Ci mo and McDonnell, 1997; Kno e al.,
2009; Kno and Blodau, 2009). Sho and long e m luc ua ions o he hyd ological and
me eo ological d i e s ha e he po en ial o al e in e nal biogeochemical p ocesses, which may
cons ain he sink and sou ce unc ions o we lands o ce ain mine als, gases and solu es (Kno e
al., 2009). De i o and Hill (1997) ha e shown ha we lands a e an e icien ne sink o sul a e du ing
high low condi ions whe e high wa e ables and anoxic condi ions enhance educ i e ans o ma ion
p ocesses e.g. deni i ica ion o sul a e educ ion. Howe e , du ing ex ended d ough pe iods and
d opping wa e ables, edox condi ions wi hin we lands change as we land laye s a e being ae a ed,
leading o inc eased mine aliza ion and e-oxida ion o educed species like sul ide o ammonium,
which a e being lushed du ing s o m uno . Unde hese condi ions, we lands can u n in o an
episodic sou ce o ni a e o sul a e (De i o and Hill, 1997).
In ca chmen s, upland a eas and ipa ian we lands a e usually connec ed hyd ologically, meaning ha
wa e o igina ing om upland a eas has o pass h ough he ipa ian we lands i s be o e i can each
he s eams o i e s ia subsu ace low. G oundwa e om upland a eas usually has a e y di e en
chemical signa u e compa ed o he po e wa e o he we land. In compa ison, po e wa e in he
we lands g oundwa e om upland a eas is o en en iched in oxidized species like sul a e, ni a e o
oxygen, whe eas in con as o we land a eas ca bon loadings a e usually low. Along low pa hs,
whe e upland g oundwa e is exposed o he anoxic condi ions wi hin we lands, compounds like
ni a e o sul a e can be educed e icien ly (Hill e al., 2000; McMahon, 2001). Howe e , in ensi e

[13]
INTRODUCTION
ain all o snowmel may esul in he gene a ion o e y as low componen s like su ace o shallow
subsu ace low ((De i o and Hill, 1997; Lischeid e al., 2007) wi hin ipa ian we lands. These as
low componen s ha e e y low subsu ace esidence imes and he po en ial o apidly anspo
wa e o igina ing om hillslope a eas o he s eams by sho -ci cui ing o bypassing he anoxic a eas
o we lands (Wiging on e al., 1990; Mu doch and S odda d, 1992; S odda d, 1994; DeWalle and
Swis ock, 1994). Unde such condi ions, he sink unc ion o we lands o ni ogen o sul u can be
deac i a ed empo a ily.
A emp s o desc ibe and ep esen he complex p ocesses and couplings be ween he hyd ology and
biogeochemis y o we lands in mechanis ic models is a challenging ask, as p ocesses and couplings
a e commonly a a le el o complexi y ha is beyond he capabili ies o cu en models (Hill, 1993;
Wadding on e al., 1993; Eshleman e al., 1994; Richa dson e al., 2007a). O en, below g ound
p ocesses wi hin we lands a e ea ed as a black box (Ke unen e al., 1999; Updeg a e al., 2001;
Chimne and Coope , 2003) whe e only he ans e cha ac e is ics be ween inpu and ou pu a iables
a e being conside ed, neglec ing unde lying physical laws ha go e n sys em-in e nal hyd ological
and biogeochemical p ocesses. To “unlock he black box” (Walling, 1983), i is necessa y o gain an
imp o ed unde s anding o sys em-in e nal p ocess mechanisms and undamen al mechanis ic
couplings be ween physical anspo and biogeochemical eac ions (Bu and Pinay, 2005), especially
in such complex en i onmen s as ipa ian we lands. This equi es spa ially-explici , physically-based
model s uc u es (Bu and Pinay, 2005; Richa dson e al., 2007b; Boano e al., 2010) which ep esen
p ocesses based on hei ac ual go e ning physical laws and which, by de ini ion, accoun o spa ial
o ganiza ion o ele an hyd ologic and biogeochemical pa ame e s. Al hough ully dis ibu ed
app oaches ha e been hea ily c i icized because o he di icul ies in adequa ely de ining p ocess
equa ions and a unique, p oblem-speci ic pa ame e iza ion ( he “equi inali y p oblem” p esen ed in
Be en, (1989) and Be en, (1993)), hey o e lexible and ex ensi e possibili ies o es ce ain
hypo heses ( he “ i ual expe imen ” concep p esen ed in Weile and McDonnell (2004)), which a e
ela ed o he na u e o in e ac ions be ween hyd ology and biogeochemis y in we land sys ems.
These app oaches can be used o pa ially elucida e he black box and in es iga e he a e o hose
elemen s and solu es, which a e a ec ed by physical anspo and biogeochemical ans o ma ion in
we land ecosys ems. This hesis con ibu es o his line o wo k.
[14]
[15]
RESEARCH OBJECTIVES AND HYPOTHESES
2 Resea ch Objec i es and Hypo heses
This hesis aims a in es iga ing undamen al in e ac ions be ween hyd ology and biogeochemis y in
we land ecosys ems wi h he pu pose o gain a be e unde s anding o how nu ien cycling, in e nal
biogeochemical p ocess dis ibu ions, solu e mobiliza ion and solu e expo a e a ec ed by such
in e ac ions. A majo ocus o his hesis is o es ablish an in e disciplina y modeling amewo k
whe e hyd ological and biogeochemical p ocesses a e add essed equally and whe e undamen al
in e ac ions and eedback mechanisms be ween a we land’s hyd ology and biogeochemis y can be
ep esen ed in a physically-based model. The i e s udies, which a e p esen ed as pa o his hesis,
use a combina ion o ield in es iga ions, i ual expe imen s and ca chmen scale nume ical
modeling o add ess he di e en esea ch objec i es and hypo heses.
S udy 1 ocuses on he e ec s o su ace mic o- opog aphy on hyd ological p ocess dynamics and
in e ac ions ha go e n su ace-subsu ace exchange and uno gene a ion in ipa ian we lands.
Speci ically, s udy 1 uses a i ual expe imen app oach o in es iga e: (1) he ole o a hummocky
opog aphy o we lands on s eam discha ge gene a ion; (2) he e ec o mic o- opog aphy on
ypically-obse ed non-linea ela ionships be ween discha ge and wa e able dep h and (3) he
connec ion be ween su ace low gene a ion and clima ic and hyd ological bounda y condi ions. In
s udy 2, he p e iously p esen ed i ual we land model, is subsequen ly used o de elop a coupled
hyd ological/biogeochemical model which is being used in ano he i ual expe imen o in es iga e
how subsu ace low pa e ns, induced by mic o- opog aphy, a ec hyd ological anspo and
biogeochemical ans o ma ion p ocesses o edox-sensi i e solu es wi hin we lands. The main
esea ch hypo hesis o s udy 2 is o explo e whe he a complex, h ee-dimensional subsu ace low
ield, as a esul o mic o- opog aphy con olled su ace/subsu ace low exchange, c ea es
biogeochemical condi ions ha acili a e he o ma ion o local p ocess ho spo s o we land- ypical
edox eac ions, e en in soils wi h uni o m soil p ope ies.
Rep esen ing small-scale in e ac ions be ween hyd ology and biogeochemis y o we land
ecosys ems, as p esen ed in s udy 2, in a coupled physically-based modeling app oach has p o en o
be compu a ionally e y demanding, esul ing in low compu a ional e iciencies and ex emely long
simula ion imes. The main objec i e o s udy 3 he e o e is o de elop a echnique how e ec s o
mic o- opog aphy on sub-su ace low pa e ns, uno gene a ion and biogeochemical p ocess
pa e ns can be ep esen ed mo e e icien ly in physically-based models. Once es ablished, such an
al e na i e ep esen a ion can be used o accoun o e ec s o mic o- opog aphy in la ge scale
models like in wa e shed o egional models. S udy 4 is mainly based on da a om a ield campaign
on DOC expo o a small o es ed wa e shed wi h ipa ian we lands. The impac s o sho e m
luc ua ions in hyd ological and me eo ological bounda y condi ions on DOC a ia ions in uno a e
in es iga ed. He e, he main esea ch objec i es a e (1) o iden i y he spa ial o igin o DOC in uno ,
[16]
RESEARCH OBJECTIVES AND HYPOTHESES
(2) o iden i y hyd ological low pa hs which a e impo an o DOC mobiliza ion and (3) o
in es iga e implica ions o sho e m a ia ions o DOC in uno o he calcula ion o annual DOC
expo a es.
Runo gene a ion mechanisms a he ca chmen scale a e in es iga ed in s udy 5, whe e a “Hyd aulic
Mixing-Cell” me hology (HMC) is used o ack o e land and s eam uno gene a ion mechanisms
o a ain a meaning ul sepa a ion o s eam low hyd og aph o he Lehs enbach. Objec i es o s udy 5
a e (1) o es whe he he HMC me hod, de eloped and p esen ed ea lie by Pa ing on e al. (2011),
can p incipally be used in gene al o iden i y and quan i y ele an uno gene a ion mechanisms in
complex nume ical low models and mo e speci ically (2) o in es iga e he spa ial o igin and ela i e
con ibu ion o di e en uno componen s in he Lehs enbach ca chmen .
[17]
MATERIALS AND METHODS
3 Ma e ials and Me hods
3.1 S udy Si e
Field expe imen s and nume ical modeling we e ca ied ou in he Lehs enbach ca chmen . The
ca chmen is loca ed close o he ci y o Weisens ad in no h eas e n Ba a ia, Ge many (50°08’38’’N,
11°51’41’’E). Ele a ions o he si e a y be ween 877m abo e sea le el o upslope a eas and 690m
abo e sea le el o he ou le o he ca chmen . Mean annual p ecipi a ion, o he 4.2 km² la ge
Lehs enbach ca chmen , is a ound 1150 mm wi h a mean empe a u e o ~5°C (Ge s be ge , 2001).
The main egional aqui e o he Lehs enbach ca chmen (a ound 40 m hick) is made up o egoli hic
ma e ial o igina ing om wea he ing o he g ani ic bed ock (Lischeid e al., 2002). Hyd ologically,
he ca chmen can be sepa a ed in o wo dis inc uni s as illus a ed in Figu e 2: Nea ly one- hi d o
he o al a ea o he ca chmen can be classi ied as ipa ian we lands su ounding all majo s eams.
Pea o ming we lands ha e p edomina ely de eloped in he opog aphic dep essions owa ds he
cen e o he bowl-shaped ca chmen , whe e con e ging g oundwa e low (Figu e 2) a o s
condi ions ha lead o he accumula ion o pea . Fo he main we lands, a e age pea hickness a ies
be ween 0.3m and 1.2m. The we lands a e locally sepa a ed om he deepe g oundwa e sys em by a
basal clay laye o a iable ex en . Annual luc ua ions o g oundwa e le els in he we land’s main
zones a e limi ed o he uppe 0.2 m, bu may inc ease down o 0.8m below soil su ace du ing e y
ex ended d ough pe iods. Wa e con en o he a iably sa u a ed zone wi hin he we lands is
compa ably high, which a o s anoxic condi ions (abo e 80% wa e sa u a ion acco ding o Paul e al.
(2006), Es op-A agonés e al. (2012) and Es op-A agonés and Blodau (2012)). Ex ended a eas o he
we lands, especially in he lowe pa s o he ca chmen close o he ou le (Schlöppne b unnen II), a e
cha ac e ized by a p onounced mic o- opog aphy (Figu e 3); sequences o hollow and hummock
s uc u es, buil by he we land’s ypical ege a ion (Ca ex os a a, C. Canesccens, E iopho um
agina um, Na dus s ic a, Molinia coe uela, Ag os is sp., Sphagnum allax, B achy hecium i ula e
and A ichum undula um acco ding o Kno e al. (2008)). Such hummocky opog aphies a e quie
common in pea lands (Nungesse , 2003) and e idence om ch ono-s a ig aphic s udies indica es ha
such s uc u es (hummocks and hollows) may pe sis ela i ely unchanged o cen u ies o e en
millennia (Godwin and Conway, 1939; Conway, 1948; Tolonen, 1971; Ba be , 1981). P e ious
s udies pe o med in he Lehs enbach ca chmen indica ed ha impo an mechanisms and p ocesses
con olling s eam low gene a ion and solu e expo a e loca ed in he nea -s eam we land a eas
(Lischeid e al., 2002; Alewell e al., 2007; Lischeid, 2008).
A ound wo- hi ds o he a ea o he Lehs enbach ca chmen is co e ed by o es (mainly No way
Sp uce popula ions, (Ge s be ge , 2001)). Hyd ologic condi ions in he o es ed a eas, loca ed mainly
in he upslope a eas o he ca chmen (Figu e 2), clea ly di e om hose wi hin he ipa ian
we lands. Long e m g oundwa e obse a ions o he upslope a eas show pe manen ly deep

[18]
MATERIALS AND METHODS
g oundwa e le els, 5-10m below he land su ace and an ex ended unsa u a ed zone wi h compa ably
low wa e con en s. In con as o he wa e sa u a ed condi ions wi hin he we lands, he upslope a eas
can be classi ied as ae a ed o es soils. The o es ed a eas ep esen he main echa ge zones o he
deepe g oundwa e sys em, as e lec ed by downwa d hyd aulic g adien s in he unsa u a ed zone.
The e is no clea e idence o p onounced la e al lows abo e he g oundwa e able (in e low) in
hese a eas wi h deep wa e able.
Figu e 2: Concep ual model o he Lehs enbach ca chmen . The o e all hyd ology o he ca chmen is
con olled by he s uc u e o he basin. Da k g ey a eas ep esen o es ed zones and ligh g ey a eas
we lands, which occupy almos 1/3 o he 4.2 km² ca chmen a ea.
Figu e 3: Pic u e o he Schlöppne b unnen II ield si e (loca ed in he lowe pa o he ca chmen ,
close o he ca chmen ’s ou le ) aken du ing a s o m low e en in sp ing 2009. The
Schlöppne b unnen II si e is cha ac e ized by a p onounced mic o- opog aphy (hollow and hummock
s uc u es) and belongs o he co e we land a eas o he Lehs enbach ca chmen .
[19]
MATERIALS AND METHODS
3.2 Hyd ological Modeling
Hyd ological modeling as pa o his hesis was pe o med using a spa ially-explici , physically-based
modeling concep , whe e su ace and subsu ace hyd ology is ep esen ed using he code
Hyd oGeoSphe e (HGS, p esen ed in The ien e al. (2008)). HGS is a ully-in eg a ed ini e elemen
su ace-subsu ace low model. Va iably sa u a ed subsu ace low in po ous media is simula ed by
sol ing he Richa ds equa ion in h ee dimensions (3D):
󰇛󰇜Γ
 

󰇛󰇜Eq. 1

󰇛󰇜 Eq. 2
Whe e  [-] ep esen s he olume ic ac ion o he o al po osi y occupied by he p ima y
con inuum (po ous o ac u ed medium) and q [LT-1] he luid lux.  [L3 L-3T-1] ep esen s he
olume ic luid exchange be ween he subsu ace domain and all o he ypes o domains suppo ed
by he model (e.g. su ace domain). Fluid exchange wi h he ou side o he simula ion domain is
ep esen ed by Q [L3 L-3T-1], which is a olume ic lux pe uni olume ep esen ing sou ce (posi i e)
and sinks (nega i e). θs [-] and Sw [-] ep esen he sa u a ed wa e con en and he deg ee o sa u a ion
espec i ely. Fu he mo e, he luid lux q is gi en by Eq. 2 whe e  [-] ep esen s he ela i e
pe meabili y o he medium as a unc ion o he wa e sa u a ion Sw, Ksa [LT-1] is he sa u a ed
hyd aulic conduc i i y o he medium,  [L] is he p essu e head and z [L] he ele a ion. Fo
ep esen a ion o a iably sa u a ed low, commonly used unc ions inco po a ed in o HGS a e hose
p esen ed in Van Genuch en (1980b) and B ooks and Co ey (1964) o al e na i ely, soil e en ion
cha ac e is ics can also be handled h ough he use o abula da a inpu i ield measu emen s a e
a ailable (The ien e al., 2008). O e land- o s eam low in 2D is ep esen ed by he di usion wa e
app oxima ion o he dep h-a e aged dynamic wa e equa ions (The ien e al., 2008):
󰇛󰇜Γ
 Eq. 3
Wi hin he di usi e wa e equa ion, he e w i en in ec o ial no a ion, do [L] ep esen s he su ace
low wa e dep h; qo [LT-1] he wa e lux on he su ace; [T-1] he luid exchange a e wi h he
subsu ace; Qo [LT-1] he olume ic low a e pe uni a ea ep esen ing ex e nal sinks (nega i e) o
sou ces (posi i e);  [-] he su ace po osi y and ho [L] he wa e su ace ele a ion. Su ace–
subsu ace coupling is implemen ed using he conduc ance concep :
Γ
 󰇛󰇜Eq. 4
[20]
MATERIALS AND METHODS
The conduc ance concep assumes ha he exchange lux be ween he su ace and he subsu ace 
[T-1] depends on he g adien ac oss a coupling in e ace h-ho [L] (h [L] ep esen s he subsu ace
wa e head and ho [L] he wa e su ace ele a ion), he hickness o he in e ace  [L] (coupling
leng h), i s ela i e pe meabili y  [-] and he e ical sa u a ed hyd aulic conduc i i y  [LT-1]
(The ien e al., 2008). All go e ning equa ions o su ace- and subsu ace low a e sol ed
simul aneously ia a con ol olume, ini e-elemen app oach (The ien e al., 2008). HGS has been
applied o e a wide ange o spa ial scales anging om plo and i e each scales (Jones e al., 2006;
B ook ield e al., 2009) o e he scale o wa e sheds (Jones e al., 2008; Li e al., 2008) up o he scale
o con inen s (Lemieux e al., 2008a; Lemieux e al., 2008b; Lemieux e al., 2008c). As pa o his
hesis, HGS was used o simula e hyd ological low p ocesses and su ace/subsu ace low
in e ac ions on wo di e en scales: On he plo scale nume ical low modeling (using HGS) was used
o ep esen he highly dynamic low p ocesses wi hin he ipa ian we lands (s udy 1, 2, 3) o he
Lehs enbach ca chmen . An in eg a ed pe spec i e on hyd ological low p ocesses, ele an o he
ca chmen scale uno gene a ion and solu e expo s, was he mo i a ion o se ing up a nume ical
ca chmen scale low model o he Lehs enbach a ea (s udy 4+5).
3.2.1 Vi ual We land Modeling (S udy 1, 2 and 3)
The concep ual idea behind he plo scale modeling is simila o he i ual expe imen s p oposed by
Weile and McDonnell (2004). The objec i es o he s udies 1-3 a e add essed h ough i ual
modeling expe imen s. The nume ical model is used as a i ual we land, in which pe ec p ocess
knowledge is assumed (see e.g. Zehe e al. (2005)). Vi ual we land modeling in ol es mo e han only
one nume ical low model: S udy 1 and 2 use di e en model scena ios wi h di e en , geos a is ically
gene a ed 3D ealiza ions o he hummocky mic o- opog aphy. S udy 3 in ol es geos a is ically
de i ed, 2D ep esen a ions o mic o- opog aphy, which we e used in subsequen model scena ios. All
nume ical low models (s udy 1-3) as pa o he i ual we land modeling app oach we e se up o
he same spa ial model domain (se up o a 10m x 20m x 2m plo ) ep esen ing a syn he ic sec ion o
a ipa ian we land d aining in o a nea by s eam segmen (Figu e 4). Vi ual we land modeling is
desc ibed in de ail in he me hod sec ion o s udy 1 and only a b ie summa y abou he applied
echniques and me hods is gi en in his sec ion.
[21]
MATERIALS AND METHODS
Figu e 4: Geome y o he i ual we land segmen : a) plana e e ence model showing he main
d ainage di ec ion and channel loca ion; b) smoo hed ealiza ion o he we lands hummocky mic o-
opog aphy; c) c oss sec ion (Y=5m) o he mic o- opog aphy model.
Rep esen a ion o Mic o- opog aphy
The spa ial s uc u e o he mic o- opog aphy o a ypical we land in he Lehs enbach ca chmen was
ep esen ed using geos a is ical indica o simula ions based on Ma ko Chain models o ansi ion
p obabili ies (TPROGS-T ansi ion PRObabili y Geos a is ical So wa e p esen ed in Ca le and Fogg
(1996)). The me hod was o iginally de eloped o ealis ically ep esen aqui e he e ogenei y wi h
disc e e ansi ions be ween di e en hyd o acies (Ca le and Fogg, 1996). TPROGS has been widely
applied o g oundwa e low and anspo p oblems (e.g. Weissmann, 1999; Fleckens ein e al.,
2006; Lee e al., 2007; F ei e al., 2009). Fo a ealis ic ep esen a ion o mic o- opog aphy, he
geos a is ical model was condi ioned wi h ield da a de i ed om se e al su eyed ansec s aken
wi hin a 30m x 30m plo o he Schlöppne b unnen II si e loca ed in he Lehs enbach ca chmen . The
ou pu o he indica o simula ions was ans e ed in o an a i icial digi al ele a ion model (DEM) by
assigning he di e en indica o s o ce ain ele a ion classes. The esul ing DEM mimics he spa ial
s uc u es o he we lands mic o- opog aphy. The applica ion o geos a is ical simula ions p o ided
he possibili y o wo k wi h mul iple ealiza ions o mic o- opog aphy based on ei he he same o
di e en s uc u al p ope ies. A de ailed desc ip ion o he used geos a is ical app oach is gi en in he
me hods chap e o s udy 1. S udy 1 and 2 use model scena ios whe e mic o- opog aphy is ac ually
[28]
MATERIALS AND METHODS
was used o ep esen he whole 3D domain o he i ual we land model, which esul ed in
~1.450.000 di e en PHREEQC sub-sec ion simula ions pe low model.
Figu e 7: Concep o he applied s eam ube app oach o ep esen a ion o biogeochemis y along
isola ed subsu ace low pa hs (dashed line). An isola ed low pa h is spli in o n di e en sub-sec ions.
Each sub-sec ion i ep esen s a small each o he low pa h, o which he biogeochemical e olu ion,
depending on he hyd ological/biogeochemical bounda y condi ions, is simula ed using PHREEQC
(Pa khu s , 1995). Bounda y and ini ial condi ions a e indi idually assigned o each PHREEQC sub-
sec ion simula ion. Be ween consecu i e sub-sec ions, edox-sensi i e solu es a e exchanged we e
he i h sub-sec ion uses he inal edox chemical composi ion o he i-1 h sub-sec ion as ini ial
condi ion. X, Y and Z ep esen he spa ial coo dina es a he beginning and he end o a sub-sec ion;
Δ ep esen s he sub-sec ion’s esidence ime.
3.3.2 Implemen ed Reac ion and Bounda y Condi ions
The biogeochemical model ep esen s we land- ypical, edox-sensi i e p ocesses, which a e
implemen ed using di e en kine ic eac ions. In pa icula , he ollowing edox-sensi i e p ocesses
a e being simula ed: ae obic espi a ion, deni i ica ion, i on(III) educ ion, sul a e educ ion, i on(II)
oxida ion, ammonium oxida ion, ae obic and anae obic sul ide oxida ion. Kine ics o all educ ion
p ocesses (ae obic espi a ion, deni i ica ion, i on(III) educ ion, sul a e educ ion) whe e
mic oo ganisms use di e en elec on accep o s (oxygen, ni a e, i on(III) and sul a e) o u no e o
o ganic ma e ial a e o mula ed based on Monod kine ics (Monod, 1949). Fo eac ions ollowing
Monod kine ics, as shown in Eq. 5, he kine ic a e Rk [ML-3T-1] is calcula ed as a unc ion o he
solu es concen a ion ck [ML-3] and he eac ion speci ic cons an s μmax [ML-3T-1] and Ks,k [ML-3].

