the impact of surface water exchange on the nutrient and particle dynamics in side-arms along the...

Science of the Total Environment 328(2004) 207–218

0048-9697/04/$ - see front matter� 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2004.01.006

The impact of surface water exchange on the nutrient and particledynamics in side-arms along the River Danube, Austria

Thomas Hein*, Christian Baranyi, Walter Reckendorfer, Fritz Schiemer

Institute of Ecology and Conservation Biology, Department of Limnology, University of Vienna, Althanstrasse 14, A–1090 Vienna,Austria

Received 4 July 2003; received in revised form 20 January 2004; accepted 26 January 2004

Abstract

Results of two monitoring programs obtained in the free-flowing section of the Danube downstream of Viennawere used to evaluate the effects of river restoration designed to increase surface water inputs into side-arms.Functional descriptors like hydrochemical parameters and plankton react immediately to restored hydrologicalconditions and offer the opportunity to elucidate the hydrological control on organic processing as an importantecosystem function in fluvial landscapes. Two hydraulic parameters were estimated and linked to basic ecologicalproperties. The level of hydrological connectivity was defined as the average annual duration(days per year) ofupstream surface connection and can be used as a ‘simple to estimate’ parameter within the planning phase. Waterage, an adapted measure of residence time based on more detailed information, allow description of the temporaldevelopment in side-arms. Greater hydrological connectivity leads to lower conductivity levels and increased nutrientconcentrations due to the shift of the dominating source to river water. The contribution of river flow is indicated byhigher suspended solid concentrations in side-arms than disconnected water bodies. The phytoplankton biomass showsthe highest mean values at a duration of integration of 1 month a and decrease with increasing connectivity. They1

relationships point to a more ‘main channel like’ hydrochemical situation in the side-arms, with a medium level ofphytoplankton biomass and increased autochthonous carbon export. No evidence of eutrophication was found due tothe shift of the side-arm from an organic matter sink to a source. On a more detailed level, water age demonstratesthe temporal patterns of riverine input, the development of plankton production and the shift between hydrologicaland biological control of phytoplankton vs. riverine flow in a side-arm. The hydrologic parameters were usefulpredictors for evaluating the effects of restoration measures in river floodplain systems.� 2004 Elsevier B.V. All rights reserved.

Keywords: Autochthonous carbon; Restoration; Nutrients; Hydrological connectivity; Floodplain; Main channel; Danube

1. Introduction

Running waters are important links in the globalbiogeochemical cycles. They transport organic

*Corresponding author. Tel.:q43-1-4277-54352; fax:q43-1-4277-9542.

E-mail address: [email protected](T. Hein).

matter from terrestrial sources, produce organicmaterial within aquatic environments and degradeorganic matter on its way downstream(Hedges etal., 2000). The net effect of processes such asstorage, breakdown and transformation of matteris a basic ecological property of lotic ecosystems(Fisher et al., 1998). In large rivers, these pro-

208 T. Hein et al. / Science of the Total Environment 328 (2004) 207–218

cesses are influenced by the spatial and temporalavailability of subsystems with higher hydrologicalretention, slack water areas, such as inshore reten-tion zones, side-arms, riparian zones, floodplains,backwaters and wetlands(Reckendorfer et al.,1999; Thorp and Delong, 2002). The availabilityof certain slack water areas or dead zones iscontrolled by the riverine discharge, morphologyand the exchange condition, e.g. during lowerwater tables inshore retention can foster the proc-essing of organic matter(Schiemer et al., 2001).In contrast, during high water, floodplains andriverine wetlands play a key role for matter reten-tion (Tockner et al., 1999). In river reaches witha variety of subsystems integrated in the riverineflow or with a frequent water exchange with themain channel, the local aquatic production and theparticulate organic matter(POM) accumulationare emphasized. Therefore, higher ecosystem sta-bility can be expected(Fisher et al., 1998).

Human impacts(e.g. on river systems likeregulation) canalization and flood protection sig-nificantly reduce the retention capacities of theriverine landscape and limit the exchange of matterto short periods of high flow(Tockner et al.,1999). In particular, embankments and lateral damslead to a significant decrease of hydrologicexchange of surface waters. The riverine landscapewith secondary channels and all the various waterbodies within the floodplain were disconnectedfrom the river flow for long periods. Lotic condi-tions were reduced to the regulated main channeland lentic conditions prevail in all other waterbodies as shown for the Austrian Danube(Tockneret al., 1998). In large rivers, therefore, restorationneeds to focus on the availability of slack waterareas with an active hydrologic exchange at a widerange of discharge conditions to improve the eco-logical integrity of these riverine systems(Brad-shaw, 1996). As an example, the ‘DanubeRestoration Project’(DRP) was developed alongthe Austrian Danube to enhance the hydrologicalconnectivity between the main channel and formerside-arms by increased duration of lotic conditionsand hydrologic exchange(Tockner et al., 1998;Schiemer et al., 1999).

