Effects of an intertidal sediment fence on sediment elevation and vegetation distribution in a Venice (Italy) lagoon salt marsh

Ecological Engineering 16 (2000) 223–233

Effects of an intertidal sediment fence on sediment elevationand vegetation distribution in a Venice (Italy) lagoon salt

marsh

Francesco Scarton a,*, John W. Day Jr b, Andrea Rismondo a,Giovanni Cecconi c, Daniele Are a

a SELC, Inc., Viale Garibaldi 50, 30173 Venezia-Mestre, Italyb Coastal Ecology Institute and Department of Oceanography and Coastal Sciences, Louisiana State Uni6ersity, Baton Rouge,

LA 70803, USAc Consorzio Venezia Nuo6a, Palazzo Morosini, S.Marco 2803, Venice, Italy

Received 2 April 1999; received in revised form 5 November 1999; accepted 13 December 1999

Abstract

On a intertidal flat in the lagoon of Venice (Italy), the effects of a fence on sediment elevation and vegetationestablishment were studied throughout the years 1994–1997. With the use of a sedimentation erosion table (SET) wemeasured 5.7 cm of accumulated sediment in the protected tidal flat after 28 months (2.5 cm/year), compared with−0.7 cm (−0.3 cm/year) in a nearby, unprotected tidal flat. After a storm which damaged part of the fence, therewas a similar loss in elevation in both tidal flats; following repair, only the protected tidal flat gained elevation. After1 and 3 years, vegetation coverage (mainly due to Salicornia 6eneta, Sarcocornia fruticosa and Atriplex portulacoides)was higher along the edge of the salt marsh of the protected tidal flat compared to the control salt marsh, butdifferences were not significant. © 2000 Elsevier Science B.V. All rights reserved.

Keywords: Lagoon of Venice; Sedimentation; Erosion; Fence; Restoration

www.elsevier.com/locate/ecoleng

1. Introduction

Rising sea level is presently leading to loss ofcoastal wetlands in many estuaries and deltas, inparticular in those areas where there is a high rateof relative sea level rise due primarily to subsi-

dence (such as the Mississippi delta, Day et al.,1997; and the Rhone delta, Ibanez et al., 1999).Soil surface elevation relative to mean local waterlevel is the one of the most important factorscontrolling the colonization, maintenance or dete-rioration of intertidal vegetation (McKee andPatrick, 1988; Cahoon et al., 1995). Loss of rela-tive elevation in salt marshes leads to higher floodduration, less frequent drainage, soil anoxia and

* Corresponding author. Tel.: +39-41-610-482.E-mail address: [email protected] (F. Scarton).

0925-8574/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.

PII: S 0925 -8574 (00 )00045 -8

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233224

finally to the death of the emergent vegetation(Mendelsshon and McKee, 1988). Generally, onlythree processes which affect coastal elevation havethe potential to increase surface elevation relativeto mean water level (Boumans et al., 1997): depo-sition of suspended sediments, deposition of or-ganic matter from above ground plant productionand expansion and incorporation of belowground plant production. Any effective restora-tion technique must either reduce the submer-gence potential or enhance the processes that canincrease elevation. Intertidal sediment fences aredesigned to increase the efficiency of trappingsediments on unvegetated tidal flats, in order toraise elevation and to allow for colonization byintertidal vegetation. In The Netherlands, one ofthe first countries in which this technique wasadopted, sediment fences were built in the last 30years, and accretion rates of up to 3.5 cm/yearhave been measured following colonization byplants (Kamps, 1962; Glopper, 1981; Bouwersmaet al., 1986). Fences have been used in Germany(Lieberman et al., 1997) and in Louisiana as well,in more than 50 locations, using recycled Christ-mas trees (Boumans et al., 1997; Coalition toRestore Coastal Louisiana, 1989).

In the lagoon of Venice, a study on the effect ofsea level rise on salt marshes began in 1993 (Dayet al., 1999, 1998a,b). In the framework of thatstudy a sediment fence was built to test its effec-tiveness under local tidal and suspended sedimentconditions, in raising the level of the sedimentsurface and encouraging vegetation growth. Inthis paper we report on the results after 4 years ofmonitoring.