[29]
MATERIALS AND METHODS

  
, 
Eq. 5
In he model, Monod kine ic cons an s o he di e en educ ion p ocesses a e based on labo a o y
s udies o biodeg ada ion o o ganic chemicals ( e e ences a e lis ed in Table 2 o s udy 2) and we e
adjus ed as pa o he calib a ion p ocess. Finally, calib a ed coe icien s a e lis ed in Table 2 o s udy
2. Oxida ion p ocesses (i on(II) oxida ion, ammonium oxida ion, anae obic and ae obic sul ide
oxida ion) we e o mula ed using highe o de eac ion kine ics as lis ed in Table 2 o s udy 2. In
edox con olled sys ems like we lands, educ ion p ocesses occu sequen ially whe e mic oo ganisms
use oxygen as p ima y elec on accep o i s , be o e ni a e, i on(III) and sul a e a e being used. To
ep esen his sequen ial beha io wi hin he biogeochemical model, di e en condi ions we e
o mula ed o which he di e en educ ion p ocesses a e being ini ia ed. In he app oach p esen ed
he e, hese condi ions a e ep esen ed by c i ical concen a ions o edox-sensi i e solu es which
con ol whe he a edox p ocess is ini ia ed o no . Fo he di e en educ ion p ocesses, con olling
c i ical concen a ions a e lis ed in Table 1. Table 1 mus be ed ow-wise, whe e en ies “>0” mean
ha he co esponding edox-sensi i e eac an (column) mus be a ailable and “-“ means ha his
p ocess does no depend on he p esence o he edox-sensi i e compound. Fo example i on(III)
educ ion in he biogeochemical simula ion is only ini ia ed i : (1) Dissol ed oxygen concen a ions
all below ; (2) Mos o he ni a e is al eady deple ed and ac ual concen a ions all below

; and (3) The elec on accep o i on(III) is a ailable.
Table 1: C i ical concen a ions which a e con olling he sequen ial ini ializa ion o he edox
sequence. Values we e de i ed om ield obse a ions. Table mus be ead ow-wise (e.g.
deni i ica ion is ini ia ed i 1) oxygen con en s d op below Cc i o oxygen and 2) i ni a e is p esen ).
= 5.0 x 10-6 mol/L; 
= 4.0 x 10-7 mol/L; = 5.0 x 10-6 mol/L.
The c i ical concen a ions we e o mula ed based on e alua ion o dep h p o iles o edox-sensi i e
solu es which we e aken a he Schlöpne b unnen II si e in he Lehs enbach ca chmen (Kno and
Blodau, 2009; Kno e al., 2009). In e als o he ac i a ion o edox p ocesses a e o e lapping,
meaning ha mul iple p ocesses can occu simul aneously which can be app o ed unde labo a o y as
well as unde ield condi ions (Kno and Blodau, 2009; Kno e al., 2009).
A ailabili y o oxygen can be seen as a key componen , con olling he p ocess composi ion wi hin
we land ecosys ems. P ocesses like ae obic espi a ion o ni i ica ion only occu i oxygen is
oxygen ni a e i on(III) Sul a e
ae obic espi a ion >0 - - -
deni i ica ion 
 >0 - -
i on(III) educ ion 
 

 >0 -
sul a e educ ion 
 

 
 >0
[30]
MATERIALS AND METHODS
a ailable. O he p ocesses, like deni i ica ion i on(III)- o sul a e- educ ion a e only ini ia ed unde
anoxic condi ions whe e oxygen concen a ions a e e y low. Along a subsu ace low pa h,
a ailabili y o oxygen a ies as he hyd ological bounda y condi ions change. Wi hin he unsa u a ed
zone, deple ed oxygen is being eplaced by di usion o a mosphe ic oxygen and a ailabili y o
oxygen o mic obial ca alyzed eac ions is high. In he sa u a ed zone dissol ed oxygen
concen a ions a e low because he esupply by di usion is being inhibi ed by po e wa e , which ac s
as an e ec i e di usion ba ie . The e o e, in he biogeochemical model oxygen a ailabili y was used
as a key a iable ha ei he igge s o supp esses edox-sensi i e p ocesses. Along a sub-su ace low
pa h, a ailabili y o oxygen was coupled o he ansien p essu e heads which we e a ailable as pa
o he i ual we land modeling. Fo each PHREEQC sub-sec ion simula ion o a sub-su ace low
pa h, he co esponding p essu e head was es ima ed o he s a loca ion o he sub-sec ion. P essu e
heads we e ela ed o a ce ain oxygen concen a ion acco ding o Figu e 8. I he p essu e head o he
sub-sec ion is loca ed wi hin zone 1 (unsa u a ed zone wi h nega i e p essu e heads), he oxygen
a ailabili y is a a maximum due o he uninhibi ed di usion o a mosphe ic oxygen. Wi hin zone 2
(sa u a ed zone wi h posi i e p essu e heads), oxygen con en s a e dec easing wi h inc easing p essu e
heads ep esen ing inc easing inhibi ion o oxygen di usion wi h dep h. Oxygen concen a ions in
sub-sec ion simula ions ha a e loca ed ei he wi hin zone 1 o 2 we e se o a cons an alue
e lec ing ha apid esupply o oxygen p e en s i s deple ion by oxygen consuming p ocesses. Fo
sub-sec ions ha a e loca ed wi hin zone 3 (deepe sa u a ed zone wi h p essu e heads abo e 0.25 m)
oxygen is no assigned as a cons an bounda y condi ion. Ins ead, oxygen is se as an ini ial condi ion
whe e he esidual oxygen con en s o he p eceding sub-sec ion a e used as ini ializa ion. Wi hin zone
3, whe e a mosphe ic di usion is dis up ed, oxygen can be o ally deple ed due o oxygen consuming
p ocesses. The ela ionship shown in Figu e 8 was de i ed om obse ed oxygen-dep h p o iles aken
a he Schlöppne b unnen II si e in he Lehs enbach ca chmen (Kno e al., 2009). Aside om an
adequa e elec on accep o (e.g. oxygen, ni a e, i on(III) o sul a e), mic obial ca alyzed educ ion
p ocesses equi e a ca bon sou ce ha is a ailable o mic oo ganisms. Fo he ca bon ich sys ems
s udied he e unlimi ed a ailabili y o ca bon was assumed.
[31]
MATERIALS AND METHODS
Figu e 8: Typical oxygen dep h p o ile obse ed o a we land si e o he Lehs enbach ca chmen .
P o ile was used o assign oxygen bounda y condi ions o he di e en PHREEQC sub-sec ion
simula ions based on ansien model ou pu o he i ual we land model.
[32]
[33]
RESULTS AND DISCUSSION
4 Resul s and Discussion
4.1 E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno
gene a ion in a i ual ipa ian we land (S udy 1)
Resul s om he i ual we land modeling indica e ha hyd ological dynamics and uno gene a ion
p ocesses wi hin he ipa ian we land a e signi ican ly a ec ed by he we land’s hummocky
opog aphy. Su ace and subsu ace uno gene a ion a e in luenced by dis inc shi s be ween
su ace and sub-su ace low dominance esul ing om he in e play be ween ain all-induced
luc ua ions o he shallow wa e able and he su ace mic o- opog aphy. Su ace lows a e
cha ac e ized by a ill and spill mechanism, simila o wha has been desc ibed o shallow subsu ace
d ainage o hillslopes (Hopp and McDonnell, 2009). He e, su ace dep essions (hollows) a e illed
wi h wa e as soon as he g oundwa e le el in e sec s wi h he land su ace (e.g. du ing in ensi e
ains o m e en s).Wi h inc easing ain all in ensi y ponded dep essions s a o in e connec , o ming
dis inc su ace low ne wo ks which de elop independen ly in space and ime (as shown in Figu e 9).
These ne wo ks can apidly d ain la ge a eas o he we lands and a imes (du ing e y in ensi e
ains o m e en s) con ibu e up o 80% o he o al discha ge ha is gene a ed om we lands.
Figu e 9: Six consecu i e snapsho s o he e ol ing su ace low ne wo ks du ing he la ges low
e en o he yea (day 217 o day 218). The ed lines sepa a e di e en low ne wo ks (1-3) ha
de eloped independen ly om each o he .
Howe e , whe he such su ace low ne wo ks de elop in space and ime and whe he su ace uno
is gene a ed in he we lands depends on he his o y o he sys em. Fo ains o ms occu ing a e
ex ended d y pe iods in summe , su ace low ne wo ks may no be gene a ed because g oundwa e
le els in he we land a e oo a below he land su ace o gene a e su ace ponding. On he con a y, a

[34]
RESULTS AND DISCUSSION
ains o m o he same o e en lesse in ensi y may cause he gene a ion o signi ican su ace uno i
i occu s wi h we p econdi ions. The simula ed uno dynamics can also explain obse ed non-linea
and hys e e ic ela ionships be ween he ipa ian g oundwa e le el in he we lands and discha ge
being gene a ed om i (Figu e 10). The dynamic uno gene a ion mechanism, which is con olled
by mic o- opog aphy whe e he sys em apidly shi s be ween su ace and subsu ace low
dominance, was iden i ied as a main d i e o he obse ed non-linea dynamics. Simila non-linea
ela ionships be ween wa e able and discha ge ha e been epo ed o we lands and ipa ian zones in
o he pa s o he wo ld (e.g. Fi zge ald e al. (2003)).
Unde s anding he mechanisms ha go e n hyd ologic low pa hs and s eam low gene a ion in
ipa ian zones is impo an , because nu ien ans o ma ion and expo a e in eg ally ela ed o he
hyd ological dynamics (Gillham, 1984; De i o and Hill, 1997; Vidon and Hill, 2004; Lischeid e al.,
2007). Al hough mobiliza ion o solu es has no been explici ly simula ed in s udy 1, he mic o-
opog aphic con olled uno gene a ion can ha e signi ican implica ions o he expo o solu es
(e.g. DOC, ni a e o sul a e) om he we lands. Fas low componen s like apid su ace d ainage due
o he ex ensi e su ace low ne wo ks o shallow subsu ace low ha e he po en ial o quickly
(wi hin minu es o hou s) mobilize solu es om he uppe mos laye s (10 o 20 cm) o he we lands.
Field obse a ions (Kno and Blodau, 2009; Kno e al., 2009) o he Lehs enbach ca chmen ha e
shown ha hese supe icial laye s, which a e ypically unsa u a ed, a e ich in oxic species ha
accumula e du ing d ie pe iods such as ni a e o sul a e. Du ing ains o ms, which igge gene a ion
o apid su ace and shallow subsu ace d ainage, hese species can be lushed om he sys em. Along
hese e y as low pa hways ni a e and/o sul a e a e no being educed because deepe , anoxic
laye s a e being bypassed by he supe icial uno componen s. In i s e ec on he mobiliza ion o
edox-sensi i e solu es, his mechanism ope a es he same way as o he bypassing p ocesses ha ha e
been desc ibed o he Lehs enbach ca chmen (Lischeid e al., 2007) and o o he compa able
ecosys ems (Cu is e al., 2011). Simila dynamics apply o he mobiliza ion o DOC because i s
concen a ions a e also highes in he uppe mos laye s whe e esh o ganic ma e ial is a ailable and
he pea is less decomposed han in deepe laye s (Clemens, 2011).
The mechanis ic unde s anding on how uno is being gene a ed on he small scale in he we land
a eas and how he di e en low componen s wi h hei indi idual esponse and esidence imes
con ibu e o s eam low gene a ion is c ucial o iden i y which low pa hways a e impo an o
solu e mobiliza ion. Findings om s udy 1 we e subsequen ly used o de elop a ca chmen -scale
concep ional model o DOC mobiliza ion p esen ed as pa o s udy 4. Mo eo e , indings ha
su ace low gene a ion in he we land a eas is s ongly in luenced by mic o- opog aphy a e impo an
o simula e he ca chmen -scale hyd ological dynamics (s udy 5), because a he han as shee low,
su ace low in he ca chmen is gene a ed in disc e e su ace low ne wo ks in a h eshold-con olled
p ocess, which mus be accoun ed o in la ge scale models (s udy 5). This was done by applying he
ill s o age concep de eloped as pa o s udy 3.
[35]
RESULTS AND DISCUSSION
Figu e 10: a) Rela ionship be ween discha ge and g oundwa e le el o wo peak low e en s,
obse ed o a small ca chmen loca ed in B i ish Colombia, Canada (modi ied a e Fi zge ald e al.
(2003)).b) Simula ed ela ionship be ween g oundwa e le el and channel discha ge o he mic o-
opog aphy model. Blue illed ci cles ep esen imes when no su ace d ainage occu s, ed open
ci cles ep esen condi ions when su ace d ainage is being gene a ed; di e en scales a e used on
he x-axis o be e isibili y o hys e e ic beha io du ing low discha ges; he sequence o days 217 o
219, ep esen ing an in ense ain s o m, is depic ed by a line.
[36]
RESULTS AND DISCUSSION
4.2 Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in
we land sys ems – a i ual modeling expe imen . (S udy 2)
Resul s om pa icle acking show ha supe icial mic o- opog aphical s uc u es o he we land
cause a complex subsu ace low ield wi h shallow and deepe low cells ha anspo wa e and
solu es ac oss he model domain (Figu e 11 A). The spa ial dis ibu ion o high poin s (hummoks) and
dep essions (hollows) esul s in small-scale pa e ns o in- and ex il a ion. Hummocks gene ally
ep esen a eas o p e e en ial in il a ion and hollows zones o p e e en ial ex il a ion (Figu e 11 A).
The coexis ing deep and shallow low sys em shows dis inc ly di e en low eloci ies and subsu ace
esidence imes (Figu e 11 B). The esul ing complex edis ibu ion o wa e in he subsu ace and
esidence imes, anging om a ew days o yea s, ha e signi ican e ec s on biogeochemical p ocess
pa e ns and he spa ial dis ibu ion o edox-sensi i e compounds in he we lands. Biogeochemical
simula ions show he o ma ion o local ho spo s o edox p ocesses wi hin he we lands. They a e
he esul o he complex subsu ace low pa hs and he anspo -limi ed a ailabili y o elec on
accep o s and dono s. Ho spo s o educ ion o edox-sensi i e species (e.g. deni i ica ion, i on(III)-
and sul a e educ ion) a e p e e en ially gene a ed below local hummocks (Figu e 11 C), whe eas
oxida ion ho spo s o m in zones o upwelling wa e below hollows whe e olde , educed
g oundwa e ge s in con ac wi h a mosphe ic oxygen (Figu e 11 D).
Findings om s udy 2 mechanis ically p o e he exis ence o localized zones o highe eac i i y (ho
spo s) whe e mos o he biogeochemical u no e is accomplished wi hin we land sys em. This has
been obse ed be o e in a ious ield s udies (e.g. Jacks and No s öm, 2004; Paul e al., 2006;
Kno , 2009). Typically, he gene a ion o such ho spo s has been explained by he he e ogeneous
dis ibu ion o s a ic, physical-chemical p ope ies o he soil (Ree e e al., 2001; Holden and Bu ,
2003) o labile ca bon inpu in he hizosphe e (C ow and Wiede , 2005). Howe e , esul s om he
biogeochemical simula ions in his s udy demons a e ha he occu ence o eac i i y ho spo s does
no need o be associa ed wi h s a ic physical-chemical soil he e ogenei ies a p io i. Resul s ha e
shown ha ho spo s could heo e ically de elop e en in homogenous pea soils due o a highly
dynamic low sys em wi h (1) complex su ace/subsu ace low in e ac ions, whe e su ace mic o-
opog aphy induces a subsu ace low ield ha de ines a small-scale zona ion o in- and ex il a ion
a eas and (2) a hyd ological con ol o he biogeochemical bounda y condi ions ha ei he acili a ed
o supp essed edox p ocesses in ex- and in il a ion a eas.
These esul s p esen a new pe spec i e on biogeochemical ans o ma ion p ocesses in ipa ian
we lands, which p o ides a dynamic amewo k o explain p ocess he e ogenei y in we land soils and
a iabili y in p ocess a es o e space and ime. Fo ma ion o biogeochemical ho spo s as a esul o
he mechanisms p esen ed in his s udy may u he mo e explain how ma e ial he e ogenei y is being
gene a ed wi hin he subsu ace. Biogeochemical ho spo s may ha e he po en ial o al e he
hyd odynamic p ope ies o he pea o we land soils. The p ecipi a ion o i on oxides o example,
[37]
RESULTS AND DISCUSSION
which p e e en ially occu s a oxida ion ho spo s, can lead o a educ ion o he e ec i e po osi y and
a lowe hyd aulic conduc i i y, p o iding a nega i e eedback on oxygen pene a ion.
Fu u e wo k will ha e o add ess unde which clima ic condi ions he simula ed biogeochemical ho
spo s a e s able, because shi s in clima ic o cing due o clima e change will p obably a ec he in
s udy 1 simula ed su ace/subsu ace low in e ac ions as well as he sub-su ace low ield. This will
in u n a ec he oxygen a ailabili y and he biogeochemical p ocess dis ibu ions wi hin he
we lands. Du ing ex ended d ough pe iods o example, which a e p edic ed by clima e models o
he empe a e zones (McCa y e al., 2001), biogeochemical ho spo s a e likely o anish as he
sys em g adually shi s owa ds a mo e homogenous p ocess dis ibu ions. He e, he d opping
g oundwa e may be esponsible o he e e sal o he hyd aulic g adien s unde dep essions,
swi ching om upwelling o in il a ing condi ions. In u n oxida ion ho spo s will diminish because
esupply o educed species om upwelling g oundwa e is dis up ed.
The e ec o he biogeochemical p ocess pa chiness on solu e expo s (e.g. ni a e o sul a e) ou o
he we land a eas has also o be in es iga ed u he . Because o model limi a ions i was so a no
possible o link he in e nal biogeochemical p ocess dis ibu ions o he uno gene a ion mechanisms
p esen ed in s udy 1 in o de o explici ly simula e solu e expo s unde condi ions o ho spo
o ma ion. Such an in eg a ed simula ion would also help o u he imp o e he in s udy 4 p esen ed
concep ual model on ca chmen -scale solu e mobiliza ion.
[44]
RESULTS AND DISCUSSION
4.5 In e p e ing low gene a ion mechanisms om in eg a ed su ace wa e -
g oundwa e low models o a ipa ian we land and ca chmen (S udy 5).
The Hyd aulic Mixing-Cell (HMC) me hodology (Pa ing on e al., 2011 and Pa ing on e al., 2012)
has p o en o be a use ul ool o assessmen o ca chmen unc ioning and sepa a ion o low
hyd og aphs. Applied o he ca chmen scale model o he Lehs enbach, he HMC me hod elucida ed
he complexi y in he spa io empo al dis ibu ion o he di e en uno gene a ion mechanisms. The
di e en low componen s which we e iden i ied o domina e uno gene a ion o he Lehs enbach
ca chmen a e, (1) g oundwa e discha ge o he s eam ne wo k (GW-CH), (2) di ec ain all en e ing
he s eams (RF-CH) and (3) s eam inpu s due o sa u a ed o e land low om he we land a eas.
O e land low om he ipa ian we lands was u he sub-di ided in o a su ace low ac ion
o igina ing om g oundwa e ex il a ion (GW-WL) and o e land low gene a ed om ain all alling
on o en i ely wa e sa u a ed a eas o he we lands (RF-WL). Rela i e con ibu ions o he i e
di e en uno gene a ion mechanisms a e acked in ime and space by he HMC ou ine. The HMC
ou ine was applied o a la ge s o m e en (13 h- 21s July, 2001) as well as o he en i e 2001
hyd ological yea (11/01/2000 – 10/31/2001). Resul s o he s o m e en a e shown in Figu e 14. The
GW-CH componen (panel A) domina es uno gene a ion o e la ge a eas o he s eam ne wo k
p io o he s o m e en du ing low low condi ions. A he peak o he s o m, GW-CH gene a ion is
o mino impo ance as o he gene a ion mechanisms a e ac i a ed (RF-CH, GW-WL and RF-WL in
panel B, C and D, espec i ely). Mos o e land low ha con ibu s o s eam discha ge du ing he
s o m e en is gene a ed due o ain all, which is di ec ly alling on o he ully wa e sa u a ed we land
a eas as indica ed by he high ela i e ac ion o he RF-WL componen in panel B. On an annual
basis, o al s eam wa e lea ing he ca chmen a he ou le (shown in Figu e 5), acco ding o he
HMC analysis, consis s o 67.9% o wa e o igina ing om g oundwa e inpu s (GW-CH), 12.6% o
di ec ain all o he s eam ne wo k and 19.5% o sa u a ed o e land low om he we land a eas
(GW-WL + RF-WL). Howe e , o e land low was iden i ied o be only ele an du ing e y in ensi e
ains o m e en s and is only gene a ed in signi ican p opo ions in he a eas o we land ha a e close
o he ca chmen ou le . Acco ding o he HMC analysis, no o e land low is gene a ed in he o es ed
a eas because ain all quickly in il a es he e and echa ges he unde lying egoli hic aqui e .