The Danube transports high amounts of nutrientsand suspended particles(Prazan, 1994; Hein et al.,

1996), which could lead to a further eutrophicationin the floodplain or lead to higher nutrient trans-formation within certain river stretches. Carbon,fixed in algal biomass in connected side arms,could either serve as a basis for the riverinecommunity corresponding to export or lead toincreased sedimentation depending on the hydrau-lic regime in the side-arm(Amoros et al., 1987).The main aim of our study was to understand howthe hydrologic exchange conditions impact the roleof side-arms as sink or source of matter andmediate the transfer of nutrients and organic mat-ter. Within the restoration projects the applicabilityof hydrologic parameters and their predictivestrength were tested. The exchange conditionswere quantified by a ‘simple-to-estimate’ para-meter: the level of hydrological connectivity ex-pressed as the average days of integration in theflow regime and its predictive value for the eval-uation of restoration programmes are discussed.The hydrologic parameter ‘water age’, an adaptedmetric of residence time measuring water exchangein multi input systems, was used to demonstratethe control of biotic processes on a more detailedlevel in one side-arm(Baranyi et al., 2002).Therefore, the main objectives of the present studywere to determine:(i) nutrient and organic matterdynamics of side-arms in relation to the durationof integration in the riverine flow regime; and(ii)to evaluate the effect of water age on abiotic andbiotic processes in a restored side-arm.

In consequence for future restoration activities,these issues were used to predict what effects ofrestoration measures on nutrient and organic mattercycling can be expected.

2. Material and methods

2.1. Study area

The Danube is one of the main drainage systemsin Europe(817 000 km). The Danube in Austria2

drains 104 000 km and is of the ninth river order.2

The flow is characterized by an alpine regime withhighly variable and stochastic patterns. The meandischarge is 1900 m s (Q : 950 m s , Q :3 y1 3 y1

95 1

5040 m s ). Like all large rivers in the industri-3 y1

alized world(Petts et al., 1989; Ward, 1998), the

209T. Hein et al. / Science of the Total Environment 328 (2004) 207–218

ecology of the Danube has been considerablyaffected by changed land-use, by pollution andmost importantly by hydro-engineering(Schiemeret al., 1999). The 50-km river reachs downstreamof Vienna, although strongly impacted by regula-tion, represents one of the last remnants of alluviallandscape in Europe and was declared a NationalPark in 1996. Its importance has been describedin a number of papers(e.g. Tockner et al., 1998;Schiemer et al., 1999). Here, the key functionalattributes of floodplains—the hydrological dynam-ics, flood pulses and bed load transport—are par-tially operative.

2.2. Floodplain restoration measures

A large-scale restoration program was initiatedto restore the hydrological connectivity of tworeaches, Regelsbrunn(RB) and Orth(OR), alongthe free-flowing stretch of the Danube downstreamof Vienna. Both floodplain segments are dominat-ed by a former river channel that was severedupstream from the Danube after the main regula-tion of the river in the 19th century. Beforerestoration, flood pulses were characterized byshort and intense upstream surface connections, ofonly a few days during spates(more than 5000m s , 3–6 d a ). Seepage and groundwater3 y1 y1

from the river have supplied the segments abovemean water. Several weirs in the side-arms havedivided the water body into distinct basins(Heileret al., 1995).

The main goal of the restoration programmes inthe Alluvial Zone National park is to increase theupstream surface connection with the river and theconnectivity within the side-arm to enhance theduration of flowing conditions. The enhancementof connectivity with the Danube was establishedby lowering the riverside embankments and byadditional artificial openings in different inflowareas(Schiemer et al., 1999). Within the side-arm,the weirs have been lowered and equipped withlarger openings. In the RB reach, surface connec-tivity with the river was established at water levels0.5 m below mean water(MW , Austrian River85

Authority, unpublished report). The restorationprogramme for the Orth section of the Danuberepresents the second restoration step by building

larger and deeper inflow areas that finally enablea level of connectivity of 300 d a . In the presenty1

study, the first phase of these restoration measuresin Orth was analyzed, with a resulting surfaceconnection in the range of 46–160 d a .y1