2. Study area

Venice Lagoon is a large (surface area is about550 km2) shallow coastal lagoon located on thenortheastern coast of the Adriatic Sea (about45°N 12°E). The lagoon originated nearly 6000years ago when rising sea level flooded the low-land coasts of the Adriatic Sea (Gatto and Car-bognin, 1981). There are two barrier islands whichseparate the lagoon from the sea and water isexchanged through three large inlets (Fig. 1). Over

the past several centuries, the lagoon has beengreatly altered resulting in important changes insediment dynamics. The Brenta and Piave rivers,which previously flowed into the lagoon, havebeen diverted into the sea to the south and northof the lagoon ecosystem. At present, only onesmall river, the Dese, discharges to the lagoon;although considerable agricultural drainage flowsinto the lagoon. Thus, riverine sediment input tothe lagoon has been almost completely eliminated.Long jetties constructed in the inlets during thelast century have reduced the import of marinesediments into the lagoon. As a result of thesealterations, there is a net loss of about 1.1×106

m3/year of sediments from the lagoon to the sea(Bettinetti et al., 1996). Most of the lagoon area isoccupied by an open waterbody (about 400 km2)which is partially vegetated by macroalgae andseagrasses. The mean depth of the lagoon is 1.1 mand the tidal range during spring tides is about 1m. There are extensive intertidal salt marshes,especially in the southwestern and northern por-tions of the lagoon. Dominant marsh species in-clude Limonium narbonense, Salicornia 6eneta,Sarcocornia fruticosa, Atriplex portulacoides, Puc-cinellia palustris, Spartina maritima and Juncusmaritimus (nomenclature follows Caniglia et al.,1997). Salt marsh area in the lagoon has fallenfrom about 12 000 to about 4000 ha between1900 and the present (Favero, 1992) due to recla-mation, erosion, and natural and human-inducedsubsidence.

The study site is a tidal flat area located adja-cent to a salt marsh on the western edge of thelagoon adjacent to Marco Polo airport (Fig. 1).The marsh edge is being eroded by wind-gener-ated waves. Vegetation at the site is mostly com-posed of P. palustris, S. fruticosa, L. narbonense,and J. maritimus. The marsh shoreline is charac-terized by two very similar small intertidal embay-ments which have a length along the shore ofabout 140 m and a width perpendicular to theshore of about 60 m. The mean water depth of thetidal flat in the vicinity of the fence is about 0.3 mand the elevation of the sediment surface is be-tween −0.10 and −0.20 m msl. By comparison,the mean elevation of the marsh behind the tidalflat is 0.38 m msl.

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233 225

The two tidal flats are similar with respect toexposure to wind and waves. The average depthof the lagoon in which waves are generated whichaffect the study site is 1.1 m. The marsh edge isexposed to wave attack generated when windsblow over about 4 km from a southerly directionand about 3 km from the east. Because of this,waves are not fetch-limited and during stormsstrong waves can be generated. In winter, stormswith strong northeast winds up to 80–90 km/h orhigher called Bora occur on average of about

once a month and winds blow for 2–3 days. TheBora winds depress water levels in the northernAdriatic and Venice Lagoon so that the marshesare flooded less often and the mean water levelcan be 30–50 cm lower in the northern lagoonthan in the southern lagoon. Because of the low-ered water level, waves generated by the Borawinds more often attack on the marsh edge belowthe level of vegetation. During the spring and fall,southern winds as high as 50–60 km/h called theScirocco lead to elevated water levels in the north-

Fig. 1. Location of study area in the Lagoon of Venice; salt marshes are in gray.

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233226

Fig. 2. Cross section of the fence built in the study area, with mean tide levels.

ern Adriatic and extensive flooding of marshes inVenice Lagoon. Waves generated during theseperiods often propagate over the marsh surfacedepositing suspended and floating material. Thickracks of debris deposited after storm activity wereobserved. The strong wave activity is causingactive frontal erosion along the exposed edge ofthe marsh vegetation at the site.

3. Methods

A sediment fence was constructed across one ofthese embayments in May 1994, following thedescription of Boumans et al. (1997). It consistedof two parallel rows of chestnut posts containingseveral layers of bundles of vegetation (Fig. 2).There were 95 posts in each row with a spacing of1.5 m, and which were driven in the sediment to adepth of about 1.2 m and extended above thesediment surface about 1.0 m. The total length ofthe fence is approximately 140 m. The posts in thetwo rows were offset from one another by 0.75 mto increase the strength of the fence. The vegeta-tion bundles were constructed of branches andstems of willow Salix spp. and popular Populusspp. 1–4 cm in diameter tied together into bun-dles about 25 cm in diameter and 3 m long. Eachlayer consisted of two rows of bundles, also offset,

which were placed between the rows of posts.There were an average of three layers of bundlesin the fence which were tied securely with nylonrope 6 mm in diameter. This structure of the fencewith the bundles has considerable open space andallows the easy passage of water and suspendedsediments but significantly lowers wave energy(Boumans et al., 1997).