[45]
RESULTS AND DISCUSSION
Figu e 14: Calcula ed s eam and o e land low gene a ion, es ima ed by applying he “hyd aulic
mixing-cell” me hodology o he Lehs enbach ca chmen model. The low gene a ion componen s
acked a e: a) g oundwa e discha ge o he channel (GW-CH), b) ain all o he channel (RF-CH), c)
g oundwa e discha ge o he we lands (GW-WL), and d) ain all o he we lands (RF-WL). Rela i e
con ibu ions (colo ed scales anging om 0 o 1) we e acked o a ypical s o m low e en .
[46]
RESULTS AND DISCUSSION
Howe e , he esponse imes o subsu ace low en e ing he s eam ne wo k (GW-CH) o ain all
seems o be e y as . This can be explained by he “p essu e wa e p opaga ion” mechanism
(Ge mann e al., 1990; Lischeid, 2008), whe e inc easing hyd aulic p essu e in he upslope a eas
mobilizes g oundwa e u he downslope (e.g. in he cen e o he bowl-shape ca chmen ).
In e es ingly, su ace low componen s om he we land a eas (GW-WL + RF-WL), which a e
usually ega ded as e y as low componen s, show a clea ly delayed esponse o ain all inpu s.
This can pa ly be explained by he h eshold-con olled “ ill and spill” su ace low gene a ion
mechanisms desc ibed in s udy 1 whe e small scale dep essions i s ha e o be illed wi h wa e
be o e any su ace low owa ds he s eam is gene a ed. The h eshold-con olled su ace low
gene a ion caused by he mic o- opog aphy has been accoun ed o in he ca chmen -scale model by
applying he ill-s o age heigh concep p esen ed in s udy 3.
Whe he simula ed luxes om he indi idual uno gene a ion p ocesses as es ima ed by he HMC
analysis ac ually ma ch wi h he mo e complex eali y is ques ionable due o simpli ica ions in he
model. Fo example, di ec ain all inpu s o he s eam ne wo k a e p esumably negligible as s eams
co e only a mino ac ion o he ca chmen a ea (Lischeid, 2008). In he ca chmen scale model,
howe e , s eam segmen s occupy compa a i ely la ge ac ions o he o al a ea because he
esolu ion o he nume ical mesh was oo coa se o adequa ely ep esen he na ow s eam channels.
This explains he la ge ac ions o he RF-CH componen (12.6% o o al discha ge pe yea ) in he
simula ed discha ge. Ne e heless, he HMC me hod in combina ion wi h nume ical modeling
p o ides a aluable ool o assess whe he o no a ca chmen model beha es in he expec ed way o ,
mo e impo an ly, he way he ca chmen p ocesses a e concep ualized. In ha sense i is a p omising
and use ul ool o a “so calib a ion” based on unde s anding o ca chmen unc ioning om eal
obse a ions.
A u u e applica ion o he HMC me hod in combina ion wi h he ca chmen -scale model o he
Lehs enbach a ea could be o sepa a e uno componen s o igina ed om o es ed si es and we land
a eas. Wa e o igina ing om he we lands and o es ed si es ha e e y di e en chemical signa u es
(e.g. oxygen sa u a ion, edox s a es o DOC loadings) which a e being mixed wi hin he s eam o he
hypo heic zone. Applying he HMC analysis o ack how much wa e in s eam uno is o igina ed
om he we lands and o es ed si es would imp o e ou unde s anding on he ela i e con ibu ions o
di e en low pa hs o s eam discha ge and solu e expo s.
[47]
CONCLUSIONS AND OUTLOOK
5 Conclusions and Ou look
Findings om his hesis ha e shown ha he combina ion o ield in es iga ions, i ual expe imen s
and ca chmen scale nume ical modeling has p o en o be a e y use ul combina ion o in es iga e
and explo e scale-dependen uno gene a ion p ocesses and couplings be ween hyd ology and
biogeochemis y. On he ca chmen scale, couplings be ween hyd ology and biogeochemis y we e
iden i ied o be e y impo an o he mobiliza ion o DOC. Flow componen s, ele an o he
gene a ion o uno in he Lehs enbach ca chmen , con ibu e di e en ly o he mobiliza ion o DOC.
Deep g oundwa e low o igina ed om he o es ed upslope a eas was iden i ied o be gene ally low
in DOC, mainly because pe cola ing wa e o o es ed si es is being e icien ly deple ed in DOC due
o so p ion and biogeochemical decomposi ion p ocesses. Fas low componen s like su ace low o
in e low, which would ha e he abili y o bypass soil laye s whe e so p ion and decomposi ion occu ,
could no be e i ied o he o es ed a eas, nei he in ield in es iga ions no in nume ical
simula ions.
Field in es iga ions and nume ical modeling indica e ha he po en ial o DOC mobiliza ion is
highes o low componen s loca ed wi hin he ipa ian we lands. Mobiliza ion o DOC wi hin he
ipa ian we lands is con olled by he in e play o (1) he ansmissi i y eedback mechanism
con olling he dep h dependen dynamics and imescales o subsu ace anspo , (2) a h eshold-
con olled su ace low gene a ion whe e, episodically, la ge amoun s o su ace wa e a e apidly
being mobilized in ex ended su ace low ne wo ks and (3) he dep h dependen a ailabili y o DOC
caused by he la e al a ia ion o DOC p oduc ion and he non-uni o m biogeochemical
ans o ma ion and deg ada ion p ocesses. Episodically, he ac i a ion o as low componen s in
shallow laye s and/o on he su ace is esponsible o he mobiliza ion o la ge amoun s o DOC,
which can explain obse ed sho e m a ia ions o DOC concen a ions in uno .
This concep ual iew on how DOC is being mobilized a he ca chmen -scale ela es physical
con olled mobiliza ion pa hways o he biogeochemical subs a e a ailabili y and includes scale-
b idging insigh s on DOC mobiliza ion and uno p oduc ion. Hyd ological and biogeochemical
p ocess in e ac ions, iden i ied o be ele an o he mobiliza ion o DOC in he Lehs enbach
ca chmen , a e, in ou opinion, o gene al signi icance and can be ans e ed o simila ecosys ems.
Howe e , his concep ional iew on how DOC is being mobilized in he Lehs enbach ca chmen has
o be u he imp o ed and e i ied. Recen ield in es iga ions (Kno , 2012) e.g. show ha
imescales o complexa ion and de-complexa ion o DOC wi h dissol ed i on in addi ion o i on
educ ion/oxida ion cycles signi ican ly con ol he a ailabili y o DOC, especially in he supe icial
laye s o he we lands. This so a has no been accoun ed o in he de eloped concep ual model.
Also, he signi icance o he in e play be ween di e en hyd ological low pa hs wi h hei indi idual
esponse and esidence imes combined wi h he spa ial he e ogenei y o biogeochemical condi ions
[48]
CONCLUSIONS AND OUTLOOK
( o es ed si es s. we lands) on nu ien cycling and solu e mobiliza ion is, in gene al, so a only s ill
poo ly unde s ood and mus be u he add essed in u u e wo k. Fi s p elimina y esul s o ni a e
(unpublished da a) o example hin ha he mobiliza ion p ocesses ac oss he ca chmen di e
signi ican ly om hose iden i ied o DOC, mainly because spa ial sou ces o ni a e and
biogeochemical ans o ma ion p ocesses along he low pa hs a e di e en o ni a e compa ed o
DOC.
On he small scale, esul s om he de eloped hyd ological and biogeochemical model, whe e sub-
su ace anspo p ocesses and kine ically con olled edox-sensi i e eac ions a e ep esen ed
equally, highligh how complex couplings be ween hyd ology and biogeochemis y can be wi hin
we land ecosys ems. One o he mos in e es ing esul s o his hesis is ha biogeochemical ho spo s
can o m e en in homogenous pea o we land soils, simply as a esul o he in e ac ions be ween a
highly dynamic, h ee-dimensional subsu ace low sys em induced by mic o- opog aphy and he
hyd ologically con olled biogeochemical bounda y condi ions ha ei he acili a e o supp ess edox-
sensi i e p ocesses. Resul s om his modeling app oach o e a new pe spec i e on biogeochemical
ans o ma ion p ocesses in ipa ian we lands which p o ides a dynamic amewo k o explain
p ocess he e ogenei y in we land soils and a iabili y in p ocess a es o e ime and space.
A nex s ep would be o app o e ha he simula ed mechanisms and in e ac ions be ween hyd ology
and biogeochemis y ac ually can esul in he o ma ion o biogeochemical ho spo s unde ield
condi ions. This is a challenging ask because cha ac e iza ion o subsu ace low pa e ns in si u,
necessa y o in es iga e in e ac ions be ween hyd ology and biogeochemis y, is e y di icul and
would equi e imp o ed expe imen al se ings. Howe e , he amewo k p esen ed as pa o his
hesis may be help ul o de elop such no el in si u expe imen s. Because o a ious limi a ions and
simpli ica ions, he hyd ological/biogeochemical modeling app oach so a is es ic ed o ela i e
simple es case scena ios. Fu u e wo k will ha e o add ess hese sho comings and imp o e he
modeling amewo k s epwise in o de o i o be applied o mo e ealis ic sys ems and o add ess
opics like he in e play be ween di e en s a ic (e.g. soil p ope ies, ege a ion pa e ns) and dynamic
con ols (e.g. low, empe a u e and ege a ion dynamics) o spa ial and empo al a ia ions in
biogeochemical p ocess ac i i ies in we lands.
Finally, his hesis has shown ha in e disciplina y esea ch e o s, combining he knowledge o
hyd ologis s and biogeochemis s, o e new pe spec i es on how ecosys ems a e unc ioning.
Howe e , a lo o knowledge gaps s ill exis and in o de o ill hese gaps and o imp o e ou
unde s anding on how nu ien s and elemen s a e cycled a a ious scales wi hin ecosys ems, i is
necessa y o u he o ganize "join ask o ces" among he di e en disciplines o de elop new
in e disciplina y app oaches whe e hyd ological and biogeochemical me hods and pe cep ions a e
being exchanged and adop ed.
[49]
REFERENCES
6 Re e ences
Alewell, C., Paul, S., Lischeid, G., S o ck, F.R., 2007. Co- egula ion o edox p ocesses in eshwa e
we lands as a unc ion o o ganic ma e a ailabili y? Science o he To al En i onmen 404 (2-
3), 335–342. doi:10.1016/j.sci o en .2007.11.001.
And ea, B., F ancesc, G., Jé ôme, L., Eusebi, V., F ancesc, S., 2006. C oss-si e compa ison o
a iabili y o DOC and ni a e c–q hys e esis du ing he au umn–win e pe iod in h ee
Medi e anean headwa e s eams: a syn he ic app oach. Biogeochemis y 77 (3), 327–349.
An oine, M., Ja aux, M., Bielde s, C., 2009. Wha indica o s can cap u e uno - ele an connec i i y
p ope ies o he mic o- opog aphy a he plo scale? Ad ances in Wa e Resou ces 32 (8), 1297–
1310. doi:10.1016/j.ad wa es.2009.05.006.
Baillie, M., Hogan, J., Ekwu zel, B., Wahi, A., Eas oe, C., 2007. Quan i ying wa e sou ces o a
semia id ipa ian ecosys em, San Ped o Ri e , A izona. Jou nal o Geophysical Resea ch 112
(G3), G03S02.
Ba be , K., 1981. Pea s a ig aphy and clima ic change: a palaeoecological es o he heo y o cyclic
pea bog egene a ion. Balkema, Ro e dam.
Belle ue, W., 2003. Tecplo Use 's Manual. Am ec Enginee ing Inc.
Be en, K., 1989. Changing ideas in hyd ology - The case o physically-based models. Jou nal o
Hyd ology 105 (1-2), 157–172.
Be en, K., 1993. P ophecy, eali y and unce ain y in dis ibu ed hyd ological modelling. Ad ances in
Wa e Resou ces 16 (1), 41–51.
Bishop, K., 1991. Episodic inc ease in s eam acidi y, ca chmen low pa hways and hyd og aph
sepe a ion. Ph.D. hesis, Camb idge Uni e si y, Camb idge, UK.
Bishop, K., Seibe , J., Köhle , S., Laudon, H., 2004. Resol ing he Double Pa adox o apidly
mobilized old wa e wi h highly a iable esponses in uno chemis y. Hyd ol. P ocess 18 (1),
185–189.
Boano, F., Dema ia, A., Re elli, R., Ridol i, L., 2010. Biogeochemical zona ion due o in ameande
hypo heic low. Wa e Resou ces Resea ch 46 (2), W02511.
B an i eun, B., Roule , N., 1998. The base low and s o m low hyd ology o a p ecamb ian shield
headwa e pea land. Hyd ological P ocesses 12 (1), 57–72.
B ons e , A., Pla e, E., 1997. Modelling o uno gene a ion and soil mois u e dynamics o
hillslopes and mic o-ca chmen s. Jou nal o Hyd ology 198 (1-4), 177–195.
B ook ield, A.E., Sudicky, E.A., Pa k, Y.J., Conan J ., B., 2009. The mal anspo modelling in a
ully in eg a ed su ace/subsu ace amewo k. Hyd ological P ocesses (23), 2150–2164.
10.1002/hyp.7282.
B ooks, P., Lemon, M., 2007. Spa ial a iabili y in dissol ed o ganic ma e and ino ganic ni ogen
concen a ions in a semia id s eam, San Ped o Ri e , A izona. Jou nal o Geophysical Resea ch
112 (G3), G03S05.
B ooks, R., Co ey, A., 1964. Hyd aulic p ope ies o po ous media. Hyd ology Pape s, Colo ado S a e
Uni e si y (Ma ch).

[50]
REFERENCES
Bu , T.P., Pinay, G., 2005. Linking hyd ology and biogeochemis y in complex landscapes. P og ess
in Physical Geog aphy 29 (3), 297–316.
Ca le, S., Fogg, G., 1996. T ansi ion P obabili y-Based Indica o Geos a is ics. Ma hema ical Geology
28 (4), 453–476.
Chapin, F., 1991. In eg a ed esponses o plan s o s ess. BioScience 41 (1), 29–36.
Chimne , R., Coope , D., 2003. In luence o wa e able le els on CO2 emissions in a Colo ado
subalpine en: an in si u mic ocosm s udy. Soil Biology and Biochemis y 35 (3), 345–351.
Ci mo, C.P., McDonnell, J.J., 1997. Linking he hyd ologic and biogeochemical con ols o ni ogen
anspo in nea -s eam zones o empe a e- o es ed ca chmen s: a e iew. Jou nal o Hyd ology
199 (1-2), 88–120.
Clemens, C., 2011. Biogeochemische Cha ak e isie ung de Wiede e nässung eines Hangmoo es im
Fich elgebi ge. Mas e Thesis, Uni e si y o Bay eu h, Bay eu h.
Conway, V., 1948. Von Pos 's wo k on clima ic hy hms. New Phy ologis 47 (2), 220–237.
C ock o d, R.H., Richa dson, D.P., 2000. Pa i ioning o ain all in o h ough all, s em low and
in e cep ion: e ec o o es ype, g ound co e and clima e. Hyd ological P ocesses 14 (16‐17),
2903–2920.
C ow, S., Wiede , R., 2005. Sou ces o CO2 emission om a no he n pea land: oo espi a ion,
exuda ion, and decomposi ion. Ecology 86 (7), 1825–1834.
Cu is, C., E ans, C., Goodale, C., Hea on, T., 2011. Wha Ha e S able Iso ope S udies Re ealed
Abou he Na u e and Mechanisms o N Sa u a ion and Ni a e Leaching om Semi-Na u al
Ca chmen s? Ecosys ems, 1–17.
Den , C., G imm, N., 1999. Spa ial he e ogenei y o s eam wa e nu ien concen a ions o e
successional ime. Ecology 80 (7), 2283–2298.
Den , C., G imm, N., Fishe , S., 2001. Mul iscale e ec s o su ace-subsu ace exchange on s eam
wa e nu ien concen a ions. Jou nal o he No h Ame ican Ben hological Socie y 20 (2), 162–
181.
De i o, K.J., Hill, A.R., 1997. Sulpha e dynamics in ela ion o g oundwa e -su ace wa e
in e ac ions in headwa e we lands o he sou he n Canadian Shield. Hyd ological P ocesses 11
(5), 485–500.
DeWalle, D., Swis ock, B., 1994. Causes o episodic acidi ica ion in i e Pennsyl ania s eams on he
no he n Appalachian Pla eau. Wa e Resou ces Resea ch 30 (7), 1955–1963.
D'odo ico, P., Laio, F., Po po a o, A., Rod iguez-I u be, I., 2003. Hyd ologic con ols on soil ca bon
and ni ogen cycles. II. A case s udy. Ad ances in Wa e Resou ces 26 (1), 59–70.
Do an, J.W., Linn, D.M., 1984. E ec o wa e - illed po e space on ca bon dioxide and ni ous oxide
p oduc ion in illed and non illed soils. Soil Science Socie y o Ame ica Jou nal 48 (6), 1267–
1272.
Eshleman, K., Polla d, J., O'B ien, A., 1994. In e ac ions be ween g oundwa e and su ace wa e in a
Vi ginia coas al plain wa e shed. 1. Hyd ological lowpa hs. Hyd ological P ocesses 8 (5), 389–
410.
[51]
REFERENCES
Es op-A agonés, C., Blodau, C., 2012. E ec s o expe imen al d ying in ensi y and du a ion on
espi a ion and me hane p oduc ion eco e y in en pea incuba ions. Soil Biology and
Biochemis y 47, 1–9.
Es op-A agonés, C., Kno , K., Blodau, C., 2012. Con ols on in si u oxygen and dissol ed ino ganic
ca bon dynamics in pea s o a empe a e en. Jou nal o Geophysical Resea ch 117 (G2),
G02002.
Fiedle , F.R., Rami ez, J.A., 2000. A nume ical me hod o simula ing discon inuous shallow low
o e an in il a ing su ace. In e na ional jou nal o nume ical me hods in luids 32 (2), 219–
239.
Fishe , S., G imm, N., Ma í, E., Holmes, R., Jones, J.J., 1998. Ma e ial spi aling in s eam co ido s:
a elescoping ecosys em model. Ecosys ems 1 (1), 19–34.
Fishe , S., Sponselle , R., He e nan, J., 2004. Ho izons in s eam biogeochemis y: lowpa hs o
p og ess. Ecology 85 (9), 2369–2379.
Fi zge ald, D.F., P ice, J.S., Gibson, J.J., 2003. Hillslope-swamp in e ac ions and low pa hways in a
hype ma i ime ain o es , B i ish Columbia. Hyd ological P ocesses 17 (15), 3005–3022.
10.1002/hyp.1279.
Fleckens ein, J., Niswonge , R., Fogg, G., 2006. Ri e -Aqui e In e ac ions, Geologic He e ogenei y,
and Low-Flow Managmen . G oundwa e 44 (6), 837–852.
F ase , C.J., Roule , N.T., Moo e, T.R., 2001. Hyd ology and dissol ed o ganic ca bon
biogeochemis y in an omb o ophic bog. Hyd ological P ocesses 15 (16), 3151–3166.
10.1002/hyp.322.
F ei, S., Fleckens ein, J.H., Kolle , S.J., Maxwell, R.M., 2009. Pa e ns and dynamics o i e -aqui e
exchange wi h a iably-sa u a ed low using a ully-coupled model. Jou nal o Hyd ology 375
(3-4), 383–393.
Ge mann, P., Ande son, M.G., Bu , T.P., 1990. Mac opo es and hyd ologic hillslope p ocesses.
P ocess S udies in Hillslope Hyd ology, 327–363.
Ge s be ge , P. (Ed.), 2001. Waldökosys em o schung in No dbaye n: Die BITÖK-
Un e suchungs lächen im Fich elgebi ge und S eige wald. Bay eu he Fo um Ökologie.
Gillham, R.W., 1984. The capilla y inge and i s e ec on wa e - able esponse. Jou nal o
Hyd ology 67 (1-4), 307–324.
Godwin, H., Conway, V., 1939. The ecology o a aised bog nea T ega on, Ca diganshi e. The
Jou nal o Ecology, 313–359.
Hauck, A., 1999. Hyd ological Cha ak e iza ion o he Lehs enbach Ca chmen . unpuplished Diploma
Thesis, Uni e si y o Bay eu h, Bay eu h.
Helio is, F., DeWi , C., 1987. Rapid Wa e Table Responses o Rain all in a No he n Pea land
Ecosys em. JAWRA Jou nal o he Ame ican Wa e Resou ces Associa ion 23 (6), 1011–1016.
Hill, A., 1993. Ni ogen dynamics o s o m uno in he ipa ian zone o a o es ed wa e shed.
Biogeochemis y 20 (1), 19–44.
Hill, A., De i o, K., Campagnolo, S., Sanmugadas, K., 2000. Subsu ace deni i ica ion in a o es
ipa ianzone: In e ac ions be ween hyd ology and supplies o ni a e and o ganic ca bon.
Biogeochemis y 51 (2), 193–223.
[52]
REFERENCES
Holden, J., Bu , T.P., 2003. Hyd aulic conduc i i y in upland blanke pea : measu emen and
a iabili y. Hyd ological P ocesses 17 (6), 1227–1237.
Hood, E., Goose , M., Johnson, S., 2006. Changes in he cha ac e o s eam wa e dissol ed o ganic
ca bon du ing lushing in h ee small wa e sheds, O egon.
Hopp, L., McDonnell, J.J., 2009. Connec i i y a he hillslope scale: Iden i ying in e ac ions be ween
s o m size, bed ock pe meabili y, slope angle and soil dep h. Jou nal o Hyd ology 376 (3), 378–
391.
Hun e , K.S., Wang, Y., an Cappellen, P., 1998. Kine ic modeling o mic obially-d i en edox
chemis y o subsu ace en i onmen s: coupling anspo , mic obial me abolism and
geochemis y. Jou nal o Hyd ology(Ams e dam) 209 (1), 53–80.
Hy önen, R., Åg en, G., Linde , S., Pe sson, T., Co u o, M., Ekblad, A., F eeman, M., G elle, A.,
Janssens, I., Ja is, P., 2007. The likely impac o ele a ed [CO2], ni ogen deposi ion, inc eased
empe a u e and managemen on ca bon seques a ion in empe a e and bo eal o es ecosys ems:
a li e a u e e iew. New Phy ologis 173 (3), 463–480.
Jacks, G., No s öm, A.C., 2004. Hyd ochemis y and hyd ology o o es ipa ian we lands. Fo es
ecology and Managemen 196 (2-3), 187–197.
Johns on, C.A., 1991. Sedimen and nu ien e en ion by eshwa e we lands: e ec s on su ace
wa e quali y. C i ical Re iews in En i onmen al Science and Technology 21 (5), 491–565.
Jones J , J., Fishe , S., G imm, N., 1995. Ve ical hyd ologic exchange and ecosys em me abolism in
a Sono an Dese s eam. Ecology, 942–952.
Jones, J., Holmes, R., Fishe , S., G imm, N., G eene, D., 1995. Me hanogenesis in A izona, USA
d yland s eams. Biogeochemis y 31 (3), 155–173.
Jones, J.P., Sudicky, E.A., B ook ield, A.E., Pa k, Y.J., 2006. An assessmen o he ace -based
app oach o quan i ying g oundwa e con ibu ions o s eam low. Wa e Resou ces Resea ch 42,
W02407. doi:10.1029/2005WR004130.
Jones, J.P., Sudicky, E.A., McLa en, R.G., 2008. Applica ion o a ully-in eg a ed su ace-subsu ace
low model a he wa e shed-scale: A case s udy. Wa e Resou ces Resea ch 44, W03407.
doi:10.1029/WR005603.
Kalbi z, K., Solinge , S., Pa k, J., Michalzik, B., Ma zne , E., 2000. Con ols on he dynamics o
dissol ed o ganic ma e in soils: a e iew. Soil Science 165 (4), 277.
Ke unen, A., Kai ala, V., Leh inen, A., Lohila, A., Alm, J., Sil ola, J., Ma ikainen, P., 1999.
Me hane p oduc ion and oxida ion po en ials in ela ion o wa e able luc ua ions in wo bo eal
mi es. Soil Biology and Biochemis y 31 (12), 1741–1749.
Kinzelbach, W., Rausch, R., 1995. G undwasse modellie ung: Eine Ein üh ung mi Übungen; mi 15
Tabellen. Bo n aege .
Kno , K., Blodau, C., 2009. Impac o expe imen al d ough and ewe ing on edox ans o ma ions
and me hanogenesis in mesocosms o a no he n en soil. Soil Biology and Biochemis y 41 (6),
1187–1198.
Kno , K., Oos e woud, M., Blodau, C., 2008. Expe imen al d ough al e s a es o soil espi a ion and
me hanogenesis bu no ca bon exchange in soil o a empe a e en. Soil Biology and
Biochemis y 40 (7), 1781–1791.
[53]
REFERENCES
Kno , K.-H., 2012. DOC-dynamics in a small headwa e ca chmen as d i en by edox luc ua ions
and hyd ological low pa hs - a e DOC expo s media ed by i on educ ion/oxida ion cycles?
Biogeosciences Discussions 9, 1–34.
Kno , K.H., Lischeid, G., Blodau, C., 2009. Dynamics o edox p ocesses in a mine o ophic en
exposed o a wa e able manipula ion. Geode ma 153 (3-4), 379–392.
Koehle , A., Mu phy, K., Kiely, G., So oco nola, M., 2009. Seasonal a ia ion o DOC concen a ion
and annual loss o DOC om an A lan ic blanke bog in Sou h Wes e n I eland.
Biogeochemis y 95 (2), 231–242.
K is ensen, K., Jensen, S., 1975. A model o es ima ing ac ual e apo anspi a ion om po en ial
e apo anspi a ion. No dic Hyd ology 6 (3), 170–188.
K use, J., Lenna z, B., Leinwebe , P., 2008. A modi ied me hod o measu ing sa u a ed hyd aulic
conduc i i y and aniso opy o en pea samples. We lands 28 (2), 527–531.
K æ ne , J., Klø e, B., 2008. Gene a ion and egula ion o summe uno in a bo eal la en. Jou nal
o Hyd ology (360), 15–30. doi:10.1016/j.jhyd ol.2008.07.009.
Lee, S.-Y., Ca le, S., Fogg, G., 2007. Geologic he e ogenei y and a compa ison o wo geos a is ical
models: Sequen ial Gaussian and ansi ion p obabili y-based geos a is ical simula ion. Ad ances
in Wa e Resou ces 30 (9), 1914–1932.
Lemieux, J., Sudicky, E.A., Pel ie , W.R., Ta aso , L., 2008a. Simula ing he impac o glacia ions on
con inen al g oundwa e low sys ems: 1. Rele an p ocesses and model o mula ion. Jou nal o
Geophysical Resea ch 113 (F3), F03017.
Lemieux, J., Sudicky, E.A., Pel ie , W.R., Ta aso , L., 2008b. Simula ing he impac o glacia ions on
con inen al g oundwa e low sys ems: 2. Model applica ion o he Wisconsinian glacia ion o e
he Canadian landscape. Jou nal o Geophysical Resea ch 113 (F3), F03018.
Lemieux, J.M., Sudicky, E.A., Pel ie , W.R., Ta aso , L., 2008c. Dynamics o g oundwa e echa ge
and seepage o e he Canadian landscape du ing he Wisconsinian glacia ion. Jou nal o
Geophysical Resea ch 113 (F1), F01011.
Li, Q., Unge , A.J., Sudicky, E.A., Kassenaa , D., Wexle , E.J., Shikaze, S., 2008. Simula ing he
mul i-seasonal esponse o a la ge-scale wa e shed wi h a 3D physically-based hyd ologic
model. Jou nal o Hyd ology (357), 317–336. doi:10.1016/j.jhyd ol.2008.05.024.
Lischeid, G., 2008. Combining Hyd ome ic and Hyd ochemical Da a Se s o In es iga ing Runo
Gene a ion P ocesses: Tau ologies, Inconsis encies and Possible Explana ions. Geog aphy
Compass 2 (1), 255–280. 10.1111/j.1749-8198.2007.00082.x.
Lischeid, G., Kolb, A., Alewell, C., 2002. Appa en ansla o y low in g oundwa e echa ge and
uno gene a ion. Jou nal o Hyd ology 265 (1-4), 195–211.
Lischeid, G., Kolb, A., Alewell, C., Paul, S., 2007. Impac o edox and anspo p ocesses in a
ipa ian we land on s eam wa e quali y in he Fich elgebi ge egion, sou he n Ge many.
Hyd ological P ocesses 21 (1), 123–132. 10.1002/hyp.6227.
Lohse, K., B ooks, P., McIn osh, J., Meixne , T., Huxman, T., 2009. In e ac ions be ween
biogeochemis y and hyd ologic sys ems. Annual Re iew o En i onmen and Resou ces 34, 65–
96.
McCa y, J.J., Canziani, O.F., Lea y, N.A., Dikken, D.J., Whi e, K.S., 2001. IPCC Clima e Change
2001: Impac s, Adap a ion and Vulne abili y. The Thi d Assessmen Repo o Wo king G oup
[60]
APPENDIX
Figu e A2: Sa u a ed hyd aulic conduc i i ies Ksa assigned o he en sub-laye s SL1-SL10 o he
we land a eas o he ca chmen scale model. Ksa alues a e exponen ially dec easing (as indica ed
by he linea dec ease using a loga i hmic X-axis) wi h dep h o mimic he ansmissi i y eedback
mechanism. Values o used Ksa - alues wi hin he we land a eas a e based on he s udy o Jacks and
No s öm, (2004).