2.3. Level of hydrological connectivity

Hydrological connectivity was defined as theaverage annual duration(days per year, 22 yearsmean) of upstream surface connection of the dif-ferent floodplain water bodies with the main chan-nel of the Danube River. Because this parameterprimarily depends on the flow pattern of the parentriver and the position of the floodplain waterbodies relative to river height(Hillman and Quinn,2002), it can easily be calculated from the riverhydrograph and the altitude of the inflow areas.The exact altitude(meter above sea level) of theinflow area and the water level data(station Orthriver km 1901.8, 01.01.1979–31.12.2001) wereprovided by the Austrian River Authority(Woe-sendorfer and Leberl, 1987). The investigatedyears 1991, 1995, 1996, 1999 and 2001 werecharacterized by high flows with at least one highwater event during the vegetation period in eachyear (Fig. 1c). The days of water levels abovemean water were 97 for 1991, 155 for 1995, 111for 1996, 166 for 1999 and 158 for 2001 betweenMarch and September in the respective years. Thesampling covered the periods of higher flows ineach year.

2.4. Water age

Water age was used as an adapted metric forresidence time within a multi-input side-arm sys-tem. It was defined as the time water has beencontained in the respective stretch in Regelsbrunn,up to any position within the system and at anypoint in time. For example, if water from the mainchannel is considered as having age 0(i.e. thewater has spent no time within the side-arm sys-tem) then the age of the water as defined aboveis some inverse measure of the ‘lotic’ character ofthe waters within the pools. Thus, low age impliesvery lotic ‘main channel-like’. This is because theimplied long time in the pools will have allowed


Fig. 1. (a) Location in Austria.(b) The floodplain of the Danube downstream of Vienna. Closed circles indicate the samplingstations in the floodplain and closed squares in the main channel. O1 to O6 and P2 to P6 were sampled before and after restoration.(c) Long term mean of Danube water level at gauge Orth(river km 1901.8). Mean"S.D. (grey lines) of the years 1980–2001 arepresented. Open circles indicate flood events(water level)515 cm) at any day in the 22-year period. Dashed lines indicate lowwater, mean water and annual high water level.

physical and biological processes significantly toalter the quality characteristics of the waters.Hence, water age is also an inverse measure of theconnectivity to the main channel, with low ageindicating high connectivity.

The software program ‘Regels 3.2’(Reckendor-fer and Steel, 2003) was used to calculate water

age for each sampling date in Regelsbrunn. Theprogram operates on a day to day basis using mainchannel to floodplain water level relationships aswell as water level to volume relationships foreach pool. It uses the volumes(derived from pooldepths and bathymetric survey data), the inflows(bank overflows, bank weirs, bank culverts and


Table 1Details for the hydrological categories. Sampling stations with the same level of connectivity and the total number of samples.Range of water level in the main channel and water temperature for all samples are presented

Integration(d a )y1 Sampling Number of Range of water Range of tempstations samples level(cm) (8C)

6 3 32 218–396 5.3–25.613 9 72 218–396 6–23.731 1 12 258–382 11.7–2646 5 33 258–382 8–21.2

160 3 9 267–337 13.8–16.6258 2 24 258–382 9–20365 2 34 218–396 6.3–19.8

ground water infiltration) and the outflows throughthe bottom of each pool to calculate water age atthe different pools. For each day, the change involume is calculated and the corresponding inflowsand outflows in the sequence of pools are proc-essed. Surface flows are calculated according tothe known height and size of inflow areas, theweir culvert dimensions and a weir coefficient(Cw) of 1.4. Groundwater infiltration from themain channel was estimated for periods beforerestoration when only infiltration drives the waterexchange between the main channel and side-armsystem. The difference between main channel andeach pool height, the length of infiltration and thedistance are the input variables. The riverine inflowentering RB exhibited an exponential relationshipwith the stage height of the river due to size andheight of the inflow areas.

2.5. Data collection

Sampling stations were located within the mainchannel and the floodplain reaches of RB and ORboth before and after restoration(Fig. 1). Themean duration of integration into the flow regimewas calculated for all stations(Table 1). The rangeof water levels and temperature showed no signif-icant difference between station and their connec-tivity level. The sampling in RB covered the years1991, 1995, 1996 and the year 1999(Heiler et al.,1995; Hein et al., 1999b, 2002). In OR the sampleswere taken in 1999 and 2001(Hein et al., 2003b).

Temperature, dissolved oxygen, conductivityand water levels were recorded in the field(port-able meters: WTW probes Series 330). Water was

glass-fibre filtered(APFyF, Millipore) within 3 hafter sampling for nutrient analysis(NH –N,4

NO –N, NO –N, N , SRP, P ) but was2 3 dissolved dissolved

used unfiltered for the determination of total phos-phorus(P ) and nitrogen(N ) (Golterman et al.,tot tot

1978). The particulate phosphorus and nitrogenwas estimated by subtraction of the totalydis-solved fraction. APFyF filters were used to deter-mine suspended particle concentrations(Hein etal., 1999b). The particulate organic fraction(POM) was determined as ash free dry weight.Chlorophyll-a concentration was analyzed accord-ing to (Lorenzen, 1967).