At the end of October 1996 (after the Octoberelevation measurement, see below), there was astrong Scirocco storm with high winds (up to 90km/h) and high waves which lasted for about 3days. Waves associated with the storm causedserious damage to the fence, along about twothirds of its length. The posts remained in placebut most of the vegetation bundles were washedout and stranded on the beach behind the fence.This event left the southernhalf of the fenced areaalmost completely unprotected from wave attack.This condition persisted until after measurementswere taken in February 1997. The fence was re-built at the end of April 1997.

Surface elevation changes were measured usinga sedimentation erosion table (SET) developed forhigh precision measurements of surface elevationchanges in wetlands and shallow waters up toabout 30 cm depth (Boumans and Day, 1993).The SET has an accuracy of about 90.2 cm. TheSET consists of a supporting pipe which is perma-

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233 227

nently installed in the field and a removable armwhich can be leveled and from which distance tothe surface of the mud is measured. The supportpipe is driven into the sediment surface untilrefusal and at the sites, the pipes were driven inapproximately 2 m. Two SET stations were ran-domly established on the tidal flat behind thesediment fence (Tessera 2a and 2b) and two sta-tions were randomly established in the controlarea without a sediment fence (Tessera 4a and 4b;Fig. 1). A small platform, close to the pipe, fromwhich measurements were made was constructedat each SET location to minimize disturbance ofthe sediment surface during sampling. Set mea-surements were taken approximately each 3months from June 1994 to November 1997. Apaired t-test was used to determine if there was asignificant change in elevation from the beginningto the end of the measurement period for each ofthe sites, and a t-test was used to detect significantdifference in elevation between Tessera 2 and 4 atthe end of the measurement period. A Spearmanrank correlation test was used to determine if thetrend in changes in elevation were significant overtime.

Observations on the vegetation occurring alongthe edge of both embayments were done in Octo-ber 1995. This is the period of maximum biomassfor the species present (Scarton et al., 1998, 2000).Sixteen random points per embayment were cho-sen, and the species and their surface coverage in

a radius of 1 m were recorded. In order to deter-mine the change in location of the marsh edge, inJuly 1997, the distance between the non-vegetatederosional edge of the marsh and the line of contin-uous vegetation were measured at 14 randomlychosen locations. Since values were not normallydistributed, percentage coverage values were arc-sin transformed and distances log10 transformed(Zar, 1996).

In the summer of 1996, five samples of superfi-cial sediment (5 cm in depth) were randomlycollected in each of the fenced and unfenced area.Grain size analysis was determined according tostandard soil analysis procedures. Bulk densityand organic matter content, calculated as theweight loss at 550°C, were also determined on thesamples. Differences between fenced and unfencedareas for these parameters was investigated usinga Student t-test.

4. Results

There was a consistent accumulation of sedi-ment behind the fence leading to an increase inmean elevation while in the unfenced area, eleva-tion change was much less and there was noconsistent trend (Fig. 3). During low tide, theaccumulated material was clearly visible on theshore side of the fence.

Fig. 3. Elevation changes for stations located behind the fence (cumulative sedimentation erosion table (SET) values for station 2aand 2b; mean, n=72, with 9S.E. is reported) and without the fence (cumulative SET values for stations 4a and 4b; mean, n=72,with 9S.E. is reported).

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233228

Fig. 4. Mean elevation changes for stations 2a and 2b (above; n=36 for each station, mean9S.E. is reported) and for stations 4aand 4b (below; n=36 for each station, mean9S.E. is reported).

In the first 6 months the fenced area (mean ofst. 2a+2b; Fig. 3) showed an elevation increaseof 2.991.9 cm, compared to an elevation of only0.590.8 cm in the control area (st. 4a+4b; Fig.3). After the first year, 3.9 cm of sediment hadaccumulated in the fenced area, with a change ofonly +0.2 cm in the unfenced area. After thisdate the elevation gain slowed down, with rates ofabout 3 cm/year observed until April 1996 (22months after the first measurement); in October1996 the observed rate was 2.5 cm/year. From theinitial measurement to October 1996, 5.7 cm hadaccumulated in the fenced area, whereas there wasa loss of 0.7 cm in the unfenced area. In theformer, the observed increase in elevation fol-lowed a significant trend (rs= 1, PB0.01),whereas in the latter the apparent decreasingtrend was not significant (rs= −0.83, n.s.) indi-cating irregular fluctuations over the first 22months. The initial elevation gain occurred

mainly at the SET pipe nearest the marsh edge (st.2a) while the station further from the marsh edge(st. 2b) gained elevation more slowly (Fig. 4). Thetwo pipes in the control area (4a and 4b) behavedin a very similar manner (Fig. 4).