[61]
APPENDIX
Table A1: O e iew o he pa ame e iza ion o he ca chmen scale model o ep esen
su ace/subsu ace low and in e ac ions o he h ee di e en zones (we lands, upslope a eas and
s eam a eas).
a. Subsu ace We lands Upslope
A eas
sa u a edhyd aulic
conduc i i y[m/d]
a iablewi h
dep h(see
Figu eA2)
0.24

po osi y[‐]0.5 0.4
speci ics o age[m‐1]0.0001 0.0001
b. Su aceWe lands Upslope
A easS eam
su aces o age[m]0.1;0.5;1.00.010.0
couplingleng h[m]0.10.10.0001
ic ionslopesXandY[m‐1/3s]8.1x10‐71.9x10‐64.0x10‐7
c. E apo anspi a ionWe lands Upslope
A eas
lea A eaIndex[‐]3.06.5
oo dep h[m]
(quad a icdecay unc ion)0.83.0

e apo a iondep h[m]
(quad a icdecay unc ion)0.50.5



[62]
[63]
CONTRIBUTIONS TO THE INCLUDED MANUSCRIPTS
8 Con ibu ions o he included manusc ip s
S udy 1
E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno gene a ion in a i ual
ipa ian we land – a modeling s udy.
Au ho s: S en F ei, Gunna Lischeid and Jan H. Fleckens ein
S en F ei: concep s, modeling, in e p e a ion and discussion o esul s,
manusc ip p epa a ion
Gunna Lischeid: commen s on manusc ip , ield si e coo dina o , ield da a
Jan H. Fleckens ein: concep s, discussion o esul , manusc ip p epa a ion
S udy 2
Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in we land sys ems – a
i ual modeling expe imen .
Au ho s: S en F ei, Klaus-Holge Kno , S e an Pei e and Jan H. Fleckens ein
S en F ei: concep s, modeling, in e p e a ion and discussion o esul s,
manusc ip p epa a ion
Klaus-Holge Kno : ield da a, concep s, in e p e a ion and discussion o esul s,
manusc ip p epa a ion
S e an Pei e : commen s on manusc ip , discussion o esul s
Jan H. Fleckens ein: in e p e a ion and discussion o esul s, manusc ip p epa a ion
S udy 3
Rep esen ing e ec s o mic o- opog aphy on uno gene a ion and subsu ace low pa e ns by
using supe icial ill s o age heigh a ia ions.
Au ho s: S en F ei, and Jan H. Fleckens ein
S en F ei: concep s, modeling, in e p e a ion and discussion o esul s, manusc ip
p epa a ion
Jan H. Fleckens ein: concep s, discussion o esul s, manusc ip p epa a ion
[64]
CONTRIBUTIONS TO THE INCLUDED MANUSCRIPTS
S udy 4
Concen a ions and luxes o dissol ed o ganic ca bon in uno om a o es ed ca chmen :
Insigh s om high equency measu emen s
Au ho s: S e an S ohmeie , Klaus-Holge Kno , Ma in Reiche , S en F ei, Jan H. Fleckens ein,
S e an Pei e and Egbe Ma zne
S e an S ohmeie : concep s, in e p e a ion and discussion o esul s, manusc ip p epa a ion,
modeling
Klaus-Holge Kno : in e p e a ion and discussion o esul s, labo a o y wo k
Ma in Reiche : ield wo k, labo a o y wo k
S en F ei: modeling, in e p e a ion and discussion o esul s, commen s on manusc ip
Jan H. Fleckens ein: commen s on manusc ip
S e an Pei e : commen s on manusc ip
Egbe Ma zne : manusc ip p epa a ion, concep s, in e p e a ion and discussion o esul s
S udy 5
In e p e ing low gene a ion mechanisms om in eg a ed su ace wa e -g oundwa e low
models o a ipa ian we land and ca chmen .
Au ho s: Daniel Pa ing on, Philip A. B unne , S en F ei, C aig T. Simmons, Ad ian D. We ne , René
The ien, Holge R. Maie , G aeme C. Dandy and Jan H. Fleckens ein
Daniel Pa ing on: concep s, coding, modeling, in e p e a ion and discussion o esul s, manusc ip
p epe a ion
Philip A. B unne : concep s, discussion o esul s, commen s on manusc ip
S en F ei: de elopmen o low models, discussion o esul s, manusc ip p epe a ion
C aig T. Simmons: discussion o esul s, commen s on manusc ip
Ad ian D. We ne : commen s on manusc ip
René The ien: concep s, discussion o esul s, commen s on manusc ip
Holge R. Maie : commen s on manusc ip
G aeme C. Dandy commen s on manusc ip
Jan H. Fleckens ein commen s on manusc ip
[65]
STUDY 1
S udy 1
E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno gene a ion
in a i ual ipa ian we land – a modeling s udy.
By S en F ei, Gunna Lischeid and Jan H. Fleckens ein
Published in Ad ances in Wa e Resou ces 33 (2010) 1388-1401

[66]
[67]
STUDY 1
Published in Ad ances in Wa e Resou ces 33 (2010) 1388-1401
E ec s o mic o- opog aphy on su ace-subsu ace exchange and uno gene a ion
in a i ual ipa ian we land – a modeling s udy.
F ei1, S., G. Lischeid2 and J. H. Fleckens ein3
1 Depa men o Hyd ology, Uni e si y o Bay eu h, Ge many
2 Leibni z Cen e o Ag icul u al Landscape Resea ch, (ZALF), Ge many
3 Depa men Hyd ogeology, Helmhol z-Cen e o En i onmen al Resea ch – UFZ, Ge many
Abs ac
In humid upland ca chmen s we lands a e o en a p ominen ea u e in he icini y o s eams and
ha e po en ial implica ions on uno gene a ion and nu ien expo . We land su aces a e o en
cha ac e ized by dis inc mic o- opog aphy (hollows and hummocks). The e ec s o such mic o-
opog aphy on su ace-subsu ace exchange and uno gene a ion o a 10 by 20 m syn he ic sec ion
o a ipa ian we land we e in es iga ed in a i ual modeling expe imen . A e e ence model wi h a
plana su ace was un o compa ison. The geos a is ically simula ed s uc u e o he mic o-
opog aphy eplica es he opog aphy o a pea - o ming ipa ian we land in a small moun ainous
ca chmen in Sou h-Eas Ge many (Lehs enbach). Flow was modeled wi h he ully in eg a ed
su ace-subsu ace code Hyd oGeoSphe e. Simula ion esul s show ha he speci ic s uc u e o he
we land su ace esul s in dis inc shi s be ween su ace and subsu ace low dominance. Su ace
dep essions ill and s a o d ain ia connec ed channel ne wo ks in a h eshold-con olled p ocess,
when g oundwa e le els in e sec he land su ace. These ne wo ks expand and sh ink in a spill and
ill mechanism when he shallow wa e able luc ua es a ound he mean su ace ele a ion unde
a iable ain all inpu s. The mic o- opog aphy e icien ly bu e s ain all inpu s and p oduces a
hyd og aph ha is cha ac e ized by subsu ace d ainage du ing mos o he yea and only empo a ily
shi s o su ace low dominance (> 80% o o al discha ge) du ing in ense ains o ms. In con as he
hyd og aph in he plana e e ence model is much “ lashie ” and mo e con olled by su ace uno . A
non-linea , hys e e ic ela ionship be ween g oundwa e le el and discha ge obse ed a he s udy si e
was ep oduced wi h he mic o- opog aphy model. Hys e esis was also obse ed in he ela ionship
be ween su ace wa e s o age and discha ge, bu o e a ela i ely na ow ange o su ace wa e
s o age alues. The e o e i was concluded ha su ace wa e s o age was a be e p edic o o he
occu ence o su ace uno han g oundwa e le els.
[68]
STUDY 1
1 In oduc ion
Ripa ian zones con ain dynamic in e aces be ween g ound- and su ace wa e lowpa hs [10, 27]. I
is impo an o unde s and he mechanisms ha go e n hyd ologic lowpa hs and s eam low
gene a ion in ipa ian zones because nu ien ans o ma ions and expo a e in eg ally ela ed o he
hyd ologic dynamics [8, 53, 18, 34]. Howe e , hese dynamics can be qui e complex [53, 27] and a e
gene ally poo ly unde s ood [49, 28].
In humid empe a e clima es ipa ian zones a e o en occupied by we lands [34, 36, 27]. Rapid
su ace and shallow subsu ace lows ypically domina e uno gene a ion in ipa ian we lands du ing
ains o ms [8, 34]. Gibson e al. [17] showed ha uno dynamics highly depend on su ace s o age
and in e ac ions be ween su ace wa e and shallow g oundwa e . K æ ne and Klø e [27] iden i ied
dis inc ly di e en uno gene a ion p ocesses wi h shi s be ween subsu ace and su ace low
dominance o low and high low e en s. Non-linea ela ionships be ween ipa ian wa e able dep h
and s eam low ha e o en been obse ed [5, 15, 38, 44]. Fo ca chmen s domina ed by ma ix low
hese ela ionships ha e been a ibu ed o he ansmissi i y eedback mechanism [4, 3, 44]. S eam
low o igina ing om ma ix low inc eases exponen ially, when he wa e able ises in o soil laye s
wi h p og essi ely inc easing la e al hyd aulic conduc i i y [3, 44]. In sys ems whe e shi s be ween
ma ix low and su ace low dominance occu , addi ional dynamics and non-linea i ies ha e been
obse ed (e.g. [27]).
Pea - o ming we lands a e o en cha ac e ized by a hummocky opog aphy wi h sequences o high
poin s (hummocks) and dep essions (hollows) a he sub-me e scale, which will a ec uno
gene a ion du ing ansi ions be ween su ace and subsu ace low dominance. E ec s o mic o-
opog aphy on in il a ion and uno gene a ion p ocesses we e i s in es iga ed by Dunne e al. [9].
They showed ha hill slope uno was con olled by an in ica e in e play be ween ain all in ensi y,
su ace low dep h, ege a ion co e and he speci ic mic o- opog aphy o he slope. Mic o-
opog aphy can a enua e and delay su ace lows [36, 27], because su ace dep essions i s need o
be illed un il a speci ic su ace wa e s o age h eshold is exceeded and hen su ace low owa ds he
s eam channel can be ini ia ed [2, 12]. T omp- an Mee eld and McDonnell [51] and T omp- an
Mee eld and McDonnell [52] e med simila h eshold dynamics in he gene a ion o subsu ace
s o m lows on bed ock su aces wi h mic o- opog aphy he " ill and spill mechanism". Qu and Du y
[40] epo ed dis inc double peaks in hyd og aphs om single ain all e en s, which hey asc ibed o
complex in e ac ions be ween small scale mic o- opog aphy con olled su ace uno in he we land
and subsu ace low.
Se e al modeling s udies ha e add essed he e ec s o mic o- opog aphy on uno dynamics. Dunne
e al. [9] used a concep ual app oach o simula e o e land low and in il a ion p ocesses o uni o m
[69]
STUDY 1
sinusoidal mic o- opog aphy. They demons a ed ha mic o- opog aphy esul ed in signi ican spa ial
a iabili y o in il a ion and su ace lows. Es e es e al. [11] and Fiedle and Rami ez [12] used
ini e di e ence solu ions o he wo-dimensional dep h-a e aged dynamic wa e equa ions o simula e
o e land low and in il a ion p ocesses on small plo s wi h mic o- opog aphy. Bo h s udies showed
ha mic o- opog aphy s ongly a ec s low di ec ions, low eloci ies and low dep hs and esul ed in
su ace low along well de ined mic o-channels. Connec i i y indica o s o su ace low on plo s
wi h mic o- opog aphy we e sys ema ically in es iga ed wi h a nume ical model by An oine e al. [2].
Each o he a o emen ioned modeling s udies we e es ic ed o su ace lows and in il a ion and did
no accoun o eedbacks be ween su ace and subsu ace low, an impo an p ocess in we lands [8,
17]. An excep ion was he s udy by Qu and Du y [40], who used a ini e elemen coupled su ace-
subsu ace low model o simula e a se ies o ain all e en s o a 0.08 km2 wa e shed in
Pennsyl ania. They demons a ed how small scale opog aphy can con ol local su ace sa u a ion and
subsequen connec i i y o su ace low pa hs leading o s eam low gene a ion. Howe e he spa ial
esolu ion o he Qu and Du y [40] model was oo coa se o accoun o mic o- opog aphy on he
sub-me e scale. Hopp and McDonnell [20] modeled he e ec s o bed ock mic o- opog aphy on
subsu ace s o m low gene a ion om hillslopes.
Ou wo k e alua es he complex hyd ologic dynamics o a ipa ian we land wi h mic o- opog aphy
h ough a i ual modeling expe imen . The pu pose o he simula ions is o examine p ocess
dynamics a he han calib a ion o a model o a speci ic ield si e. We a gue ha o accu a ely
desc ibe hese dynamics a nume ical model has o accoun o o e land low, a iably sa u a ed
subsu ace low and complex in e ac ions be ween he su ace and subsu ace domains. A ully-
in eg a ed modeling app oach simul aneously sol es all o he equa ions ha go e n he complex
in e ac ions be ween su ace and subsu ace. E icien nume ical models ha use he ully-in eg a ed
app oach ha e become a ailable in ecen yea s (e.g. [25, 50]). The ully in eg a ed, h ee-
dimensional nume ical low model Hyd oGeoSphe e [50] is used he e o examine hyd ologic
dynamics in a i ual ipa ian we land wi h dis inc mic o- opog aphy (hummocks and hollows). The
mic o- opog aphic elie is geos a is ically gene a ed o a 10m x 20m a ea a a esolu ion o
app oxima ely 0.1m based on su eyed mic o- opog aphy in a ipa ian we land o he small
expe imen al Lehs enbach ca chmen loca ed in Ge many (Figu e 1). The we lands in he ca chmen ,
which ha e a hummocky su ace opog aphy, can be classi ied as ens. The ela i e ele a ion
di e ences be ween hollows and hummocks ange be ween 0.2-0.4 m and he hollows a e gene ally
in e -connec ed. In lows om deepe g oundwa e a e locally diminished by a basal clay laye . A
se e al loca ions la e al in lows om adjacen hillslopes a e in e cep ed by small s eam channels
bounding he we lands. Mos small s eams ha e hei headwa e s in he we lands and p ac ically all
he wa e ha eaches he s eams ei he o igina es in o passes h ough he we lands.
[76]
STUDY 1
condi ion a he channel ou le . All o he bounda ies we e se o no- low bounda ies wi h he
excep ion o he uppe model su ace whe e a iable ain all a es a e applied. The ini ial g oundwa e
ele a ion was p esc ibed as 0.5 m abo e he ho izon al base o he model wi h an equilib ium p essu e
dis ibu ion abo e he wa e able. Daily p ecipi a ion was applied o he model su ace based on he
ain all eco d om he 2000 hyd ologic yea (No embe 1999 h ough Oc obe 2000). The su ace
domain was ini ialized wi h a ze o dep h o ponded wa e ep esen ing d y ini ial condi ions. The
ic ion slope o su ace low calcula ions is desc ibed using Manning's equa ion. Manning's
oughness coe icien s o he pea su ace we e uni o mly assigned as 0.03 m-1/3s o x and y; a alue
epo ed o densely ege a ed su aces [45].

[77]
STUDY 1
3 Resul s
3.1 Dynamics o uno gene a ion o s eady ain all
To in es iga e he gene al dynamics o discha ge gene a ion unde inc easing we ness, a simula ion
wi h a cons an ain all a e o 0.008 m/d was un un il he s eady s a e discha ge a he channel ou le
was a ained. The ain all a e ep esen ed condi ions o a mode a e o in ense ains o m (exceeded on
abou 40 days pe yea o a ypical hyd ologic yea ) and ensu ed ha su ace low ne wo ks could
de elop be o e he inal s eady s a e was eached. Figu e 4 shows he de elopmen o channel
discha ge and he wa e able (e alua ed a an obse a ion well a he up-s eam end o he model
domain – see Figu e 3) o he plana (uppe panel) and he mic o- opog aphy model (lowe panel)
espec i ely. In he ini ial s age o bo h simula ions channel discha ge g adually inc eased om
subsu ace in lows caused by inc easing hyd aulic g adien s owa ds he channel. The inc ease was
mo e apid in he plana model compa ed o he mic o- opog aphy model. The slowe and sligh ly
undula ing inc ease in he la e case was caused by he p og essi e o ma ion o ponds when he
wa e able in e sec s local su ace dep essions. A his poin he build-up o subsu ace g adien s
owa ds he channel was slowed. The same inpu o subsu ace heads below he ponds inc eased less
apidly as i he same amoun o wa e had in il a ed (due o he po osi y). Su ace low in he plana
model, indica ed by a s eep inc ease in channel discha ge, occu s a e app oxima ely 16 days.
Discha ge subsequen ly inc eased apidly un il he sys em a ained a s a e o equilib ium wi h cons an
discha ge a ound day 24. In he mic o- opog aphy model isola ed ponds a he su ace de eloped
connec ed channel ne wo ks, which e en ually spilled in o he main channel segmen a ound day 45.
The subsequen apid inc ease in discha ge displayed se e al kinks, which ep esen ed he
de elopmen and ma u a ion o di e en su ace low ne wo ks. The ne wo ks e en ually all p o ided
wa e o he channel when equilib ium was eached a ound day 50. G oundwa e le els a equilib ium
(e alua ed a he loca ion in he upslope cen e o he domain – see Figu e 3) we e abou 0.18 m
below he land su ace o he mic o- opog aphy model and a he land su ace o he plana model.
[78]
STUDY 1
Figu e 4: Hyd og aphs and de elopmen o he local g oundwa e le el o a simula ion wi h cons an
ain all (0.008 m/d). Resul s o he plana model a e shown a he op and o he mic o- opog aphy
model on he bo om. Channel discha ge and g oundwa e le el a e e alua ed a he channel ou le
and in an obse a ion well (as shown in Figu e 3).
[79]
STUDY 1
3.2 Runo dynamics and low componen s o a iable ain all
Figu e 5 shows he simula ed discha ge hyd og aph a he main channel ou le (lowe igh co ne o
he domain – see Figu e 3) o he mic o- opog aphy model. Discha ge is sepa a ed in o a su ace and
a subsu ace low componen . The model- o cing daily p ecipi a ion eco d is depic ed on he op
axis. The sepa a ion o low componen s was achie ed by placing "hyd og aph nodes", which acked
all low h ough a node in he g id, along he edged o he channel segmen and a he channel ou le
(see Figu e 3). The su ace low hyd og aph nodes acked su ace low in o he channel a each ime
s ep o he simula ion. The subsu ace low componen s acked all low ha exi ed he model domain
(sum o su ace and subsu ace lows). The di e ence be ween he wo componen s ep esen ed
subsu ace lows in o he channel segmen . S eady ain all in echa ged g oundwa e , he g oundwa e
le els inc eased, and he hyd aulic g adien s o he s eam inc eased, esul ing in inc eased subsu ace
lows. A e ini ial we ing o he sys em, su ace low ia su ace channel ne wo ks, was ini ia ed on
day 125. Maximum discha ge was simula ed o day 217 a e he mos in ensi e ain all e en in he
annual eco d (48 mm/d). Simula ed discha ge was gene a ed ia subsu ace low du ing mos o he
yea . Only on 52 o he 365 simula ed days was su ace low obse ed in he model. On hese 52 days,
su ace low accoun ed o up o 85% o o al channel discha ge (see Fi zge ald e al. [17] o a ield
example).
Figu e 5: Simula ed, yea ly hyd og aph o he mic o- opog aphy model. P ecipi a ion a he ield si e
o he hyd ologic yea 2000 (10/31/1999 – 11/1/2000) is shown on he op. Su ace and subsu ace
ac ions o o al channel discha ge a e shown in g een and black espec i ely.
[80]
STUDY 1
Figu e 6 (panel a) shows a ypical si ua ion du ing pe iods wi h low o in e media e ain all
in ensi ies. Wa e was al eady ponded in local dep essions (hollows) a he su ace. Howe e , ponded
a eas a e no all in e connec ed and su ace d ainage in o he channel segmen was inhibi ed by he
mic o- opog aphy. Only du ing high ain all a es (panel b) did pond a eas s a o become
in e connec ed and o m ex ended su ace low ne wo ks and mic o-channels. Unde hese condi ions
a la ge ac ion o he we land su ace d ained in o he adjacen channel. D ainage in o he channel
occu ed a wo dis inc loca ions (Figu e 6). Simila pa e ns we e obse ed in he en loca ed a he
ield si e du ing a ains o m in he sp ing o 2009 (Figu e 7).
Figu e 6: Snapsho s o he e ol ing su ace low ne wo ks o a) mode a e low condi ions (day 180)
and b) du ing peak low (day 218). Blue zones indica e ponded su ace wa e and yellow a ows
s eam aces in he su ace low ne wo ks. Snapsho s show simula ed esul s.
Figu e 7: Pic u e o he ield si e aken du ing a s o m- low e en in sp ing 2009. Channel loca ion is
ma ked by a line.
[81]
STUDY 1
Figu e 8 shows he simula ed discha ge hyd og aph o he plana model. Compa ed o he mic o-
opog aphy model, he hyd og aph gene ally showed highe peak discha ges. Su ace lows we e
gene a ed much ea lie (a ound day 55) and occu ed mo e equen ly compa ed o he mic o-
opog aphy model (75 o 365 simula ed days). Du ing he ela i ely d y summe pe iod be ween day
150 and 217, ain all in ensi ies du ing he six di e en e en s we e high enough o gene a e su ace
d ainage. The mic o- opog aphy model, in compa ison, did no show any su ace d ainage du ing his
pe iod. In he plana model su ace d ainage was no inhibi ed by mic o- opog aphic s uc u es and
could occu as shee low as soon as he wa e able in e sec ed he land su ace. In he plana model
su ace low con ibu ed up o 95% o he o al discha ge du ing indi idual e en s.
Figu e 8: Simula ed, yea ly hyd og aph o he plana e e ence model. P ecipi a ion a he ield si e
o he hyd ologic yea 2000 (10/31/1999 – 11/1/2000) is shown on he op. Su ace and subsu ace
ac ions o o al channel discha ge a e shown in g een and black espec i ely.
3.3 Non-linea i ies and hys e esis
No unique g oundwa e le el o ain all a e could be associa ed wi h he de elopmen o su ace low
ne wo ks and he onse o su ace lows. In con as he amoun o ponded su ace wa e , necessa y o
ini ia e low o he channel ia he su ace low ne wo ks and mic o-channels, was na owly de ined.
Figu e 9 (uppe plo ) shows he ela ionship be ween su ace discha ge and su ace wa e s o age
( o al amoun o ponded su ace wa e in m³ s o ed in local dep ession and low ne wo ks). Figu e 9
(uppe plo ) summa izes esul s o he 365 day simula ion o he mic o- opog aphy model. The
di e en loops ep esen di e en ajec o ies o single ain all e en s. The ajec o y o he mos