3. Results

3.1. Basic hydrologic properties of side-arms

The estimated duration of integration covered awide range of lotic situations for side-arms inregulated river systems. Periods of less than 1month were characterized by flowing conditionsonly during high water, whereas periods longerthan 160 d a refer to upstream connection closey1

to mean water. The relevance of the two side-armscan be demonstrated by the portion of dischargeflowing through them. In both side-arm systems,RB and OR, approximately 1% and up to 12% ofthe total discharge was drained through each siteat mean and high water, respectively.

3.2. The impact of hydrological connectivity

The mean values presented were obtained bysampling during the vegetation period at water


levels between approximately mean water andbankfull level(Table 1). The degree of hydrologicintegration (d a ) in the riverine flow regimey1

affected the mean geochemical situation and nutri-ent and particle concentrations in the side-arms(Table 2). Concentrations of conductivity and alka-linity were inversely related to the level of con-nectivity. Mean conductivity values decreasedsignificantly (Tamhane post hoc test,P-0.05)below 400mS cm at a duration of 160 d ay1 y1

(Fig. 2). Generally, nutrient concentrationincreased with the duration of integration andapproached the riverine values(Table 2). SRPvalues below 5mg l were measured at 30% ofy1

all samples, more often at stations with low con-nectivity levels(-47 d) at mean water("0.5 m)situations. More than 50% of these low SRP valueswere observed before the restoration measures.The SRP concentrations showed a significantincrease within the first 50 d a (Fig. 3). They1

inorganic nitrogen demonstrated a similar trend,nitrate tended to increase with a longer durationof integration than SRP. Low inorganic nitrogen(-100mg l DIN) concentrations were observedy1

only occasionally before restoration. The standarddeviation of conductivity, dissolved inorganicnitrogen and phosphorus decreased with longerduration of integration. Concentrations of particlesshowed significant relationships with duration ofintegration. The inorganic particle fractionincreased linearly towards the permanent flowingconditions(Fig. 4). Relationships of both particu-late phosphorus and particulate organic matter withduration of integration were best described by ahump-shaped curve with minimum values at amean duration of 3–5 months a(Fig. 5a,b). They1

mean chlorophylla (Chl a) concentration washighest at short integration times of 1 month andestablished stable values until permanent loticconditions(Table 2).

3.3. Water age as an eco-hydraulic determinant

Analyses of the hydrologic exchange conditionsand retention mechanisms in RB led to a time-integrated description of relevant physical andbiological processes(Fig. 6). Water age signifi-cantly explained the dynamics of inorganic nutri-

ents and particulate matter while passing theside-arm of Regelsbrunn. During high water per-iods, the input of nutrients and suspended matterdominated the side-arm system. Increasing waterage led to a rapid decline of nutrients and partic-ulate matter, while phytoplankton biomass devel-oped maximum values at water ages between 3and 10 d. At high water ages, a further decreaseof phytoplankton was evident in Regelsbrunn. Dueto the restoration, the frequency of low water ages(-1) increased five-times and the frequency ofoccurrence of medium water ages was reduced inRegelsbrunn.

4. Discussion

4.1. The effect of increased hydrologicalconnectivity

Generally, the hydrochemical conditions of thefloodplain segments approached the riverine situ-ation with increasing days of integration(Table2). The geochemical situation of the floodplainswas significantly different to the mean main chan-nel situation until 46 d a . The low frequency ofy1

upstream surface water input and the higher por-tion of seepage supply was responsible for thehigher values of geochemical parameters(Stein-inger, 2002). SRP and NH –N concentrations4

showed for all situations, means in the same rangeand a high variability in time. Potential limitationof SRP, indicated by SRP values below 5mg l ,y1

were found at situations with low connectivitylevels for 30% of all samples. The frequency ofobservations of these low values decreased afterrestoration, pointing to better resource availabilityat times of low water levels in restored side-armsystems. The biological utilization of SRP is ofmain importance as shown for the phytoplankton–SRP relationship in RB(Hein et al., 1999b). Theparticle concentrations increased towards the mainchannel, which was particularly true for the inor-ganic fraction at an integration of 258 d a .y1

Increasing connectivity led to higher particleconcentrations in side-arms, but did not raise thedissolved nutrient concentrations. Before restora-tion, nitrate and phosphate concentrations only10–25% lower than the corresponding main chan-

213T.

Hein

etal.

/Science

ofthe

TotalE

nvironment

328(2004)

207–218

Table 2Integration value and the mean"S.D. of alkalinity(Alk), conductivity(cond), SRP(soluble reactive phosphorus), P (particulate phosphorus), NH –N (ammoniapart 4

nitrogen), NO –N (nitrite nitrogen), NO –N (nitrate nitrogen), N (particulate nitrogen), PIM (particulate inorganic matter), POM (particulate organic matter), Chl-2 3 part

a (chlorophyll a)

Integration Alk cond SRP Ppart NH –N4 NO –N2 NO –N3 Npart PIM POM Chl-a(d) (mmol l )y1 (mS cm )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1 (mg l )y1

6 4.92"1.30 564"110 10"15 45"23 19"26 12"14 882"899 264"174 5.11"3.34 5.03"2.61 21"1613 3.68"0.46 458"67 16"20 46"21 27"35 21"10 1238"636 244"161 7.66"5.47 4.68"2.23 24"1831 4.19"0.35 531"39 16"19 46"14 28"29 25"5 1926"566 281"169 14.67"8.91 5.35"1.93 26"2546 3.63"0.67 437"64 13"13 42"17 25"29 15"7 951"541 190"114 15.83"9.71 4.03"2.32 18"12

160 2.72"0.16 351"18 21"8 33"20 30"21 8"2 1281"136 89"45 18.77"11.40 3.46"1.56 10"6258 2.97"0.19 370"25 17"9 66"34 82"30 25"5 1669"264 194"95 41.66"42.39 4.88"2.27 19"12365 2.86"0.25 365"39 23"13 77"48 57"41 18"6 2000"593 191"94 50.58"54.16 5.91"2.79 16"11


Table 3The independent(x) and dependent(y) variables used for regression analysis and the regression equations

Independent(x) Dependent(y) Function R value2 P value n

Integration(m a )y1 Cond(mS cm )y1 ys358q197,52=exp(y0.016=x) 0.81 P-0.05 7Integration(m a )y1 SRP(mg l )y1 ys7.43q2.27=ln(abs(x)) 0.61 P-0.05 7Integration(m a )y1 PIM (mg l )y1 ys5.5q0.12=x 0.84 P-0.01 7Integration(m a )y1 P (mg l )y1

part ys5.09y0.016=xq0.0001=x2 0.71 P-0.05 7Integration(m a )y1 POM (mg l )y1 ys46.27y0.109=xq0.0006=x2 0.83 P-0.05 7Water age ln(d) SRP(mg l )y1 ysy0.41q35.87=exp(y1.19=x) 0.60 P-0.01 60Water age ln(d) PIM (mg l )y1 ys11.4q83.48=exp(y7.8=x) 0.32 P-0.01 60Water age ln(d) Chl-a (mg l )y1 ys12.57q17.51=xy3.94=x2 0.20 P-0.01 60

Units of measurement are given in parentheses; m«months, a«year.Abbreviations: Conductivity(cond); SRP(soluble reactivephosphorus); P (particulate phosphorus); PIM (particulate inorganic matter); POM (particulate organic matter); Chl-a (chloro-part

phyll a).

Fig. 2. Relationship of duration of integration and conductivity(cond). Mean values and S.D. are plotted. Line indicates sig-nificant regression, see details in Table 3.

Fig. 4. Relationship of duration of integration and PIM(par-ticulate inorganic matter). Mean values and S.D. are plotted.Line indicates significant regression, see details of regressionin Table 3.

Fig. 3. Relationship of duration of integration and SRP(solublereactive phosphorus). Mean values and S.D. are plotted. Lineindicates significant regression, see details of regression inTable 3.

nel concentrations were frequently observed dueto the seepage supply with nutrient rich water fromthe main channel(Hein et al., 1999b; Tockner etal., 1999). The organic particle as well as theparticulate phosphorus concentrations can beexplained by higher portion of autochthonous pro-duction at less than 1 month of integration andhigh concentration of allochthonous organic matterat prolonged integration(Hein et al., 2003a,b).The phytoplankton biomass is not affected byprolonged integration and all mean biomass valuespresented are in the same range as in the mainchannel(Riedler and Schagerl, 1998). Neverthe-less the temporal variability of the chlorophyll-aconcentration is very high and impacted by both


Fig. 5. (a) Relationship of duration of integration and Ppart

(particulate phosphorus). Mean values and S.D. are plotted.Line indicates significant regression.(b) Relationship of dura-tion of integration and POM(particulate organic matter). Meanvalues and S.D. are plotted. Line indicates significant regres-sion, see details of regressions in Table 3.