[82]
STUDY 1
in ense ains o m o he simula ed yea is discussed in mo e de ail (ma ked by a line in Figu e 9 uppe
plo ). This p ecipi a ion e en (48 mm/d) occu ed igh a e an ex ended d ie pe iod (day 150 o
217, see Figu e 5) on day 217 ollowed by only 7.2 mm/d on day 218. In he beginning o he ain
s o m in il a ing ainwa e exclusi ely echa ged g oundwa e (no su ace discha ge). Wi h ising
g oundwa e le els, local dep essions we e illed wi h wa e and inc easingly mo e wa e was s o ed
on he soil su ace (inc easing su ace s o age wi hou su ace low in he channel). La e , he illed
dep essions s a o in e connec un il a c i ical su ace s o age alue (~5.6m³) was exceeded. The
esul ing low ne wo k was subsequen ly la ge enough o p o ide i s su ace low o he channel.
The su ace low apidly inc eased un il i eached a s able a e o ~3.8 m³/d. A e ha su ace low
ab up ly inc eased as su ace s o age exceeded ano he c i ical alue (~7.2m³). This was caused by a
second su ace low ne wo k ha de eloped and d ained independen ly om he i s one. Tha is
illus a ed in Figu e 10 by di e en snapsho s, aken o i e di e en ime s eps.
[83]
STUDY 1
Figu e 9: Rela ionship be ween su ace s o age and channel discha ge o he mic o- opog aphy
model (uppe panel) and he plana model (lowe panel). The black line depic s he peak low e en
a ound day 218. Sc i (max) - Sc i (min) (uppe and lowe panels) ep esen s he c i ical ange o su ace
wa e s o age, wi hin which su ace lows occu in he yea ly simula ions.
[84]
STUDY 1
Figu e 10: Six consecu i e snapsho s o he e ol ing su ace low ne wo ks du ing he la ges low
e en o he yea (day 217 o day 218). The ed lines sepa a e di e en low ne wo ks (1-3) ha
de eloped independen ly om each o he .
The lines on he model su ace in Figu e 10 delinea e di e en su ace low ne wo ks, de e mined by
an analysis o o e land low s eam aces. The i s snapsho shows he si ua ion igh a he
beginning o he ains o m on day 217. Due o he p eceding d ie pe iod only a ew, isola ed
dep essions we e illed wi h wa e . A e he onse o ain all, addi ional dep essions we e illed wi h
wa e and he isola ed ponded a eas began o in e connec (second snapsho ). A e 12 hou s and 10
minu es ( hi d snapsho ) ne wo ks 1 and 2 each hei maximum ex en , al hough su ace d ainage o
he channel has no ye been ini ia ed. Al hough he hi d snapsho seems o sugges ha he e we e
small ponded low b idges connec ing he wo ne wo ks, an analysis o s eam aces showed ha
he e was no su ace wa e exchange. A e 12 hou s and 20 minu es ( ou h snapsho ), he i s low
ne wo k s a ed o d ain in o he channel causing he i s inc ease in o al uno (Figu e 9 uppe
panel). 7 hou s and 10 minu es la e ( i h snapsho ), he second low ne wo k was ac i a ed and s a s
o spill in o he channel esul ing in he second apid inc ease in discha ge (Figu e 9 uppe panel).
Consequen ly peak discha ge (a he end o day 217) occu ed when bo h ne wo ks we e connec ed o
he channel. A ha ime, he hi d zone was s ill no connec ed o ei he o he wo o he ne wo ks.
This a ea was cha ac e ized by dep essions, which emained isola ed om he low ne wo ks and
whe e he ponded su ace wa e was immobile du ing e en s. The p ocess o g owing (du ing ain all
e en s) and sh inking ne wo ks (du ing low ecessions) was esponsible o he di e en clockwise
loops ha can be seen in he ela ionship shown in Figu e 9 (uppe panel). In con as o he we ing
p ocess he d ying cycle du ing low ecessions p oceeds much mo e uni o mly (Figu e 9 uppe
panel), because i was no cha ac e ized by he same s epwise h eshold beha io as he we ing
[85]
STUDY 1
p ocess. This dis inc ly di e en beha io o he sys em du ing we ing and d ying esul ed in he
obse ed hys e esis loops. The same beha io was also e iden in he de elopmen o subsu ace and
su ace low h oughou he e en (Figu e 12). Subsu ace low showed a g adual inc ease du ing
we ing o he sys em, whe eas su ace lows we e ini ia ed a dis inc h esholds when speci ic low
ne wo k s a ed o spill in o he channel. In con as ecession o su ace low was much mo e g adual.
Figu e 11: Simula ed ela ionship be ween g oundwa e le el and channel discha ge o he mic o-
opog aphy model. Blue illed ci cles ep esen imes when no su ace d ainage occu s, ed open
ci cles ep esen condi ions when su ace d ainage is being gene a ed, di e en scales a e used on
he x-axis o be e isibili y o hys e e ic beha io du ing low discha ges, he sequence o days 217 o
219, ep esen ing an in ense ain s o m, is depic ed by a line.
[92]
STUDY 1
Figu e 14: Rela ionship be ween discha ge and g oundwa e le el o wo peak low e en s obse ed
o a small ca chmen loca ed in B i ish Colombia, Canada (modi ied a e Fi zge ald e al. (2003)).
4.3 Limi a ions and cons ain s
The i ual sys ems simula ed he e a e a simpli ica ion o complex ield si ua ions. Pea soils a e
a ely homogeneous and hyd aulic conduc i i ies a e usually non-uni o m. Re en ion cha ac e is ics
o pea soils can be hys e e ic. Hence non-linea esponse o eal sys ems may be caused by se e al
easons. Fu he mo e highe e ec i e hyd aulic conduc i i ies in some pea soils (e.g. caused by
p e e en ial low) may so e icien ly d ain a we land ha o a gi en ain all a e su ace ponding
ne e occu s. Signi ican in lows om adjacen hillslopes o om deepe g oundwa e may a ec he
dynamics o low in he pea . Mos o hese aspec s we e in en ionally excluded om his s udy o
highligh he e ec s o he mic o- opog aphy. Al hough his limi s he deg ee o which he esul s can
be gene alized, i p o ides a new insigh in o he p ocess dynamics caused by dis inc su ace mic o-
opog aphy, which is no uncommon o pea - o ming we lands. The s uc u e o mic o- opog aphy,
hyd aulic conduc i i ies o he pea and ain all a es we e aken om a ipa ian en in an
expe imen al wa e shed in Ge many and a e belie ed o be ep esen a i e o o he hummocky
ipa ian we lands in humid clima es. The e o e simula ion esul s can p o ide new insigh s in o he
dynamics o uno gene a ion in such sys ems ha may help o explain o he obse ed non-linea
sys em esponses (e.g. [13]).
5 Conclusions
Hyd ologic sys ems ypically show complex non-linea s eam low esponse o ain all inpu s.
Deciphe ing he p ocesses ha cause he obse ed esponse is usually di icul due o he s ongly

[93]
STUDY 1
non-linea beha io o hyd ologic sys ems [57]. Using physically-based nume ical models as
con olled eplica es o na u al sys ems o conduc " i ual expe imen s" [55] can be a use ul ool o
elucida e indi idual p ocesses and hei in e dependencies (see also Zehe e al. [57]). This app oach
was used he e o in es iga e he e ec s o su ace mic o- opog aphy on uno gene a ion in a i ual
ipa ian we land in a humid clima e. Simula ion esul s e eal complex h eshold p ocesses wi h
s epwise expansions and con ac ions o su ace low ne wo ks ha go e n s eam low gene a ion.
Dis inc ly di e en beha io o he sys em du ing we ing and d ying esul s in a p onounced
clockwise hys e esis in he non-linea ela ionship be ween s eam low and ipa ian g oundwa e
le el ha esembles simila ela ionships obse ed in he ield. Simula ions o di e en mic o-
opog aphies and o a plana e e ence model show clea di e ences in he shape o he non-linea
ela ionship and demons a e how s eam low is mode a ed by he mic o- opog aphy. The plana
model does no show signi ican hys e esis in he s eam low-wa e able ela ionship. Resul s om a
model wi h smalle mean leng h o he mic o- opog aphic s uc u es (1/2 o he o iginal model)
sugges ha o dec easing size o he s uc u es he esponse o he sys em app oaches ha o he
plana model. A compa ison o he model esul s wi h esul s p esen ed by Fi zge ald e al. ([13])
om a ield s udy in a humid ipa ian we land in Canada, sugges s ha he simula ed dynamics migh
p o ide a consis en explana ion o he obse ed beha io o he sys em. We hypo hesize ha he
simula ed hyd ologic dynamics in we lands wi h a de ined mic o- opog aphy can esul in a la ge
ange o subsu ace esidence imes and dynamic mixing be ween su ace and subsu ace wa e o
di e en age and po en ially impac wa e quali y. P elimina y pa icle acking simula ions, which
will be p esen ed in a ollow-up pape , suppo his hypo hesis. To wha deg ee he simula ed
dynamics could p o ide a new amewo k o in e p e he common a iabili y in s eam wa e
chemis y du ing e en s ha is desc ibed in Ki chne 's double pa adox [24] emains o be
in es iga ed. Fu u e wo k will also add ess o wha deg ee simpli ied concep ual ep esen a ions o
su ace s uc u es in nume ical models (e.g. by de ining a ill s o age heigh o la ge model cells)
can mimic he e ec s o he mic o- opog aphy on su ace low and su ace-subsu ace exchange.
Ackonwledgemen s
The au ho s would like o hank he anonymous e iewe s o cons uc i e commen s, which helped o
imp o e he inal manusc ip . This s udy was unded by he Ge man Resea ch Founda ion (DFG,
g an FL 631/6-2). Thei inancial suppo is g ea ly app ecia ed. The au ho s also hank Rob
MacLa en, Young-Jin Pa k, And ea B ook ield and Ed Sudicky a he Uni e si y o Wa e loo, Canada
o hei in aluable help wi h he ins and ou s o he nume ical code Hyd oGeoSphe e.
[94]
STUDY 1
Re e ences
[1] Alewell, C.; Paul, S.; Lischeid, G.; S o ck, F. R. (2007), Co- egula ion o edox p ocesses in
eshwa e we lands as a unc ion o o ganic ma e a ailabili y?, Science o he To al
En i onmen 404(2-3):335–342, doi:10.1016/j.sci o en .2007.11.001.
[2] An oine, M.; Ja aux, M.; Bielde s, C. (2009), Wha indica o s can cap u e uno - ele an
connec i i y p ope ies o he mic o- opog aphy a he plo scale?, Ad ances in Wa e
Resou ces 32(8):1297–1310, doi:10.1016/j.ad wa es.2009.05.006.
[3] Bishop, K.; Seibe , J.; Köhle , S.; Laudon, H. (2004), Resol ing he double pa adox o apidly
mobilized old wa e wi h highly a iable esponses in uno chemis y, Hyd ol. P ocess 18185–
189.
[4] Bishop, K.H. ,(1991) ,Episodic inc ease in s eam acidi y, ca chmen low pa hways and
hyd og aph sepe a ion, Ph.D. hesis, Camb idge Uni e si y, Camb idge, UK.
[5] B an i eun, B. A.; Roule , N. T. (1998), The base low and s o m low hyd ology o a p ecamb ian
shield headwa e pea land, Hyd ological P ocesses 12(1):57–72.
[6] B ook ield, A. E.; Sudicky, E. A.; Pa k, Y. J.; Conan J ., B. (2009), The mal anspo modelling in
a ully in eg a ed su ace/subsu ace amewo k, Hyd ological P ocesses(23):2150–2164,
10.1002/hyp.7282.
[7] Ca le, S.F.; Fogg, G.E. (1996), T ansi ion P obabili y-Based Indica o Geos a is ics, Ma hema ical
Geology 28(4):453–476.
[8] De i o, K. J.; Hill, A. R. (1997), Sulpha e dynamics in ela ion o g oundwa e -su ace wa e
in e ac ions in headwa e we lands o he sou he n Canadian Shield, Hyd ological P ocesses
11(5):485–500.
[9] Dunne, T.; Zhang, W.; Aub y, B. F. (1991), E ec s o ain all, ege a ion, and mic o opog aphy on
in il a ion and uno , Wa e Resou ces Resea ch 27(9):2271–2285.
[10] Du al, T. P.; Hill, A. R. (2006), In luence o s eam bank seepage du ing low- low condi ions on
ipa ian zone hyd ology, Wa e Resou ces Resea ch 42(10):10425,
doi:10.1029/2006WR004861.
[11] Es e es, M.; Fauche , X.; Galle, S.; Vauclin, M. (2000), O e land low and in il a ion modelling
o small plo s du ing uns eady ain: nume ical esul s e sus obse ed alues, Jou nal o
Hyd ology 228(3-4):265–282.
[12] Fiedle , F. R.; Rami ez, J. A. (2000), A nume ical me hod o simula ing discon inuous shallow
low o e an in il a ing su ace, In e na ional jou nal o nume ical me hods in luids 32(2):219–
239.
[13] Fi zge ald, D. F.; P ice, J. S.; Gibson, J. J. (2003), Hillslope-swamp in e ac ions and low
pa hways in a hype ma i ime ain o es , B i ish Columbia, Hyd ological P ocesses 17(15):3005–
3022, 10.1002/hyp.1279.
[14] Fleckens ein, J.H.; Niswonge , R.G.; Fogg, G.E. (2006), Ri e -Aqui e In e ac ions, Geologic
He e ogenei y, and Low-Flow Managmen , G oundwa e 44(6):837–852.
[15] F ase , C. J.D.; Roule , N. T.; Moo e, T. R. (2001), Hyd ology and dissol ed o ganic ca bon
biogeochemis y in an omb o ophic bog, Hyd ological P ocesses 15(16):3151–3166,
10.1002/hyp.322.
[16] F ei, S.; Fleckens ein, J. H.; Kolle , S. J.; Maxwell, R. M. (2009), Pa e ns and dynamics o i e –
aqui e exchange wi h a iably-sa u a ed low using a ully-coupled model, Jou nal o Hyd ology
375(3-4):383–393.
[17] Gibson, J. J.; P ice, J. S.; A a ena, R.; Fi zge ald, D. F.; Maloney, D. (2000), Runo gene a ion
in a hype ma i ime bog- o es upland, Hyd ol. P ocess 142711–2730, doi: 10.1002/hyp.1279.
[18] Gilliam, J. W. (1994), Ripa ian we lands and wa e quali y, Jou nal o En i onmen al Quali y
23(5):896.
[19] Hauck, Ane e ,(1999) ,Hyd ological Cha ak e iza ion o he Lehs enbach Ca chmen ,
unpuplished Diploma Thesis, Depa men o Ecological Modelling, Uni e si y o Bay eu h,
Bay eu h.
[95]
STUDY 1
[20] Hopp, L.; McDonnell, J. J. (2009), Connec i i y a he hillslope scale: Iden i ying in e ac ions
be ween s o m size, bed ock pe meabili y, slope angle and soil dep h, Jou nal o Hyd ology,
doi:10.1016/j.jhyd ol.2009.07.047.
[21] Jones, J. P.; Sudicky, E. A.; B ook ield, A. E.; Pa k, Y. J. (2006), An assessmen o he ace -
based app oach o quan i ying g oundwa e con ibu ions o s eam low, Wa e Resou ces
Resea ch 42, W02407, doi:10.1029/2005WR004130.
[22] Jones, J. P.; Sudicky, E. A.; McLa en, R. G. (2008), Applica ion o a ully-in eg a ed su ace-
subsu ace low model a he wa e shed-scale: A case s udy, Wa e Resou ces Resea ch 44,
W03407, doi:10.1029/WR005603.
[23] Kendall, K. A.; Shanley, J. B.; McDonnell, J. J. (1999), A hyd ome ic and geochemical app oach
o es he ansmissi i y eedback hypo hesis du ing snowmel , Jou nal o Hyd ology 219(3-
4):188–205, doi:10.1016/S0022-1694(99)00059-1.
[24] Ki chne , J. W. (2003), A double pa adox in ca chmen hyd ology and geochemis y, Hyd ological
P ocesses 17(4):871–874.
[25] Kolle , J.S.; Maxwell, R.M. (2006), In eg a ed su ace-g oundwa e low modeling: A ee-su ace
o e land low bounda y condi ion in a pa allel g oundwa e low model, Ad ances in Wa e
Resou ces 29(7):945–958.
[26] K use, J.; Lenna z, B.; Leinwebe , P. (2008), A modi ied me hod o measu ing sa u a ed
hyd aulic conduc i i y and aniso opy o en pea samples, We lands 28(2):527–531.
[27] K æ ne , J.; Klø e, B. (2008), Gene a ion and egula ion o summe uno in a bo eal la en,
Jou nal o Hyd ology(360):15–30, doi:10.1016/j.jhyd ol.2008.07.009.
[28] Laudon, H.; Seibe , J.; Kohle , S.; Bishop, K. (2004), Hyd ological low pa hs du ing snowmel :
cong uence be ween hyd ome ic measu emen s and oxygen 18 in mel wa e , soil wa e , and
uno , Wa e Resou ces Resea ch 40(3):W03102, doi:10.1029/2003WR002455.
[29] Lee, S.-Y.; Ca le, S.F.; Fogg, G.E. (2007), Geologic he e ogenei y and a compa ison o wo
geos a is ical models: Sequen ial Gaussian and ansi ion p obabili y-based geos a is ical
simula ion, Ad ances in Wa e Resou ces 30(9):1914–1932.
[30] Li, Q.; Unge , A. J.A.; Sudicky, E. A.; Kassenaa , D.; Wexle , E. J.; Shikaze, S. (2008),
Simula ing he mul i-seasonal esponse o a la ge-scale wa e shed wi h a 3D physically-based
hyd ologic model, Jou nal o Hyd ology(357):317–336, doi:10.1016/j.jhyd ol.2008.05.024.
[31] Lischeid, G.; Lange, H.; Mo i z, K.; Bü che , H. (2004), Dynamics o uno and uno chemis y
a he Lehs enbach and S eink eu z ca chmen , In: Ma zne , E. (ed ). Biogeochemis y o
Fo es ed Ca chmen s a. Changing En i onmen : A. Ge man Case S udy, Ecological S udies,
Vol.172 ,Sp inge , Heidelbe g, pp. 399–436.
[32] Lischeid, G.; Bi e sohl, J. (2008), T acing biogeochemical p ocesses in s eam wa e and
g oundwa e using non-linea s a is ics, Jou nal o Hyd ology 357(1-2):11–28,
doi:10.1016/j.jhyd ol.2008.03.013.
[33] Lischeid, G.; Kolb, A.; Alewell, C. (2002), Appa en ansla o y low in g oundwa e echa ge and
uno gene a ion, Jou nal o Hyd ology 265(1-4):195–211.
[34] Lischeid, G.; Kolb, A.; Alewell, C.; Paul, S. (2007), Impac o edox and anspo p ocesses in a
ipa ian we land on s eam wa e quali y in he Fich elgebi ge egion, sou he n Ge many,
Hyd ological P ocesses 21(1):123–132, 10.1002/hyp.6227.
[35] McLa en, R.G. (2008), GRID BUILDER - A p e-p ocesso o 2D, iangula elemen , ini e-
elemen p og ams, G oundwa e Simula ions G oup, Uni e si y o Wa e loo, Wa e loo.
[36] Me cal e, R. A.; Bu le, J. M. (1999), Semi-dis ibu ed wa e balance dynamics in a small bo eal
o es basin, Jou nal o Hyd ology 226(1-2):66–87.
[37] Moldan, F.; W igh , R. F. (1998), Episodic beha iou o ni a e in uno du ing six yea s o
ni ogen addi ion o he NITREX ca chmen a Gå dsjön, Sweden, En i onmen al Pollu ion
102(1S1):439–444, 10.1016/S0269-7491(98)80066-3.
[38] Molena , J.; Gascuel-Odoux, C. (2002), Modelling low and ni a e anspo in g oundwa e o
he p edic ion o wa e a el imes and o consequences o land use e olu ion on wa e quali y,
Hyd ological P ocesses 16(2):479–492, 10.1002/hyp.328.
[96]
STUDY 1
[39] P ice, J. S.; McLa en, R. G.; Rudolph, D. L. (2009), Landscape es o a ion a e oil sands mining:
concep ual design and hyd ological modelling o en econs uc ion, In e na ional Jou nal o
Mining, Reclama ion and En i onmen 99999(1):1–15.
[40] Qu, Y.; Du y, C. J. (2007), A semidisc e e ini e olume o mula ion o mul ip ocess wa e shed
simula ion, Wa e Resou ces Resea ch 43(8):W08419, doi:10.1029/2006WR005752.
[41] Rice, K. C.; B icke , O. P. (1995), Seasonal cycles o dissol ed cons i uen s in s eamwa e in
wo o es ed ca chmen s in he mid-A lan ic egion o he eas e n USA, Jou nal o Hyd ology
170(1-4):137–158, doi:10.1016/0022-1694(95)92713-N.
[42] Schlo zhaue , S. M.; P ice, J. S. (1999), Soil wa e low dynamics in a managed cu o e pea
ield, Quebec: Field and labo a o y in es iga ions, Wa e Resou ces Resea ch 35(12):3675–
3683.
[43] Seibe , J.; Bishop, K.; Rodhe, A.; McDonnell, J. J. (2003), G oundwa e dynamics along a
hillslope: A es o he s eady s a e hypo hesis, Wa e Resou . Res 39(1):1014,
doi:10.1029/2002WR001404, 2003.
[44] Seibe , J.; G abs, T.; Köhle , S.; Laudon, H.; Win e dahl, M.; Bishop, K. (2009), Linking soil–and
s eam-wa e chemis y based on a ipa ian low-concen a ion in eg a ion model, Hyd ology and
Ea h Sys em Sciences, 13:2287-2297.
[45] Shen, H. W.; Julien, P. Y. (1993), E osion and sedimen anspo , Handbook o Hyd ology.12.1-
12.61.
[46] Spence, C. (2007), On he ela ion be ween dynamic s o age and uno : A discussion on
h esholds, e iciency, and unc ion, Wa e Resou ces Resea ch 43(12):W12416.
[47] Spence, C.; Woo, M. (2003), Hyd ology o suba c ic Canadian shield: soil- illed alleys, Jou nal
o Hyd ology 279(1-4):151–166.
[48] Sudicky, E. A.; Jones, J. P.; Pa k, Y. J.; B ook ield, A. E.; Colau i, D. (2008), Simula ing complex
low and anspo dynamics in an in eg a ed su ace-subsu ace modeling amewo k,
Geosciences Jou nal 12(2):107–122.
[49] Taylo , C. H. (1997), Runo p ocesses in empe a e headwa e we lands, Ecology o We lands
and Associa ed Sys ems.(SK Majumda , EW Mille and FJ B enne , eds.) pp.169–181.
[50] The ien, R.; McLa en, R.G.; Sudicky, E.A.; Panday, S.M. (2008), Hyd oGeoSphe e A Th ee-
dimensional Nume ical Model Desc ibing Fully-in eg a ed Subsu ace and Su ace Flow and
Solu e T anspo (Manual), G oundwa e Simula ions G oup, Uni e si y o Wa e loo.
[51] T omp- an Mee eld, H. J.; McDonnell, J. J. (2006), Th eshold ela ions in subsu ace s o m low:
1. A 147-s o m analysis o he Panola hillslope, Wa e Resou ces Resea ch 42(2):W02410.
[52] T omp- an Mee eld, H. J.; McDonnell, J. J. (2006), Th eshold ela ions in subsu ace s o m low:
2. The ill and spill hypo hesis, Wa e Resou ces Resea ch 42(2):W02411.
[53] Vidon, P. G.F.; Hill, A. R. (2004), Landscape con ols on ni a e emo al in s eam ipa ian zones,
Wa e Resou ces Resea ch 40(3):W03201, doi:10.1029/2003WR002473.
[54] Vogel, H. J.; Ho mann, H.; Ro h, K. (2005), S udies o c ack dynamics in clay soil:: I.
Expe imen al me hods, esul s, and mo phological quan i ica ion, Geode ma 125(3-4):203–211.
[55] Weile , M.; McDonnell, J. (2004), Vi ual expe imen s: a new app oach o imp o ing p ocess
concep ualiza ion in hillslope hyd ology, Jou nal o Hyd ology 285(1-4):3–18.
[56] Weissmann, G.S. ,(1999) ,Towa d new models o subsu ace he e ogenei y: An allu ial an
sequence s a ig aphic amewo k wi h ansi ion p obabili y geos a is ics, unpublished Ph.D.
hesis, Hyd ologic Sciences G adua e G oup, Uni e s iy o Cali o nia, Da is.
[57] Zehe, E.; Becke , R.; Bá dossy, A.; Pla e, E. (2005), Unce ain y o simula ed ca chmen uno
esponse in he p esence o h eshold p ocesses: Role o ini ial soil mois u e and p ecipi a ion,
Jou nal o Hyd ology 315(1-4):183–202.
[58] Zehe, E.; Blöschl, G. (2004), P edic abili y o hyd ologic esponse a he plo and ca chmen
scales: Role o ini ial condi ions, Wa e Resou . Res 40(10):W10202.
[59] Zehe, E.; Elsenbee , H.; Lindenmaie , F.; Schulz, K.; Blöschl, G. (2007), Pa e ns o p edic abili y
in hyd ological h eshold sys ems, Wa e Resou . Res 43(7):W07434.
[60] Zehe, E.; Si apalan, M. (2009), Th eshold beha iou in hyd ological sys ems as (human) geo-
ecosys ems: mani es a ions, con ols, implica ions, Hyd ol. Ea h Sys . Sci 131273–1297.
[97]
STUDY 2
S udy 2
Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in we land
sys ems – a i ual modeling expe imen .
By S en F ei, Klaus-Holge Kno , S e an Pei e and Jan H. Fleckens ein
Published in Jou nal o Geophysical Resea ch Le e s - Biogeosiences (in p ess)