Fig. 6. Relationship of water age(lnq1) and Chl-a (Chloro-phyll a), PIM (particulate inorganic matter)—broken line andSRP (soluble reactive phosphorus)—dotted line. Details ofregressions are in Table 3.

discharge conditions and season(Riedler et al.,unpublished data).

4.2. Water age reflecting the control of majorprocesses in riverine systems

The water exchange between the main channeland its side-arm and storage is quantified by thewater age for RB(Schiemer et al., 1999). Recentstudies have demonstrated the importance of waterage for the availability of inorganic nutrients andsuspended solids(Hein et al., 1999a), and bothplankton biomass and composition(Hein et al.,1999a; Baranyi et al., 2002) in Regelsbrunn.

The duration of integration and the amount ofwater flowing through the floodplain segment setsthe frame for the potential retention and transfor-

mation of organic matter. The biological response,the productivity, is shown by the increase inplankton biomass. At low water ages the dischargethrough RB increases up to 600 m s and thus,3 y1

the input and transport of nutrients and particlesdominate in the floodplain segment of Regels-brunn. Increasing water age and discharge below100 m s lead to an increase in phytoplankton3 y1

biomass. Higher water ages()20 d) during timesof low hydrologic exchange lead to an increase ofmetazooplankton biomass and thereby decrease thephytoplankton biomass by zooplankton grazing(Keckeis et al., 2003).

The frequent surface water exchange betweenside-arms and the main channel increases thepotential for plankton production and provides asignificant source of autochthonous organic matter,which can support riverine food webs(Thorp andDelong, 2002). Connected floodplains, with a res-ervoir or storage effect associated with POC pro-duction, can be of considerable importance in theprocessing and cycling of organic carbon for thewhole river network(Raymond and Bauer, 2001).Closely linked to an increase in organic matterproduction is also a higher rate of nutrient uptakeand more intense nutrient spiraling by planktonbiota within these side-arms(Tockner et al., 2000).


4.3. Days of integration and water age as tools topredict ecosystem functions

The aim of sustainable river restoration pro-grams is in general to improve the habitat availa-bility and restore key ecosystem functions(Meyer,1997). For the riverine landscape, the hydrologicexchange is a key process(Schiemer et al., 1999).Emphasis of the restoration measures along theAustrian Danube has been on the restoration ofthe surface water exchange of the floodplain seg-ments with the main channel. Nutrient transfor-mation, particle dynamics and autochthonousproduction in these segments and the interactionwith the main channel as important ecosystemfunctions are controlled primarily by the surfacewater exchange(Schiemer et al., 1999). For nutri-ent transformations, especially nitrogen, ground-water exchange is also likely to be of importance.

Within the presented restoration projects, animportant aim was to develop also eco-hydraulicpredictors for ecosystem functions, which can beassessed during the construction planning(Schie-mer et al., 2001). Identified, indicative eco-hydrau-lic parameters are days of integration, estimatedsimply by the height of the inflow areas and waterage quantifying the riverine input and the storagewithin the side-arm. Both eco-hydraulic parameterscan be used for future restoration measures offeringthe possibility to predict changes in the floodplainsegment and also estimating the effects of anymeasure for the main channel. This informationprovides the basis for evaluating local changes andthe value for the whole river stretch.

4.4. Restoration efforts

In large regulated river systems, restorationefforts try to increase structural elements with ahigher hydrologic retention or an intense hydrolog-ic exchange with the main channel on differentscales. The presented side-arm reconnections pro-vide an example of restoration on the reach scalewhere rare elements of the former braided reachwere introduced again. In both side-arms anincrease of the connectivity levels was achievedby the restoration measures. The main water supplyof each floodplain segment shifted from a seepage

to a surface water dominated system afterrestoration.

Restoring the hydrological connectivity andincreasing the duration of lotic conditions maintaina higher overall productivity since a balancebetween the retention, transformation and exportof organic matter is established. Especially thespatial configuration and temporal availability ofside-arms over a broad range of discharge condi-tions contribute to the overall ecosystem function-ing and increase also species diversity as shownalso for restoration works at side-arms of theRhone River(Henry et al., 2002). Comparing theresults in RB to the pre-restoration conditionsdescribed in Tockner et al.(1999), during loticphases with maximum plankton production, theexport of more labile POM to the main channelwould be expected(Hein et al., 2003a,b). Con-nected side-arms increase, therefore, the totalorganic matter retention and transformation inregulated rivers and can be of considerable impor-tance for the support of riverine biota as alsoshown for inshore retention structures(Schiemeret al., 2001).

5. Conclusion

Restoration concepts for large river–floodplainsystems require a profound insight into ecologicalfunctioning. The evaluation of restoration measurescan be improved by the development of eco-hydraulic predictors, which are easy to estimateduring the planning phase and have a high poten-tial to predict the changes on important ecosystemfunction like matter processing and organic matterproduction.

The mean duration of integration, as one ofthese eco-hydraulic predictors, was significantlyrelated to geochemical conditions, nutrients andparticle concentration in side-arms. A level ofconnectivity higher than 46 d a leads to lowery1

conductivity levels and higher nutrient concentra-tions due to the shift of the dominating source toriver water. The shift from dominating seepagesupply to integration in the riverine flow is indi-cated by the increase of suspended solids concen-trations in the side-arms. The phytoplanktonbiomass shows the highest mean values at a


duration of integration of 1 month a andy1

decrease with increasing duration. The presentedrelationships point to a more ‘main channel like’hydrochemical situation in the side-arms, with amedium level of phytoplankton biomass andincreased autochthonous carbon export to the mainchannel(Hein et al., 2003a).

On a more detailed level, water age demon-strates the temporal patterns of riverine input, thedevelopment of plankton production and the shiftbetween hydrological and biological control ofphytoplankton vs. riverine flow in a side-arm.

Both eco-hydraulic parameters can be used topredict the functional response of re-connectedside-arms. The change in water exchange in theside-arms determines the function as a sink orsource for matter of a particular side-arm after therestoration measures. The parameter water agedescribes the temporal dimension of matter reten-tion and plankton production in a side-arm.

Acknowledgments

This work was funded by the Austrian ScienceFund(grant� P11720 bio), by the Austrian RiverAuthority (in Regelsbrunn the project ‘Gewaes-servernetzung Regelsbrunn’) and by the NationalPark Authority (Life98NATyAy005422 in Orth).A. Aschauer, G. Heiler, C. Holarek, H. Kraill, P.Riedler and M. Schagerl contributed to the datacollection and analysis. The authors would like tothank T.J. Battin and two anonymous reviewersfor their comments on an earlier version of themanuscript.

References

Amoros C, Roux AL, Reygrobellet JL, Bravard JP, Pautou G.A method for applied ecological studies of fluvial hydro-systems. Regul Rivers Res Manage 1987;1:17–36.

Baranyi C, Hein T, Holarek C, Keckeis S, Schiemer F.Zooplankton biomass and community structure in a DanubeRiver floodplain system: effects of hydrology. FreshwaterBiol 2002;47:473–482.

Bradshaw AD. Underlying principles of restoration. Can. J.Fish. Aquat. Sci. 1996;53:3–9.

Fisher SG, Grimm NB, Martı E, Holmes RM, Jones JB Jr.´Material spiraling in stream corridors: a telescoping ecosys-tem model. Ecosystems 1998;1:19–34.

Golterman HL, Clymo RS, Ohnstad MAM. Methods forphysical and chemical analysis of freshwater. IPB handbook.Oxford: Blackwell Scientific, 1978.

Hedges JI, Mayorga E, Tsamakis E, McClain ME, Aufdenkam-pe A, Quay P, Richey JE, Benner R, Opsahl S, Black B,Pimentel T, Quintanilla J, Maurice L. Organic matter inBolivian tributaries of the Amazon River: a comparison tothe lower mainstream. Limnol Oceanogr 2000;45:1449–1466.

Heiler G, Hein T, Schiemer F. Hydrological connectivity andflood pulses as the central aspects for the integrity of ariver–floodplain system. Regul Rivers Res Manage1995;11:351–361.

Hein T, Schagerl M, Heiler G, Schiemer F. Chlorophyll-a andhydrochemical dynamics in a backwater system of theDanube, controlled by hydrology. Arch Hydrobiol Suppl1996;113:463–470.

Hein T, Baranyi C, Heiler G, Holarek C, Riedler P, SchiemerF. Plankton communities controlled by hydrology in twofloodplain systems and the River Danube. Arch HydrobiolSuppl Large Rivers 1999a;11:439–452.

Hein T, Heiler G, Pennetzdorfer D, Riedler P, Schagerl M,Schiemer F. The Danube restoration project: functionalaspects and planktonic productivity in the floodplain system.Regul Rivers Res Manage 1999b;15:259–270.

Hein T, Baranyi C, Reckendorfer W. Bedeutung der hydrolo-gischen Vernetzung fur die hydrochemische Situation der¨Augewasser im Bereich OrthyDonau.(Importance of hydro-¨logical connectivity for the hydrochemical situation of thebackwaters at OrthyDonau in German with abstract inEnglish), NP Donauauen, Vienna, 2002, p. 52.