[98]
[99]
STUDY 2
Published in Jou nal o Geophysical Resea ch Le e s – Biogeosiences (in p ess)
Su ace mic o- opog aphy causes ho spo s o biogeochemical ac i i y in we land
sys ems – a i ual modeling expe imen .
F ei1, S., Kno 1, K.H., Pei e 1, S., and Fleckens ein2. J.H.
1 Depa men o Hyd ology, Uni e si y o Bay eu h, Ge many
2 Depa men Hyd ogeology, Helmhol z-Cen e o En i onmen al Resea ch – UFZ, Ge many
Abs ac
We lands p o ide impo an ecohyd ological se ices by egula ing luxes o nu ien s and pollu an s
o ecei ing wa e s, which can in u n mi iga e ad e se e ec s on wa e quali y. Tu no e o edox-
sensi i e solu es in we lands has been shown o ake place in dis inc spa ial and empo al pa e ns,
commonly e e ed o as ho spo s and ho momen s. Despi e he impo ance o such pa e ns o
solu e luxes he mechanis ic unde s anding o hei o ma ion is s ill weak and hei exis ence is o en
explained by a ia ions in soil p ope ies and di usi e anspo only. He e we show ha su ace
mic o- opog aphy in we lands can cause he o ma ion o biogeochemical ho spo s solely by he
ad ec i e edis ibu ion o in il a ing wa e as a esul o complex subsu ace low pa e ns. Su ace
and subsu ace lows a e simula ed o an idealized sec ion o a ipa ian we land using a ully
in eg a ed nume ical code o coupled su ace-subsu ace sys ems. Biogeochemical p ocesses and
anspo along ad ec i e subsu ace low pa hs a e simula ed kine ically using he biogeochemical
code PHREEQC. Dis inc pa e ns o biogeochemical ac i i y (exp essed as eac ion a es) de elop in
esponse o mic o- opog aphy induced subsu ace low pa e ns. Simula ed e ical po e wa e
p o iles o a ious edox-sensi i e species esemble p o iles obse ed in he ield. This mechanis ic
explana ion o ho -spo o ma ion complemen s he mo e s a ic explana ions ha ela e ho spo s
solely o spa ial a iabili y in soil cha ac e is ics and can accoun o spa ial as well as empo al
a iabili y o biogeochemical ac i i y, which is needed o assess u u e changes in he biogeochemical
u no e o we land sys ems.
[100]
STUDY 2
1 In oduc ion
We lands p o ide impo an ecohyd ological se ices in many moun ainous headwa e ca chmen s.
They s o e signi ican amoun s o ca bon as pea , and ac as e ec i e nu ien sinks e.g. o sul u ,
phospho us and ni ogen [Le Kellogg and B idgham, 2003; Paul e al., 2006; Tauchni z e al., 2010].
Redox condi ions and he co esponding biogeochemical p ocesses in hese we lands la gely con ol
he sou ce and sink unc ions o pea -soil domina ed ca chmen s [Bishop e al., 2004; Lischeid e al.,
2007]. P ocess ac i i ies in such we lands a e spa ially nonuni o m, hough, and ha e been ound o
o m dis inc ho spo s [Jacks and No s öm, 2004], i. e. a eas o pa ches ha show disp opo ionally
high eac ion a es ela i e o he su ounding a eas [McClain e al., 2003; Mo is and Wadding on,
2011]. Such ho spo s a e no easily iden i ied in he sca e o spa io empo al da ase s and hence hei
ele ance o ne ma e u no e is assumed o be unde es ima ed [Richa dson e al., 2007; McClain
e al., 2003; Vidon e al., 2010]. Va ious s udies ha e obse ed la ge a ia ions in he spa ial
dis ibu ion o edox-sensi i e solu es wi hin we land soils [Jacks and No s öm, 2004; McMahon
and Chapelle, 2008] on he scale o ansec s (10-50m) [Jacks and No s öm, 2004] as well as in he
me e and sub-me e ange [Kno and Blodau, 2009; Mi chell and B an i eun, 2005; Wachinge e
al., 2000]. I seems ob ious ha complex anspo and ans o ma ion p ocesses wi hin he
subsu ace a e main d i e s o he obse ed spa ial he e ogenei y in solu e concen a ions. Al hough
s udies ha e poin ed a po en ial e ec s o subsu ace low dynamics in we lands on solu e
concen a ions, e.g. by enhanced mixing due o hyd aulic g adien e e sals [Ree e e al., 2006] and
he o ma ion o ho spo s has concep ually been linked o anspo p ocesses [McClain e al., 2003]
anspo and biogeochemical ans o ma ions a e a ely combined mechanis ically o explain such
phenomena. Recen s udies in we lands ha e mainly a ibu ed he o ma ion o ho spo s o la e al
a ia ions in local physico-chemical a iables such as soil ex u e, composi ion, mois u e o
empe a u e [B uland and Richa dson, 2005; Mo is and Wadding on, 2011] o he local a ailabili y
o ce ain eac an s such as ni a e o DOC [B uland e al., 2006]. Di e ences in hese p ope ies may
e.g. a ise om di e en deg ees o pea decomposi ion, pea compac ion, ege a ion o su ace mic o-
opog aphy [Ga ni and Kenne h, 1990; Cheng e al., 2011; B uland and Richa dson, 2005].
This pe spec i e, howe e , does no conside ha mic obial p ocesses a e dynamic and dependen on
a iable hyd ologic and biogeochemical bounda y condi ions. The close links be ween he
mechanisms con olling biogeochemical ac i i y in we lands and he hyd ological p ocesses occu ing
wi hin he we land ha e been highligh ed in se e al s udies [Mo is and Wadding on, 2011; Mi chell
and B an i eun, 2005]. Field s udies [Kno e al., 2009; Kno and Blodau, 2009] demons a ed a
apid change o p edominan edox p ocesses (i.e. i on(III)-, sul a e educ ion and me hanogenesis) in
a we land exposed o luc ua ions o hyd ological bounda y condi ions du ing manipula ion o he
wa e le el. We lands in moun ainous ca chmen s a e o en cha ac e ized by apidly luc ua ing bu
[101]
STUDY 2
shallow wa e le els [De i o and Hill, 1997; Lischeid e al., 2007]. Such hyd ological condi ions
acili a e as low componen s like sa u a ion excess o e land low and shallow subsu ace lows
[F ei e al., 2010; Holden and Bu , 2003]. The dynamics o hese low componen s a e impo an
con ols on mobiliza ion o dissol ed solu es (e.g. dissol ed o ganic ca bon o ni a e) om we lands
[Alewell e al., 2007; Lischeid e al., 2007; Hin on e al., 1998; Dosskey and Be sch, 1994] bu hei
e ec on he biogeochemical p ocesses and dis ibu ion o edox-sensi i e solu es is s ill poo ly
unde s ood and a ely add essed [Shabaga and Hill, 2010]. Pa ly his is because i is nea ly
impossible o di ec ly in es iga e and cha ac e ize he complex, dynamic subsu ace hyd ology in he
ield. The e o e he in e p e a ion o ield obse a ions (e.g. dep h p o iles o edox-sensi i e solu es)
may be poo ly cons ained, e.g. i biogeochemical u no e a es a e calcula ed based on he
assump ion ha esupply o dissol ed elec on accep o s/dono s wi hin ipa ian we lands is only
di usion limi ed [Bee and Blodau, 2007; Clymo and B yan , 2008]. This simpli ica ion may hold
ue o some si es [Bee and Blodau, 2007] and o de ined lab incuba ions [Kno and Blodau,
2009], bu i neglec s ha anspo and u no e o edox-sensi i e solu es a many na u al si es
occu s wi hin a complex, h ee-dimensional (3D) subsu ace low ield ha is subjec o a iable
bounda y condi ions. This esul s in dis inc low pa hs along which biogeochemical eac ions can
occu , con olled by he indi idual kine ics o each p ocess [Kno and Blodau, 2009; Hill, 2000;
B o elli e al., 2011]. An imp o ed mechanis ic model o he o ma ion and occu ence o
biogeochemical ho spo s he e o e needs o accoun o low and anspo p ocesses and how hey
a e a ec ed by changes in hyd ologic bounda y condi ions. This is o pa icula impo ance i such a
model is used o assess he e ec s o clima e change whe e induced shi s in he equency o in ense
ains o ms o ex ended d ough s [Hun ing on, 2006] ha e he po en ial o signi ican ly al e he
bounda y condi ions wi hin we lands.
Vi ual expe imen s [Weile and McDonnell, 2004, 2006] ha e p o en o be a sui able ool o
in es iga e complex hyd ologic p ocesses and eedback mechanisms be ween hyd ology and
biogeochemis y [F ei e al., 2010; Boano e al., 2010; Jakobsen, 2007]. In his s udy, we use i ual
modeling expe imen s o in es iga e how complex subsu ace low pa e ns induced by su ace mic o-
opog aphy a ec he subsu ace anspo o edox-sensi i e solu es and he esul ing spa ial
dis ibu ion o biogeochemical p ocess ac i i ies wi hin a hummocky we land. We es he hypo hesis
ha he complex subsu ace low- ield c ea es biogeochemical condi ions in he subsu ace ha
acili a e he o ma ion o local p ocess ho spo s e en in soils wi h uni o m soil p ope ies. To
add ess his objec i e, he nume ical simula ions o complex su ace and subsu ace low p ocesses in
he hypo he ical sec ion o he ipa ian we land wi h p onounced mic o- opog aphy (hollows and
hummocks) as desc ibed by F ei e al., [2010], is combined wi h ad ec i e pa icle acking and
mul i-species biogeochemical simula ions in a sequen ial s eam ube app oach. The main edox
eac ions ypically ound in pea - o ming we lands a e simula ed along indi idual subsu ace low
[108]
STUDY 2
Figu e 2: Typical oxygen dep h p o ile based on obse a ions om a ipa ian we land si e in he
Lehs enbach ca chmen . P o ile was used o assign oxygen bounda y condi ions o he di e en
PHREEQC sub-sec ion simula ions based on ansien low model ou pu .
All ep esen ed educ i e p ocesses (de ailed in o ma ion a e gi en in he nex pa ag aph) a e ea ed
as eac ions ca alyzed by mic oo ganisms, compa able o e.g. he p ocess model o me hane
p oduc ion in we lands as shown by Sege s and Kengen [1998]. These ypes o eac ions depend on
he p esence o (a) an adequa e elec on accep o (e.g. oxygen, ni a e, i on(III) o sul a e) and (b) a
sou ce o labile ca bon ha is a ailable o mic oo ganisms. As a simpli ica ion o educe model
complexi y, we conside ed he elec on accep o as he limi ing ac o o he p esence o he
indi idual ca alyzed edox- eac ions. We hink ha his is a easonable app oxima ion, since he
supply o labile ca bon (e.g. ace a e) may be assumed o be coupled o he o ganic ma e
mine aliza ion a e, as usually no in e media es (e.g. om e men a ion) accumula e [Sege s and
Kengen, 1998]. The biogeochemical simula ions we e hus pe o med based on ha concep ,
implemen ing an unlimi ed ca bon sou ce as BC o all sub-sec ion simula ions and limi ing p ocess
a es solely by hei kine ic pa ame e s. By dynamically assigning he biogeochemical bounda y
condi ions o each indi idual sub-sec ion he whole sequence o sub-sec ion simula ions o one
subsu ace low pa h, can be iewed as a con inuous simula ion o he edox-chemical e olu ion o a
small wa e pa cel ha ca ies dissol ed edox-sensi i e solu es and is anspo ed along ha speci ic
low pa h.

[109]
STUDY 2
Implemen ed Reac ions and Kine ics
Fo each sub-sec ion, PHREEQC simula es edox p ocesses as kine ic eac ions based on he assigned
bounda y (BCi) and ini ial condi ions (ICglobal/FCi-1). Implemen ed p ocesses a e shown in Table 2. All
educ ion p ocesses a e o mula ed based on Monod kine ic eac ions acco ding o Equa ion 4:
kks
k
k
k
kCK
C
d
dC
R

,
max,

(4)
He e Rk [ML-3 T
-1] is he kine ic a e o he co esponding educ ion eac ion k (

4,3,2,1k)
acco ding o Table 2. kmax,

[ML-3T-1] ep esen s he maximal speci ic g ow h a e ( o k=1 ae obic
espi a ion, k=2 de-ni i ica ion, k=3 i on(III)- educ ion, k=4: sul a e educ ion) and Ks,k [ML-3]
ep esen s he subs a e sa u a ion cons an (i.e. subs a e concen a ion o k=1 oxygen, k=2 ni a e,
k=3 i on(III), k=4 sul a e a hal kmax,

). Ck [ML-3] is he co esponding concen a ion o he elec on
accep o ( o k=1: oxygen, k=2: ni a e, k=3: i on(III); k=4: sul a e). Monod kine ic coe icien s (
kmax,

and Ks,k) o all educ ion p ocesses a e based on alues epo ed o biodeg ada ion o o ganic
chemicals in aqui e s [Appelo and Pos ma, 2005; Bekins e al., 1998; Schi me e al., 1999;
MacQua ie e al., 1990; Ecke and Appelo, 2002; Kelly e al., 1996; Goldsmi h and Balde son,
1988] and we e la e modi ied and adjus ed as pa o he calib a ion p ocess. Simula ed dep h p o iles
o edox-sensi i e compounds (ni a e, sul a e and i on(II)) we e calib a ed by sys ema ic a ia ion o
he Monod coe icien s o bes i obse ed da a aken a he s udy si e [Kno and Blodau, 2009;
Kno e al., 2009]. Calib a ed Monod coe icien s a e lis ed in Table 2. Fo all p ocesses whe e
o ganic ca bon is being decomposed, o ganically bound ni ogen is being eleased acco ding o he
Red ield a io [Red ield, 1934]. Oxida ion p ocesses (k=5 i on(II) oxida ion, k=6 ni i ica ion, k=7
ae obic sul ide oxida ion and k=8 anae obic sul ide oxida ion) we e o mula ed using highe o de
eac ion kine ics as lis ed in Table 2.
[110]
STUDY 2
Table 2: Implemen ed p ocesses and he equi alen eac ion speci ic kine ic a e. Reduc ion
p ocesses a e o mula ed based on Monod ype eac ion kine ics.
P ocess Ra e Coe icien s Re e ence
ae obic
espi a ion
acco ding o equa ion 4 1max,

= 1.6 x 10-9
mol/Ls
Ks,1 = 2.9 x 10-6 mol/L
modi ied and calib a ed
a e [Appelo and
Pos ma, 2005; Bekins
e al., 1998; Schi me
e al., 1999;
MacQua ie e al.,
1990; Ecke and
Appelo, 2002; Kelly e
al., 1996; Goldsmi h
and Balde son, 1988]
deni i ica ion acco ding o equa ion 4 2max,

= 1.06 x 10-9
mol/Ls
Ks,2 = 2.0 x 10-6 mol/L
modi ied and calib a ed
a e [Appelo and
Pos ma, 2005; Bekins
e al., 1998; Schi me
e al., 1999;
MacQua ie e al.,
1990; Ecke and
Appelo, 2002; Kelly e
al., 1996; Goldsmi h
and Balde son, 1988]
i on(III)
educ ion
acco ding o equa ion 4 3max,

= 1.5 x 10-12
mol/Ls
Ks,3 = 2.94 x 10-6 mol/L
modi ied and calib a ed
a e [Appelo and
Pos ma, 2005; Bekins
e al., 1998; Schi me
e al., 1999;
MacQua ie e al.,
1990; Ecke and
Appelo, 2002; Kelly e
al., 1996; Goldsmi h
and Balde son, 1988]
sul a e educ ion acco ding o equa ion 4 3max,

= 0.5 x 10-10
mol/Ls
Ks,4 = 2.5 x 10-6 mol/L
modi ied and calib a ed
a e [Appelo and
Pos ma, 2005; Bekins
e al., 1998; Schi me
e al., 1999;
MacQua ie e al.,
1990; Ecke and
Appelo, 2002; Kelly e
al., 1996; Goldsmi h
and Balde son, 1988]
i on(II)
oxida ion )()()( 2
25
5
5
  FecOpOHaA
d
dC
R A5 = 8 x 10 13 min-1a m-1 [Appelo and Pos ma,
2005; S umm and
Mo gan, 1995]
ammonium
oxida ion )()( 246
6
6OcNHcA
d
dC
R  A6 =5 x 106 (mol/L)-1 a-1
[Billen, 1982; an
Cappellen and Wang,
1996]
ae obic sul ide
oxida ion )()( 27
7
7OcHScA
d
dC
R  A7 = 1.6 x 105 (mol/L)-1
a-1
[Mille o e al., 1987;
an Cappellen and
Wang, 1996]
anae obic
sul ide oxida io
oxida ion
)()( 3
8
8
8
  FecHScA
d
dC
R A8 = 8 x 103 (mol/L)-1 a-
1
[Pyzik and Somme ,
1981; an Cappellen
and Wang, 1996]
[111]
STUDY 2
In edox con olled sys ems like we lands, educ ion p ocesses can be expec ed o occu sequen ially
due o he modynamic easons (e.g. [Ach nich e al., 1995]). Oxygen is used as p ima y elec on
accep o , and a e deple ion ni a e, subsequen ly i on(III) and inally sul a e a e being educed.
Fu he elec on accep o s, such as manganese [Nealson and Sa a ini, 1994] o o ganic molecules
[Lo ley e al., 1996] we e no conside ed in his s udy. To make su e ha he educ ion p ocesses
p oceed sequen ially in he biogeochemical simula ions, speci ic edox condi ions we e de ined.
These condi ions a e ep esen ed by c i ical concen a ions o edox-sensi i e solu es which con ol
whe he a edox p ocess can be ini ia ed o no . C i ical concen a ions Cc i [ML-3] o oxygen, ni a e
and I on(III) we e de i ed based on obse ed dep h p o iles o edox-sensi i e compounds [Kno
and Blodau, 2009; Kno e al., 2009; Es op-A agonés and Blodau, 2012] Fo example, he c i ical
concen a ion o oxygen Cc i O2 is he esidual concen a ion o oxygen unde which deni i ica ion is
being ini ia ed, which was es ima ed om obse ed dep h p o iles and ield da a. C i ical
concen a ions o oxygen, ni a e and i on(III) a e lis ed in Table 3. The ows o Table 3 ep esen he
condi ions unde which he di e en educ ion p ocesses a e ini ia ed. En ies mus be ead ow-wise,
whe e en ies “>0” mean ha he co esponding edox-sensi i e eac an (column) mus be p esen
and “-“ means ha his p ocess is independen om he p esence o his speci ic compound. Fo
example i on(III) educ ion in he biogeochemical simula ion is ini ia ed i : (1) Dissol ed oxygen
concen a ions all below Cc i O2; (2) Mos o he ni a e is al eady deple ed whe e concen a ions o
ni a e all below Cc i NO3; (3) The elec on accep o i on(III) is a ailable. In e als o he ac i a ion o
educ ion p ocesses a e o e lapping which means ha mul iple p ocesses can occu simul aneously in
he simula ion; his was also obse ed in labo a o y and unde ield condi ions [Kno and Blodau,
2009; Kno e al., 2009].
Table 3:. C i ical concen a ions which a e con olling he sequen ial ini ializa ion o he edox
sequence. Values we e de i ed om ield obse a ions. Table mus be ead ow wise (e.g.
deni i ica ion is ini ia ed i 1. oxygen con en s d op below Cc i de i ed o oxygen and 2. i ni a e is
p esen ).
Oxygen Ni a e i on(III) sul a e
ae obic espi a ion >0 - - -
deni i ica ion < Cc i O2 >0 - -
i on(III) educ ion < Cc i O2 < Cc i NO3 >0 -
sul a e educ ion < Cc i O2 < Cc i NO3 < Cc i Fe3+ >0
Cc i O2 5.0 x 10-6 mol/L
Cc i NO3 4.0 x 10-7 mol/L
Cc i Fe3+ 5.0 x 10-6 mol/L
[112]
STUDY 2
Simpli ying Model Assump ions
To educe he complexi y o he ep esen ed sys em and o main ain a ac able model he ollowing
simpli ying assump ions we e made: (1) Soil speci ic pa ame e s (sa u a ed hyd aulic conduc i i y,
po osi y and e en ion cu es o a iably sa u a ed low) a e uni o m wi hin he model domain o
sepa a e he e ec s o mic o- opog aphy on subsu ace low dynamics om possible impac s o
he e ogenei y. (2) By simula ing biogeochemical eac ions along isola ed subsu ace low pa hs, i is
assumed ha he e is no in e ac ion be ween di e en low pa hs whe e wa e and/o solu es a e
exchanged due o hyd odynamic dispe sion (mechanic dispe sion + di usion). (3) Subsu ace low
pa hs a e de i ed based on a ansien low ield esul ing om yea ly model uns. Pa icle acking is
pe o med o a 25 yea pe iod by epea ing he yea ly ou pu o he low model wen y- i e imes.
This assumes ha he e a e no in e -annual changes in he basic p ope ies o he subsu ace low ield
(dis ibu ion o low pa hs and RTs). (4) In he biogeochemical simula ions a ailabili y o DOC, as
he p ima y elec on sou ce o mic obially ca alyzed eac ions (ae obic espi a ion, deni i ica ion,
i on(III)- and sul a e educ ion) was assumed o be non-limi ing. (5) E ec s o ege a ion and i s
po en ial in luence on subsu ace low and edox p ocesses, i.e. due o oo espi a ion o exuda ion
and e apo anspi a ion, a e no conside ed. (6) I on(III) species in he biogeochemical simula ions a e
ea ed as solu es only, which a e ad ec i ely anspo ed wi hin he subsu ace domain and no as
immobile solids bound o he pea ma ix. (7) In he biogeochemical simula ions, bioa ailabili y o all
in ol ed species is no a ec ed by e.g. complexa ion wi h DOC.
[113]
STUDY 2
3 Resul s
3.1 Subsu ace low pa e ns
Subsu ace low pa hs o he wo mic o- opog aphy models and he plana e e ence model a e
shown in Figu e 3 (A-D). Fo he model wi h a mean-leng h o 0.5m (ml-0.5m) low-pa hs a e shown
o he en i e 3D model domain (A) as well as o he 2D ansec loca ed acc oss he cen e o he 3D
model domain (dashed line in A). The 2D low ields o he ansec s ep esen p ojec ions o he 3D
low pa hs in o a 2D plane ( low componen s in he y di ec ions a e neglec ed). In con as o he
plana e e ence model, bo h mic o- opog aphy models showed complex dis ibu ed subsu ace low
pa hs whe e coexis ing shallow and deep low cells de eloped in 3D. This is a common phenomenon
caused by opog aphy and was i s desc ibed by To h, [1962] o egional g oundwa e low sys ems
bu can be ound o lows in sys ems wi h p onounced opog aphy o e a ange o scales [Wö man e
al., 2006; S onedahl e al., 2010]. Shallow low cells a e mos p onounced o he model wi h a mean
leng h o he su ace s uc u es o 0.5m (B) and a e associa ed wi h he dominan su ace s uc u es
(la ges hummocks). A eas cha ac e ized by shallow low cells a e ou lined wi h ed do ed lines in
Figu e 3B. Wa e in il a ing in hese a eas ela i ely quickly e u ns o he land su ace, a els
sho e dis ances and is cha ac e ized by sho subsu ace esidence imes (Figu e 4 A and B). In
con as deepe low cells, which de elop o a eas whe e wa e in il a es deep in o he subsu ace
p edomina ely a loca ions ha a e loca ed a away om he channel segmen , ha e longe a el
dis ances (o en spanning he en i e ex en o he model domain) and show signi ican ly longe
esidence imes as also e lec ed in he wa e ages ( esidence ime in he subsu ace since in il a ion)
plo ed o he cen al 2D ansec in Figu e 5B. Deepe low cells a e con olled by he gene al
hyd aulic g adien ac oss he model domain. The low ield o he mic o- opog aphy model wi h a
mean leng h o 0.25m (ml-0.25m) shows no clea sepa a ion be ween shallow and deep low cells
because he opog aphic a ia ions a e oo small o c ea e su icien a ia ions in subsu ace hyd aulic
po en ials ha could induce signi ican shallow low cells (Figu e 3 C). Simila ly in he plana
e e ence model (Figu e 3 D) low pa hs a e ela i ely uni o m in space wi h low di ec ions almos
pa allel o he plana land su ace.