Hein T, Baranyi C, Herndl GJ, Wanek W, Schiemer F.Allochthonous and autochthonous particulate organic matterin floodplains of the River Danube: the importance ofhydrological connectivity. Freshwater Biol 2003a;48:220–232.

Hein T, Baranyi C, Reckendorfer W. Einfluss von Offnungs-¨

maßnahmen auf die hydrochemische Situation und dieplanktischen Prozesse in einem dynamischen Ausystem.(Impact of re-connection on the hydrochemical conditionsand plankton in a dynamic backwater in German). VerhZoo Bot Ges 2003b; in press.

Henry CP, Amoros C, Roset N. Restoration ecology of riverinewetlands: a 5-year post-operation survey on the RhoneRiver, France. Ecol Eng 2002;18:543–554.

Hillman TJ, Quinn GP. Temporal changes in macroinvertebrateassemblages following experimental flooding in permanentand temporary wetlands in an Australian floodplain forest.River Res Appl 2002;18:137–154.

Keckeis S, Baranyi C, Hein T, Holarek C, Riedler P, SchiemerF. The significance of zooplankton grazing in a floodplainsystem of the River Danube. J Plankton Res 2003;25:243–253.

Lorenzen CJ. Determination of chlorophyll and phaeopig-ments: spectrometric equations. Limnol Oceanogr1967;12:343–346.


Meyer JL. Conserving ecosystem function. In: Pickett STA,Ostfeld RS, Shachak M, Likens GE, editors. The ecologicalbasis of conservation heterogeneity, ecosystems and biodiv-ersity, vol. 136–145. New York: Chapman and Hall, 1997.

Petts GE, Moller H, Roux AI. Historical changes in largealluvial rivers: Western Europe. Chichester: Wiley, 1989.

Prazan H. Zur Ermittlung der Schwebstoffuhrung der Donau¨in Osterreich.(Determining suspended sediment transport¨

in the River Danube in Austria in German). Oesterr Wasser-und Abfallwirtschaft 1994;7–8:181–194.

Raymond PA, Bauer JE. Riverine export of aged terrestrialorganic matter to the North Atlantic Ocean. Nature2001;409:497–500.

Reckendorfer W, Keckeis H, Winkler G, Schiemer F. Zoo-plankton abundance in the River Danube, Austria: thesignificance of inshore retention. Freshwater Biol1999;41:583–591.

Reckendorfer W, Steel A. Auswirkungen der hydrologischenVernetzung zwischen Fluss und Au auf Hydrologie, Mor-phologie und Sedimente.(Effects of hydrological connectiv-ity on hydrology, morphology and sediments in German).Verh Zoo Bot Ges 2003; in press.

Riedler P, Schagerl M. Pelagic primary production and relatedparameters in the River Danube near Vienna(Austria). ArchHydrobiol Suppl 1998;115:139–151.

Schiemer F, Baumgartner C, Tockner K. Restoration of flood-plain rivers: the Danube restoration project. Regul RiversRes Manage 1999;15:231–244.

Schiemer F, Keckeis H, Reckendorfer W, Winkler G. The‘inshore retention concept’ and its significance for largerivers. Arch Hydrobiol Suppl 2001;135:509–516.

Steininger A. Hydrological connectivity between surface andsubsurface systems in riverine floodplains—a key factor incontrolling groundwater nutrient dynamics and the structureof hypogean animal assemblages. M.Sc. Thesis Universityof Vienna, Vienna, 2002.

Thorp JH, Delong AD. Dominance of autochthonous auto-trophic carbon in food webs of heterotrophic rivers. Oikos2002;96:543–550.

Tockner K, Schiemer F, Ward JV. Conservation by restoration:the management concept for a river–floodplain system onthe Danube River in Austria. Aquat Conserv Mar FreshwaterEcosyst 1998;8:71–86.

Tockner K, Pennetzdorfer D, Reiner N, Schiemer F, Ward JV.Hydrological connectivity, and the exchange of organicmatter and nutrients in a dynamic river–floodplain system.Freshwater Biol 1999;41:521–535.

Tockner K, Malard F, Ward JV. An extension of the floodpulse concept. Hydrol Proc 2000;14:2861–2883.

Ward JV. Riverine landscapes: biodiversity patterns, distur-bance regimes and aquatic conservation. Biol Conserv1998;83:269–278.

Woesendorfer H, Leberl S. Uferzonen der Donau von Wienbis zur Marchmundung.(Shore zones of the Danube¨between Vienna and the confluence with the Morava. inGerman) Austrian River Authority, 1987.

the impact of surface water exchange on the nutrient and particle dynamics in side-arms along the...

Documents