[114]
STUDY 2
Figu e 3: Subsu ace low pa hs de i ed om pa icle acking. A) Flow pa hs o he 3D domain o he
mic o- opog aphy ealiza ion wi h a mean leng h o 0.5 m. B)-D) Flow pa hs p ojec ed o a c oss
sec ion a he cen e o he 3D domain (yellow dashed line in A) o he wo mic o- opog aphy models
and he plana e e ence model. Ou lined a eas ( ed do ed lines) in B) ep esen he ypical down and
upwelling mo emen o he shallow low sys em induced by su ace mic o- opog aphy. Yellow do ed
lines ep esen wo low pa hs (in il a ing a X = 0.4 m and X = 6.8 m) e lec ing long and sho
subsu ace esidence imes o which he biogeochemical e olu ion is shown in Figu e 4. The model
domain is 10 m x 20 m x 2 m.
[115]
STUDY 2
3.2 Biogeochemical e olu ion along low pa hs
Figu e 4 (E-H) depic s exempla ily he esul s o he biogeochemical simula ions, shown o wo
selec ed subsu ace low pa hs o he ml-0.5m mic o- opog aphy model. The wo low pa hs,
beginning a loca ion X = 0.4 m and X = 6.8 m (shown as yellow do ed lines in Figu e 3B), ep esen
he deep and shallow low cells espec i ely. Resul s o he deep low pa h a e shown in Figu e 4 A,
C, E, G and o he shallow one in Figu e 4 B, D, F, H. Bo h low pa hs s a in he unsa u a ed zone
whe e p essu e heads a e nega i e (C and D). Fo he unsa u a ed zone, dissol ed oxygen
concen a ions a e cons an ly high (E and F) due o unlimi ed di usi e supply o a mosphe ic oxygen.
Ae obic espi a ion is he dominan p ocess wi hin he unsa u a ed zone. The high u no e o o ganic
ma e ial and he associa ed elease o o ganically bound ni ogen wi hin he unsa u a ed zone esul s
in inc easing concen a ions o ammonium (G and H), which is in u n oxidized o ni a e due o
ni i ica ion (E and F). When he low pa hs each he sa u a ed zone (p essu e heads become
posi i e), oxygen con en s a e dec easing and u no e due o ae obic espi a ion wi h associa ed
elease and oxida ion o ammonium a e slowed down (E and F). Oxygen con en s a e ini ially
luc ua ing in he sa u a ed zone because o p essu e head a ia ions (i.e. wa e able luc ua ions due
o ain e en s), which a e coupled o he oxygen bounda y condi ion as shown in Figu e 2. Once he
low pa hs each a dep h below he wa e able o abou 0.25 m (p essu e heads >= 0.25 m) oxygen
become limi ing and is comple ely deple ed a e ~90 days o he deep and a e ~100 days o he
shallow low pa h. Unde anoxic condi ions, inc easing concen a ions o educed species (e.g.
i on(II) o sul ide) indica es ha he sys em sequen ially shi s o de-ni i ica ion, i on(III)- and sul a e
educ ion (Figu e 4 G and H). A e 250 days, he deep low pa h is in a comple ely educed s a e
whe e all oxidized species a e deple ed (Figu e 4 E and G) and condi ions emains educed un il he
low pa h eeme ges a he su ace and he wa e ex il a es. The shallow low pa h eaches
comple ely educed condi ions a e 200 days, bu sho ly be o e ex il a ion oxygen becomes
a ailable again and oxida ion p ocesses a e eac i a ed (Figu e 4 F and H). The eason why e-
oxida ion only occu es a he end o he shallow low pa h is ela ed o he co esponding ex il a ion
loca ion. The shallow low pa h ends in a shallow, wa e illed dep ession whe e ponded wa e heigh s
a e low enough (p essu e heads < 0.25 m) o a mosphe ic oxygen o di use in o he uppe mos
laye s o he pea so ha oxygen is in con ac wi h he upwelling educed wa e . In con as , he deep
low pa h which ex il a es in o he s eam channel, whe e ponded wa e dep hs a e oo la ge o allow
esupply o oxygen by di usion; no eoxida ion o educed species is obse ed. Anima ion 1 and 2
(auxilia y ma e ial) show simila esul s whe e edox condi ions a e changing along wo isola ed
subsu ace low pa hs (deep and shallow) ex ac ed om he 3D model domain.
[116]
STUDY 2
Figu e 4: Resul s o biogeochemical simula ions along wo di e en low pa hs. A,C,E,G ep esen a
deep low pa h wi h long subsu ace esidence ime and B,D,E,F a low pa h o he shallow low
sys em. Subsu ace low eloci ies, p essu e heads, low dep hs and a el dis ances as shown in
A,B,C,D we e de i ed om nume ical low modeling and we e used as hyd ologic bounda y condi ions
o he biogeochemical simula ions. Addi ionally, oxygen a ailabili y (E, F) was coupled o he
p essu e head dynamics (C,D), indi idually o each low pa h. E and F show he e olu ion o
oxidized species (ni a e, i on(III) and sul a e) in ime no malized o hei co esponding ini ial
concen a ions C =0 and G and H he e olu ion o educed species (ammonium, i on(II) and sul ide)
no malized o hei inal concen a ions C =max. How edox cond ions a e changing in ime is also
shown in Anima ion 1 and 2 (auxilia y ma e ial).
[117]
STUDY 2
3.3 Spa ial pa e ns o ho spo s
In he p e ious pa ag aph, esul s o he biogeochemical simula ions o wo selec ed subsu ace low
pa hs we e p esen ed in he ime domain. A ep esen a ion in space is depic ed in Figu e 5 and was
gene a ed by in e pola ing local species concen a ions and eac ion a es om he biogeochemical
model o all low pa hs in o he 3D spa ial domain o he low model. In Figu e 5 he esul s o he
p ocess o sul a e educ ion in he model wi h ml = 0.5 m a e p esen ed as an example and plo ed
o he cen al ansec aligned along Y = 5 m (dashed line in Figu e 3 A). Panel C and D shows
simula ed sul a e educ ion and sul ide oxida ion a es whe eas panel E and F show he co esponding
concen a ions o he eac ion p oduc (sul ide) and educ (sul a e). Flow pa hs a e shown in panel A
and he age o subsu ace wa e ( esidence ime in he subsu ace since in il a ion de i ed om
pa icle acking) is depic ed in panel B. In he c oss sec ion, a eas o in ensi e sul a e educ ion (ho
spo s) a e isible as well as a eas whe e sul a e educ ion is p ac ically inac i e (panel C). The la e
a eas a e mainly associa ed wi h zones o upwelling subsu ace wa e ha is in a educed s a e and
deple ed o sul a e (plo E and F). They a e p e e en ially loca ed below local dep essions. Fo a eas
o in il a ion, p e e en ially loca ed below local hummocks, ho spo s (panel C) o sul a e educ ion
can de elop because he in il a ing wa e , o igina ing om he oxygena ed unsa u a ed zone, is ich
in sul a e which can be educed when mo e educing condi ions a e encoun e ed a inc easing dep h
(panel E). This gene al pa e n wi h local educ ion ho spo s below hummocks and an inhibi ion o
absence o educ ion p ocesses below dep essions, is also e iden o all o he edox-sensi i e species
(e.g. see plo s in he supplemen Figu e A1-A7). In compa ison, oxida ion p ocesses (i on(III)- ,
ae obic sul ide oxida ion) show a e e sed pa e n, whe e local ho spo s a e p e e en ially gene a ed
below dep essions whe e olde upwelling wa e , ich in educed species, comes in con ac wi h
a mosphe ic oxygen (panel D). In in il a ing a eas, oxida ion p ocesses a e p ac ically inac i e as he
eshly in il a ed wa e ca ies p edominan ly oxidized species.
[124]
STUDY 2
o be highly a iable in space and ime in we lands wi h a hummocky opog aphy, depending on he
clima ic bounda y condi ions [F ei e al., 2010]. Du ing in ensi e ain all e en s, su ace s o age and
uno gene a ion in we lands wi h shallow wa e able can be con olled by a dynamic ill and spill
mechanism [F ei e al., 2010]. Dep essions a e illed wi h wa e due o ising g oundwa e le els
du ing onse o ain all. Wi h las ing ain all, isola ed ponded dep essions s a o in e connec wi h
each o he building ex ended su ace low ne wo ks [F ei e al., 2010; An oine e al., 2009]. These
su ace low ne wo ks can e icien ly d ain la ge ac ions o he we land's su ace. A imes mo e
han 80% o he gene a ed s eam discha ge may o igina e om his ype o su ace low [F ei e al.,
2010]. Du ing high wa e able condi ions, as di usion o a mosphe ic oxygen in o he subsu ace
sys em is limi ed o a eas o high ele a ion (hummocks), which emain unsa u a ed a he su ace.
Du ing wa e able ecessions and dec easing su ace ponding, di usion o a mosphe ic oxygen,
below dep essions wi h lowe su ace ponding, becomes mo e e ec i e in e ms o inc easing a es
o esupply, which igge s oxida ion p ocesses o upwelling condi ions. Gene ally ield da a on
oxygen supply in we lands, i s coupling o wa e able dynamics and pea p ope ies a e sca ce
[A agonès and Blodau., 2012], s essing he impo ance o i ual modeling s udies. A special
condi ion can de elop du ing ex ended d y pe iods, whe e dep essions become disconnec ed om he
declining wa e able. Below hese disconnec ed dep essions hyd aulic g adien s may e e se,
swi ching om upwelling o in il a ing condi ions. In u n oxida ion ho spo s will diminish because
esupply o educed species om upwelling g oundwa e is dis up ed. I is easonable o assume ha
du ing d ough s ho spo pa e ns will become less p onounced and may e en ually anish as he
sys em g adually shi s owa ds a mo e homogenous dis ibu ion o p ocess ac i i ies.
In eal we land sys ems p obably mo e han one mechanism will be esponsible o he o ma ion o
biogeochemical ho spo s [McClain e al., 2003] and a clea sepa a ion o he in luence o one speci ic
p ocess is almos impossible unde ield condi ions. The simula ions p esen ed he e, howe e ,
demons a e ha he e ogeneous p ocess pa e ns in hummocky we lands can be explained by he
complex e-dis ibu ion o edox-sensi i e solu es in space as being con olled by mic o- opog aphy
induced, subsu ace anspo p ocesses and al e na ing biogeochemical bounda y condi ions.
Fu he mo e, he p esen ed concep shows ha biogeochemical ho spo s can be gene a ed wi hou
e e ence o ma e ial he e ogenei ies which o en a e ha dly obse able in ho izon ally ela i ely
homogenous pea soils [Mo is and Wadding on, 2011; Holden and Bu , 2003; Ree e e al., 2001;
Ree e e al., 2006; Clymo, 1984]. O cou se he p esen ed concep neglec s impo an aspec s o eal
ield condi ions. E ec s o he we lands ege a ion like oo wa e up ake and i s in luence on
subsu ace low o he special biogeochemical condi ions wi hin he hizosphe e [C ow and Wiede ,
2005; Kno e al., 2008; Wachinge e al., 2000] a e no conside ed as well as he po en ial e ec s o
dispe sion he a ailabili y o elec on accep o s and dono s. Hyd odynamic dispe sion may cause a
smea ing e ec whe e he bounda ies be ween ho spo s and su ounding a eas a e no as sha p and

[125]
STUDY 2
clea ly exp essed as in an ad ec i ely domina ed sys em, because solu es a e also e-dis ibu ed along
concen a ion g adien s (di usion) and ans e sally and longi udinally along he ad ec i e low
di ec ions (dispe sion). The biogeochemical simula ions we e pe o med using 5-day ime s eps,
which was necessa y because o compu a ional cons ain s du ing he low modeling (e.g. memo y
o e low, s o age limi a ions). Howe e , i is known ha hyd ological e en s a ime scales o hou s
(e.g. single ains o m e en s) can in luence he biogeochemical p ocesses wi hin we lands, as e.g.
demons a ed o pulses o N2O emission [Goldbe g e al., 2010] o high ins an aneous CO2
p oduc ion [Deppe e al., 2010] a e we ing. Dynamics a hese ime scales, howe e , we e no he
main ocus o his wo k and a his poin canno be ully accoun ed o in he p esen modeling
app oach because o compu a ional limi a ions. Fu he i is known ha o ganic ca bon in we lands
ypically consis s o a ac ion o labile componen s ha can be easily u ilized by mic o-o ganisms
(mos ly wi hin shallow laye s) and mo e ecalci an componen s (mo e abundan in deepe laye s)
[Ya i and Lang, 1990; Reiche e al., 2010; Moo e e al., 2007]. Labile o ganic ca bon is no
uni o mly a ailable as is assumed in ou app oach. Howe e , he e a e wo main easons why we
hink ha ou assump ion o unlimi ed ca bon supply is none heless easonable. Fi s ly, labile o ganic
ca bon a ailabili y is highe in shallow pea laye s, in which mos o he modeled p ocesses occu ,
mos ly due o inpu s om he ege a ion and high e men a ion ac i i y in he hizosphe e [Kno e
al., 2008; Wachinge e al., 2000; Reiche e al., 2010]. Secondly, we did no include me hanogenesis,
o which he supply o elec on dono s will be he key con ol, as he ubiqui ous CO2 may se e as
elec on accep o [Ach nich e al., 1995]. Field obse a ions sugges ed ha i al e na i e elec on
accep o s we e p esen , he espec i e p ocess p oceeded, while unde me hanogenic condi ions,
espi a o y ac i i y slowed down and pa ly ceased [Bee and Blodau, 2007; Kno e al., 2009].
Ne e heless, he p ocess a e, cons an in his case, depends on he quali y o o ganic ma e used and
is no uni e sal bu subs a e speci ic. The applica ion o he Red ield a io o simula e elease o
o ganic bound ni ogen due o decomposi ion o o ganic ma e ial in e es ial ecosys ems was
p obably a weak model assump ion. Recen li e a u e epo ed ha C:N:P a ios in e es ial
ecosys ems a y depending on ege a ion ypes, bu on he global scale a e age a abou 186:13:1 o
soil biomass and 60:7:1 o soil mic obial biomass [Cle eland and Lip zin, 2007]. In ou
biogeochemical model we assumed ha he majo i y o o ganic ca bon a ailable o mic obes
o igina es om ege a ion and e men ed plan ma e ial p ocessed by mic oo ganisms. The Red ield
a io is, howe e , na owe han he global a e age obse ed o soil biomass (106:16:1 compa ed o
186:13:1) and ni ogen elease would be o e es ima ed by ou model. Tha means ha he
concen a ions o ammonia, a es o ni i ica ion and hus also ni a e pools a ailable o
deni i ica ion may also be o e es ima ed. Ne e heless, his should ansla e in o sligh ly longe
phases o ni i ica ion o subsequen deni i ica ion only, hus no undamen ally al e ing spa ial
pa e ns o he model ou pu .
[126]
STUDY 2
4.2 Compa ison wi h ield obse a ions
Despi e hese simpli ica ions, he p esen ed model is capable o ep oducing spa ial a ia ions in po e
wa e concen a ions o edox-sensi i e solu es in he ield (Figu e 10). Ve ical concen a ion p o iles
we e measu ed in po e wa e om six di e en loca ions a he Lehs enbach ield si e, o an a ea,
which is compa able in size o he spa ial domain o he low model (10m x 20m) [Goldbe g e al.,
2010; Kno e al., 2009]. Simula ed maxima in ni a e concen a ions a e ound a a dep h o ~0.1m
and no di ec ly a he su ace, which ag ees wi h measu ed da a. The obse ed shi o ni a e
concen a ion maxima has been explained as a esul o plan up ake om he uppe laye s, as plan
co e o en leads o apid deple ion o ni a e concen a ions [Sil an e al., 2005]. Howe e ou
biogeochemical simula ions sugges an addi ional explana ion o he inc eased ni a e concen a ions
a shallow dep h: As shown o he c oss sec ions (Figu e 6 C) high ni i ica ion a es a e limi ed o a
ela i ely hin laye whe e u no e o ammonium o ni a e is highes . This laye o highe eac i i y
is he esul o he e ical anspo o wa e , which is being en iched wi h ammonium as i passes he
unsa u a ed zone. Because ni i ica ion a es unde ae obic condi ions depend on he local a ailabili y
o ammonium, highe ammonium concen a ions esul in highe ni i ica ion a es, which can be
ound di ec ly abo e he de-ni i ica ion zone whe e anae obic condi ions igge apid ni a e
educ ion. Simila indings we e epo ed o di e en ield s udies [Regina e al., 1999; Goldbe g e
al., 2010]. Measu ed dep h p o iles as shown in Figu e 10 a e o en used o calcula e biogeochemical
u no e a es based on a simpli ied app oach ea ing we lands as di usion limi ed sys ems whe e he
esupply o dissol ed elec on accep o s/dono s is solely con olled by di usion [Bee and Blodau,
2007; Clymo and B yan , 2008]. Howe e , model esul s show ha ad ec i e anspo can be an
impo an componen especially o sligh ly sloping we lands wi h mic o- opog aphy and can
signi ican ly a ec he spa ial a ailabili y and e-dis ibu ion o elec on accep o s and dono s wi hin
he subsu ace. Ve ical concen a ion p o iles simula ed in his s udy sugges ha dep h a ia ions in
he concen a ions o edox-sensi i e solu es obse ed in he ield a e p obably he esul o a complex
in e play be ween h ee-dimensional ad ec i e anspo p ocesses and biogeochemical eac ions,
which a e in u n con olled by mic o- opog aphy mode a ed in e ac ions be ween su ace and
subsu ace low p ocesses and do no a ise om pu e di usion and eac ions alone.
[127]
STUDY 2
Figu e 10: Obse ed and simula ed a ia ions o dep h p o iles o edox-sensi i e species (ni a e,
i on(II) and sul a e). G ey a eas ep esen en elopes o p edic ed dep h p o iles and he black lines
(mean +/- s anda d de ia ion) ac ual ield obse a ions aken simul aneously a six di e en loca ions
o an a ea which is compa able o he model 20 m x 10 m domain a he ield si e in he Lehs enbach
ca chmen .
[128]
STUDY 2
5 Conclusions and Implica ions
A he landscape scale, ipa ian we lands a e commonly assumed o be zones o enhanced
biogeochemical ans o ma ions (e.g. deni i ica ion) due o anae obic condi ions and la ge ca bon
supplies [Johns on, 1991]. Field s udies, howe e , ha e shown ha biogeochemical condi ions wi hin
we lands can be qui e di e se whe e mos o he biogeochemical u no e may be accomplished in
localized zones o highe eac i i y (ho spo s) [Paul e al., 2006; Kno e al., 2009; Kno and
Blodau, 2009; Fenne e al., 2011]. Unde ield condi ions, di e en p ocesses and mechanisms can
lead o he o ma ion o ho spo s [McClain e al., 2003] depending on he scale o in e es . Howe e ,
explaining such ho spo s solely by he he e ogeneous dis ibu ion o s a ic, physico-chemical
p ope ies o he soil [Ree e e al., 2001; Holden and Bu , 2003] may be oo simplis ic. Ou
simula ions indica e ha biogeochemical ho spo s can o m e en in homogeneous pea soils as a
esul o a dynamic subsu ace low sys em wi h (1) complex su ace/subsu ace in e ac ions whe e
su ace mic o- opog aphy induces a subsu ace low ield ha is cha ac e ized by a small-scale
zona ion o in- and ex il a ion and (2) hyd ological con ols o he biogeochemical bounda y
condi ions ha ei he acili a e o supp ess edox p ocesses in ex- and in il a ion a eas. Hence he
occu ence o eac i i y ho spo s does no need o be associa ed wi h s a ic he e ogenei ies in
physico-chemical soil p ope ies a p io i. In ac , he o ma ion o biogeochemical ho spo s in
we land sys ems may ha e he po en ial o al e he hyd odynamic p ope ies o he pea and he e o e,
ypically obse ed ma e ial he e ogenei y may esul om p ocesses desc ibed in his s udy. The
p ecipi a ion o i on oxides e.g., which p e e en ially occu s a oxida ion ho spo s, can lead o a
educ ion o he e ec i e po osi y and a lowe hyd aulic conduc i i y, p o iding a nega i e eedback
on oxygen pene a ion; o in a eas o educ ion ho spo s e.g. i on sul ides may become en iched ha
could be eoxidized upon mo e se e e d ying. Ou esul s o e a new pe spec i e on biogeochemical
ans o ma ion p ocesses in ipa ian we lands ha p o ides a dynamic amewo k o explain p ocess
he e ogenei y in we land soils and a iabili y in p ocess a es o e ime and space. Fu u e wo k will
ha e o add ess he in e play be ween di e en s a ic (e.g. soil p ope ies, ege a ion pa e ns) and
dynamic con ols (e.g. low, empe a u e & ege a ion dynamics) o spa ial and empo al a ia ions in
biogeochemical p ocess ac i i ies in we lands. I is clea ha a mechanis ic unde s anding o he links
be ween hyd ologic dynamics and biogeochemical ans o ma ions will be c ucial o an assessmen
o clima e change impac s on we land unc ions and associa ed ecosys ems se ices. The wo k
p esen ed he e can se e as s a ing poin o such an assessmen by p o iding an explo a i e,
mechanis ic modeling amewo k o in es iga e po en ial shi s in hyd ological and biogeochemical
p ocesses including changes in eedback mechanisms caused by changes in clima ic o cing.
[129]
STUDY 2
Acknowledgmen s
This s udy was unded by he Ge man Reasea ch Founda ion (DFG, g an FL 631/6-2). Thei
inancial suppo is g ea ly app ecia ed. We would like o acknowledge he cons uc i e commen s
om Ca olyn Oldham and an anonymous e iewe , which g ea ly helped o imp o e he inal
pape .The au ho s also hank Rob MacLa en, Young-Jin Pa k, And ea B ook ield and Ed Sudicky a
he Uni e si y o Wa e loo, Canada o hei in aluable help wi h he ins and ou s o he nume ical
code Hyd oGeoSphe e.

[130]
STUDY 2
Re e ences
Ach nich, C., F. Bak and R. Con ad (1995), Compe i ion o elec on dono s among ni a e educe s,
e ic i on educe s, sul a e educe s, and me hanogens in anoxic paddy soil, Biology and
Fe ili y o Soils 19(1), 65–72.
Alewell, C., S. Paul, G. Lischeid and F. R. S o ck (2007), Co- egula ion o edox p ocesses in
eshwa e we lands as a unc ion o o ganic ma e a ailabili y?, Science o he To al
En i onmen 404(2-3), 335–342, doi:10.1016/j.sci o en .2007.11.001.
An oine, M., M. Ja aux and C. Bielde s (2009), Wha indica o s can cap u e uno - ele an
connec i i y p ope ies o he mic o- opog aphy a he plo scale?, Ad ances in Wa e Resou ces
32(8), 1297–1310, doi:10.1016/j.ad wa es.2009.05.006.
Appelo, C. A. and D. Pos ma (2005), Geochemis y, g oundwa e and pollu ion, Taylo & F ancis.
Bee , J. and C. Blodau (2007), T anspo and he modynamics cons ain belowg ound ca bon u no e
in a no he n pea land, Geochimica e Cosmochimica Ac a 71(12), 2989–3002.
Bekins, B. A., E. Wa en and E. M. Godsy (1998), A Compa ison o Ze o O de , Fi s O de , and
Monod Bio ans o ma ion Models, G ound Wa e 36(2), 261–268.
Belle ue, W. (2003), Tecplo Use 's Manual, Am ec Enginee ing Inc.
Billen, G. (1982), An idealized model o ni ogen ecycling in ma ine sedimen s, Am. J. Sci 282, 512–
541.
Bishop, K., J. Seibe , S. Köhle and H. Laudon (2004), Resol ing he Double Pa adox o apidly
mobilized old wa e wi h highly a iable esponses in uno chemis y, Hyd ol. P ocess 18(1),
185–189.
Boano, F., A. Dema ia, R. Re elli and L. Ridol i (2010), Biogeochemical zona ion due o
in ameande hypo heic low, Wa e Resou ces Resea ch 46(2), W02511.
B ook ield, A. E., E. A. Sudicky, Y. J. Pa k and B. Conan J . (2009), The mal anspo modelling in
a ully in eg a ed su ace/subsu ace amewo k, Hyd ological P ocesses(23), 2150–2164,
10.1002/hyp.7282.
B o elli, A., O. Ca anza-Diaz, L. Rossi and D. Ba y (2011), Design me hodology accoun ing o he
e ec s o po ous medium he e ogenei y on hyd aulic esidence ime and biodeg ada ion in
ho izon al subsu ace low cons uc ed we lands, Ecological Enginee ing(37), 758–770.
B uland, G., C. Richa dson, and S. Whalen (2006), Spa ial a iabili y o deni i ica ion po en ial and
ela ed soil p ope ies in c ea ed, es o ed, and pai ed na u al we lands, We lands 26(4), 1042–
1056.
B uland, G. and C. Richa dson (2005), Hyd ologic, edaphic, and ege a i e esponses o
mic o opog aphic ees ablishmen in a es o ed we land, Res o a ion Ecology 13(3), 515–523.
Ca le, S. and G. Fogg (1996), T ansi ion P obabili y-Based Indica o Geos a is ics, Ma hema ical
Geology 28(4), 453–476.
Cheng, Y., M. S iegli z, G. Tu k and V. Engel (2011), E ec s o aniso opy on pa e n o ma ion in
we land ecosys ems, GEOPHYSICAL RESEARCH LETTERS 38(4), L04402.
Cle eland, C. and D. Lip zin (2007), C:N:P s oichiome y in soil: is he e a “Red ield a io” o he
mic obial biomass?, Biogeochemis y 85(3), 235-252
Clymo, R. S. (1984), The limi s o pea bog g ow h, Philosophical T ansac ions o he Royal Socie y
o London. B, Biological Sciences 303(1117), 605.
Clymo, R. S. and C. L. B yan (2008), Di usion and mass low o dissol ed ca bon dioxide, me hane,
and dissol ed o ganic ca bon in a 7-m deep aised pea bog, Geochimica e Cosmochimica Ac a
72(8), 2048–2066.
[131]
STUDY 2
C ow, S. and R. Wiede (2005), Sou ces o CO2 emission om a no he n pea land: oo espi a ion,
exuda ion, and decomposi ion, Ecology 86(7), 1825–1834.
Deppe, M., Kno K-H., McKnigh D. and Blodau C (2010) E ec s o sho - e m d ying and i iga ion
on CO2 and CH4 p oduc ion and emission om mesocosms o a no he n bog and an alpine en.
Biogeochemis y 100: 89-103.
De i o, K. J. and A. R. Hill (1997), Sulpha e dynamics in ela ion o g oundwa e -su ace wa e
in e ac ions in headwa e we lands o he sou he n Canadian Shield, Hyd ological P ocesses
11(5), 485–500.
Dosskey, M. and P. Be sch (1994), Fo es sou ces and pa hways o o ganic ma e anspo o a
blackwa e s eam: a hyd ologic app oach, Biogeochemis y 24(1), 1–19.
Nealson, K. and D. Sa a ini (1994), I on and manganese in anae obic espi a ion: en i onmen al
signi icance, physiology, and egula ion, Annual Re iews in Mic obiology 48(1), 311–343.
Ecke , P. and C. A. Appelo (2002), Hyd ogeochemical modeling o enhanced benzene, oluene,
e hylbenzene, xylene (BTEX) emedia ion wi h ni a e, Wa e Resou ces Resea ch 38(8), 1130.
Es op-A agonés, C. and C. Blodau (2012), Con ols on in si u oxygen and DIC dynamics in pea s o a
empe a e en., Jou nal o Geophysical Resea ch – Biogeosciences 117(G2), G02002.
Fenne , N., R. Williams, H. Tobe man, S. Hughes, B. Reynolds and C. F eeman (2011),
Decomposi ion ho spo s in a ewe ed pea land: implica ions o wa e quali y and ca bon
cycling, Hyd obiologia, 1–16.
F ei, S., G. Lischeid and J. H. Fleckens ein (2010), E ec s o mic o- opog aphy on su ace-subsu ace
exchange and uno gene a ion in a i ual ipa ian we land-a modeling s udy, Ad ances in
Wa e Resou ces 33(11), 1388–1401.
Ga ni, A. and N. Kenne h (1990), Hyd aulic cha ac e is ics o ou pea lands in Minneso a, Canadian
Jou nal o Soil Science 70(2), 239–253.
Goldbe g, S., K. Kno , C. Blodau, G. Lischeid and G. Gebaue (2010), Impac o al e ing he wa e
able heigh o an acidic en on N2O and NO luxes and soil concen a ions, Global Change
Biology 16(1), 220–233.
Goldsmi h, C. D. and R. K. Balde son (1988), Biodeg ada ion and g ow h kine ics o en ichmen
isola es on benzene, oluene, and xylene, Wa e Science and Technology WSTED 4 20(11/12).
Hill, A. R. (2000), S eam chemis y and ipa ian zones, S eams and g ound wa e s, 83–110.
Hin on, M. J., S. L. Schi and M. C. English (1998), Sou ces and lowpa hs o dissol ed o ganic
ca bon du ing s o ms in wo o es ed wa e sheds o he P ecamb ian Shield, Biogeochemis y
41(2), 175–197.
Holden, J. and T. P. Bu (2003), Hyd aulic conduc i i y in upland blanke pea : measu emen and
a iabili y, Hyd ological P ocesses 17(6), 1227–1237.
Hun ing on, T. G. (2006), E idence o in ensi ica ion o he global wa e cycle: Re iew and
syn hesis, Jou nal o Hyd ology 319(1-4), 83–95.
Inamda , S., J. Rupp and M. Mi chell (2009), G oundwa e lushing o solu es a we land and hillslope
posi ions du ing s o m e en s in a small glacia ed ca chmen in wes e n New Yo k, USA,
Hyd ological P ocesses 23(13), 1912–1926.
Jacks, G. and A. C. No s öm (2004), Hyd ochemis y and hyd ology o o es ipa ian we lands,
Fo es ecology and Managemen 196(2-3), 187–197.
Jakobsen, R. (2007), Redox mic oniches in g oundwa e : A model s udy on he geome ic and kine ic
condi ions equi ed o concomi an Fe oxide educ ion, sul a e educ ion, and me hanogenesis,
Wa e Resou ces Resea ch 43(12), W12S12.
[132]
STUDY 2
Johns on, C. A. (1991), Sedimen and nu ien e en ion by eshwa e we lands: e ec s on su ace
wa e quali y, C i ical Re iews in En i onmen al Science and Technology 21(5), 491–565.
Jones, J. P., E. A. Sudicky, A. E. B ook ield and Y. J. Pa k (2006), An assessmen o he ace -based
app oach o quan i ying g oundwa e con ibu ions o s eam low, Wa e Resou ces Resea ch
42, W02407, doi:10.1029/2005WR004130.
Kelly, W. R., G. M. Ho nbe ge , J. S. He man and A. L. Mills (1996), Kine ics o BTX
biodeg ada ion and mine aliza ion in ba ch and column sys ems, Jou nal o con aminan
hyd ology 23(1-2), 113–132.
Kno , K. H., B. Glase and C. Blodau (2008), Fluxes and 13C iso opic composi ion o dissol ed
ca bon and pa hways o me hanogenesis in a en soil exposed o expe imen al d ough ,
Biogeosciences 5: 1457–1473.
Kno , K. H., G. Lischeid and C. Blodau (2009), Dynamics o edox p ocesses in a mine o ophic en
exposed o a wa e able manipula ion, Geode ma 153(3-4), 379–392.
Kno , K. and C. Blodau (2009), Impac o expe imen al d ough and ewe ing on edox
ans o ma ions and me hanogenesis in mesocosms o a no he n en soil, Soil Biology and
Biochemis y 41(6), 1187–1198.
Le Kellogg and S. D. B idgham (2003), Phospho us e en ion and mo emen ac oss an omb o ophic-
mine o ophic pea land g adien , Biogeochemis y 63(3), 299–315.
Lischeid, G., A. Kolb, C. Alewell and S. Paul (2007), Impac o edox and anspo p ocesses in a
ipa ian we land on s eam wa e quali y in he Fich elgebi ge egion, sou he n Ge many,
Hyd ological P ocesses 21(1), 123–132, 10.1002/hyp.6227.
Lo ley, D., J. Coa es, E. Blun -Ha is, E. Phillips and J. Woodwa d (1996), Humic subs ances as
elec on accep o s o mic obial espi a ion, Na u e 382(6590), 445–448.
MacQua ie, K. T., E. A. Sudicky and E. O. F ind (1990), Simula ion o biodeg adable o ganic
con aminan s in g oundwa e : 1. Nume ical o mula ion in p incipal di ec ions, Wa e
Resou ces Resea ch 26(2), 207–222.
McClain, M., E. Boye , C. Den , S. Ge gel, N. G imm, P. G o man, S. Ha , J. Ha ey, C. Johns on
and E. Mayo ga (2003), Biogeochemical ho spo s and ho momen s a he in e ace o
e es ial and aqua ic ecosys ems, Ecosys ems 6(4), 301–312.
McMahon, P. B., and F. H. Chapelle (2008), Redox p ocesses and wa e quali y o selec ed p incipal
aqui e sys ems, G ound Wa e 46(2), 259–271.
Mille o, F. J., S. Hubinge , M. Fe nandez and S. Ga ne (1987), Oxida ion o H2S in seawa e as a
unc ion o empe a u e, pH, and ionic s eng h, En i onmen al Science & Technology 21(5),
439–443.
Mi chell, C. and B. B an i eun (2005), Hyd ogeomo phic con ols on educ ion-oxida ion condi ions
ac oss Bo eal upland Pea land in e aces, Ecosys ems 8(7), 731–747.
Moo e, T., J. Bubie and L. Bledzki (2007), Li e decomposi ion in empe a e pea land ecosys ems:
he e ec o subs a e and si e, Ecosys ems 10(6), 949–963.
Mo is, P. J. and J. M. Wadding on (2011), G oundwa e esidence ime dis ibu ions in pea lands:
Implica ions o pea decomposi ion and accumula ion, Wa e Resou ces Resea ch 47(2),
W02511.
Pa khu s , D. (1995), Use 's guide o PHREEQC: A compu e p og am o specia ion, eac ion-pa h,
ad ec i e- anspo , and in e se geochemical calcula ions, U.S. GEOLOGICAL SURVEY
WA e -Resou ces In es iga ions Repo 95-4227.
Paul, S., K. Küsel and C. Alewell (2006), Reduc ion p ocesses in o es we lands: acking down
he e ogenei y o sou ce/sink unc ions wi h a combina ion o me hods, Soil Biology and
Biochemis y 38(5), 1028–1039.
[133]
STUDY 2
Pyzik, A. J. and S. E. Somme (1981), Sedimen a y i on monosul ides: kine ics and mechanism o
o ma ion, Geochimica e Cosmochimica Ac a 45(5), 687–698.
Red ield, A. (1934), On he p opo ions o o ganic de i a i es in sea wa e and hei ela ion o he
composi ion o plank on, James Johns one memo ial olume 176, 92.
Ree e, A. S., Di Siegel and P. H. Glase (2001), Simula ing dispe si e mixing in la ge pea lands,
Jou nal o Hyd ology 242(1-2), 103–114.
Ree e, A., R. E ensen, P. Glase , D. Siegel and D. Rosenbe y (2006), Flow pa h oscilla ions in
ansien g ound-wa e simula ions o la ge pea land sys ems, Jou nal o Hyd ology 316(1-4),
313–324.
Regina, K., J. Sil ola and P. Ma ikainen (1999), Sho ‐ e m e ec s o changing wa e able on N2O
luxes om pea monoli hs om na u al and d ained bo eal pea lands, Global Change Biology
5(2), 183–189.
Reiche, M., G. Gleixne and K. Küsel (2010), E ec o pea quali y on mic obial g eenhouse gas
o ma ion in an acidic en, Biogeosciences 7(1), 187–198.
Richa dson, M. C., B. A. B an i eun, V. B. Robinson and P. A. G anie o (2007), Towa ds simula ing
biogeochemical ho spo s in he landscape: A geog aphic objec -based app oach, Jou nal o
Hyd ology 342(1-2), 97–109.
Schi me , M., B. J. Bu le , J. W. Roy, E. O. F ind and J. F. Ba ke (1999), A ela i e-leas -squa es
echnique o de e mine unique Monod kine ic pa ame e s o BTEX compounds using ba ch
expe imen s, Jou nal o con aminan hyd ology 37(1-2), 69–86.
Sege s, R. and S. W. Kengen (1998), Me hane p oduc ion as a unc ion o anae obic ca bon
mine aliza ion: a p ocess model, Soil Biology and Biochemis y 30(8-9), 1107–1117.
Shabaga, J. and A. Hill (2010), G oundwa e - ed su ace low pa h hyd odynamics and ni a e
emo al in h ee ipa ian zones in sou he n On a io, Canada, Jou nal o Hyd ology 388(1), 52–
64.
Sil an, N., E. Tui ila, V. Ki unen, H. Vasande and J. Laine (2005), Ni a e up ake by E iopho um
agina um con ols N2O p oduc ion in a es o ed pea land, Soil Biology and Biochemis y 37(8),
1519–1526.
S onedahl, S., J. Ha ey, A. Wö man, M. Salehin and A. Packman (2010), A mul iscale model o
in eg a ing hypo heic exchange om ipples o meande s, Wa e Resou ces Resea ch 46(12),
W12539.
S umm, W. and J. J. Mo gan (1995), Aqua ic chemis y: chemical equilib ia and a es in na u al
wa e s, JOHN WILEY & SONS, NEW YORK, NY 10158(USA). 1995.
Tauchni z, N., R. Meissne , S. Be nsdo and U. Wegene (2010), Ni ogen Fluxes o a Slope Mi e in
he Ge man Ha z Moun ains, Wa e , Ai , & Soil Pollu ion 205(1), 107–112.
The ien, R., R. McLa en, E. Sudicky and S. Panday (2008), Hyd oGeoSphe e A Th ee-dimensional
Nume ical Model Desc ibing Fully-in eg a ed Subsu ace and Su ace Flow and Solu e
T anspo (Manual), Uni e si y o Wa e loo.
To h, J. (1962), A heo y o g oundwa e mo ion in small d ainage basins in cen al Albe a, Canada,
Jou nal o Geophysical Resea ch 67(11), 4375–4387.
Van Cappellen, P. and Y. Wang (1996), Cycling o i on and manganese in su ace sedimen s: a
gene al heo y o he coupled anspo and eac ion o ca bon, oxygen, ni ogen, sul u , i on,
and manganese, Ame ican-Jou nal ok Science 296, 197–243.
Vidon, P., C. Allan, D. Bu ns, T. P. Du al, N. Gu wick, S. Inamda , R. Low ance, J. Okay, D. Sco
and S. Sebes yen (2010), Ho Spo s and Ho Momen s in Ripa ian Zones: Po en ial o Imp o ed
Wa e Quali y Managemen 1, JAWRA Jou nal o he Ame ican Wa e Resou ces Associa ion
46(2), 278–298.
[140]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A5: Resul s o he biogeochemical simula ions shown o he ni i ica ion o he mic o-
opog aphy scena io wi h he mean leng h 0.5 m. PHREEQC simula ions we e pe o med along he
low pa hs shown in A. Resul s we e in e pola ed in o he 2D c oss sec ions. B shows he age
dis ibu ion in yea s o subsu ace low de i ed om backwa d pa icle acking. Ni i ica ion a es in
mol/Ls a e shown in C. Ammonium concen a ions and ni a e concen a ions in mol/L a e shown in D
and E espec i ely. Dissol ed oxygen concen a ions in mol/L a e shown in F.

[141]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A6: Resul s o he biogeochemical simula ions shown o he ae obic sul ide oxida ion o he
mic o- opog aphy scena io wi h he mean leng h 0.5 m. PHREEQC simula ions we e pe o med along
he low pa hs shown in A. Resul s we e in e pola ed in o he 2D c oss sec ions. B shows he age
dis ibu ion in yea s o subsu ace low de i ed om backwa d pa icle acking. Ae obic sul ide
oxida ion a es in mol/Ls a e shown in C. Sul ide concen a ions and sul a e concen a ions in mol/L
a e shown in D and E espec i ely. Dissol ed oxygen concen a ions in mol/L a e shown in F.
[142]
STUDY 2 – SUPPLEMENTARY MATERIAL
.
Figu e A7: Resul s o he biogeochemical simula ions shown o he anae obic sul ide oxida ion o he
mic o- opog aphy scena io wi h he mean leng h 0.5 m. PHREEQC simula ions we e pe o med along
he low pa hs shown in A. Resul s we e in e pola ed in o he 2D c oss sec ions. B shows he age
dis ibu ion in yea s o subsu ace low de i ed om backwa d pa icle acking. Anae obic sul ide
oxida ion a es in mol/Ls a e shown in C. Sul ide concen a ions, i on(III) and i on(II) concen a ions in
mol/L a e shown in D-F espec i ely.
[143]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A8: Top iew o he mic o- opog aphy scena ios and he plana e e ence. Mic o- opog aphy
is shown in ca ego ies, ed o hummock and blue o hollow s uc u es. Addi ionally, o he plana
e e ence he linea slope is shown. Black a eas o he ide sides ep esen a eas o ae obic
espi a ion ho spo s ela i e o hei su oundings. Ae obic espi a ion can only occu i oxygen is
p esen . Below hummocks a a iably sa u a ed zone wi h high oxygen con en s is s able whe e
p e e en ial ae obic espi a ion occu s. Zones below hollows usually a e wa e sa u a ed whe e no
oxygen is a ailable o ae obic espi a ion.
[144]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A9: Top iew o he mic o- opog aphy scena ios and he plana e e ence. Mic o- opog aphy
is shown in ca ego ies, ed o hummock and blue o hollow s uc u es. Addi ionally, o he plana
e e ence he linea slope is shown. Black a eas o he ide sides ep esen a eas o p e e en ial
i on(III) educ ion (ho spo s) ela i e o hei su oundings. The pa chy pa e n de elops because
I oon(III) educ ion p e e en ially occu s below hummock s uc u es because o highe i on(III)
abundance. Below hollows upwelling wa e is ich in educed i on species (I on(II)).
[145]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A10: Top iew o he mic o- opog aphy scena ios and he plana e e ence. Mic o- opog aphy
is shown in ca ego ies, ed o hummock and blue o hollow s uc u es. Addi ionally, o he plana
e e ence he linea slope is shown. Black a eas o he ide sides ep esen a eas o p e e en ial
ammonium oxida ion (ho spo s) ela i e o hei su oundings.

[146]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A11: Top iew o he mic o- opog aphy scena ios and he plana e e ence. Mic o- opog aphy
is shown in ca ego ies, ed o hummock and blue o hollow s uc u es. Addi ionally, o he plana
e e ence he linea slope is shown. Black a eas o he ide sides ep esen a eas o p e e en ial
i on(II) oxida ion (ho spo s) ela i e o hei su oundings.
[147]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A12: Top iew o he mic o- opog aphy scena ios and he plana e e ence. Mic o- opog aphy
is shown in ca ego ies, ed o hummock and blue o hollow s uc u es. Addi ionally, o he plana
e e ence he linea slope is shown. Black a eas o he ide sides ep esen a eas o p e e en ial
sul ide oxida ion (ho spo s) ela i e o hei su oundings.
[148]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A13:. Fence plo s showing he zones o p e e en ial sul a e educ ion o he whole 3D domain
o he mean leng h 0.25 m model.
[149]
STUDY 2 – SUPPLEMENTARY MATERIAL
Figu e A14: Fence plo s showing he zones o p e e en ial sul a e educ ion o he whole 3D domain
o he plana e e ence model