After the storm in October 1996 (which dam-aged the fence) and before repairing the fence,there was a similar average elevation loss in bothareas; 1.7 cm in the fenced area and 1.8 cm in thecontrol site (Fig. 3). Because of the direction ofthe winds, the southern part of the marsh edge inthe fenced area was more strongly affected bywaves. Because of this, the elevation loss wasmuch more pronounced in station 2a (Fig. 4)which was located closer to the marsh edge in thesouthern part of the area. This is also the areawhere the fence was damaged. The two measure-ments taken after the fence was repaired showthat on average, there was no elevation change.No strong storms occurred between June and

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233 229

November 1997, which would deliver a strongsediment pulse to the area. When considered indi-vidually (Fig. 4), station 2b which is more in thecenter of the fenced area and in the area wherethe fence was less protected gained almost 3 cmbetween June 1997 and November 1997, whilestation 2a near the marsh edge continued to looseelevation.

Overall, at the protected tidal flat there was anelevation increase of 2.3 cm from the beginning(paired t-test, t=8.3, 71 d.f., PB0.001), althoughwithout a significant trend over the 41 months ofobservation (rs= −0.31, n.s.); at the other, theelevation loss was 4.5 cm (paired t-test, t=33.6,71 d.f., PB0.001) with a clear decreasing trend(rs= −0.93, PB0.001). The difference betweenthe two observed final values of elevation change(+2.3 and −4.5 cm) is highly significant (t=22.6, 142 d.f., PB0.001).

Considering the layer of sediment trapped inthe protected area (5.7 cm) and that lost in theunprotected one (0.7 cm), there was a net gain of6.4 cm before the storm. The fenced area wasabout 1500 m2. Therefore, about 100 m3 of sedi-ments were trapped in the fenced area in 28months. It is also interesting to consider howmuch sediment would have been trapped if thefence had survived the storm. If we assume thatthe fence would have prevented sediment loss,then the volume of sediment trapped would havebeen about 112 m3.

Grain size analysis did not show any significantdifferences between bottom superficial layers in

the area protected by the fence and the unpro-tected one (Fig. 5). Mean percentages for sand,silt and clay for both sites were 43, 55 and 2%,respectively. Organic matter content was slightlyhigher in the fenced tidal flat (3.5 vs. 2.5%), aswas the bulk density (0.88 vs. 0.79 g/cm3); none ofthese differences were significant (Mann–WitneyU-test, P\0.05 in both cases). These results indi-cated that the fence did not result in a selectivetrapping of a specific sediment particle size.

Vegetation coverage was higher along the edgeof the fenced bay (29.3 against 18.7%) but thedifference was not significant (Student t-test, 30d.f., t=1.0, P\0.05). There were more speciespresent in the fenced area; S. fruticosa, A. portula-coides and S. 6eneta were present. In the unfencedarea, only S. 6eneta was present. No vegetationgrowth has occurred in either the protected tidalflat or in the unprotected one. The distance be-tween the edge of the salt marsh and the vegeta-tion was 7.5 m93.3 along the protected bay and9.1 m92.9 in the exposed one; again, this differ-ence was not significant (Student t-test, 26 d.f.,t=1.4, P\0.05).

5. Discussion and conclusions

The results show that the sediment fence causeda rapid increase in sediment accumulation whichled to an increase in the elevation of the sedimentsurface of nearly 6.0 cm over the first 2 years ofthe study. The trend was that vegetation coverage

Fig. 5. Grain size analysis of sediment samples from the tidal flat protected by the fence and from the unprotected (n=5 for each,mean9S.E. is shown).

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233230

and number of species were higher in the fencedarea. It was believed that the enhancement ofsediment accumulation was a result of decreasedwave energy which increased sediment deposition,reduced resuspension, and enhanced consolida-tion of the surface sediments (Anderson et al.,1981; Nagai et al., 1984; Ward et al., 1984; Shy-bayama et al., 1986; Boumans et al., 1997). Theloss of sediment due to the storm was probably aresult of erosion of the bottom and export ofsuspended sediment. There may also have beensome fluid mud near the bottom which flowedback through the fence.

The fence was effective in rapidly increasing soilelevation in the protected tidal flats. The valuesobserved for the first 2 years of study (between 2.5and 4.9 cm/year) are in agreement with thosereported by Boumans et al. (1997) for the Missis-sippi delta. They measured sediment elevationincreases of 1.7–3.3 cm/year in areas protected byfences very similar to those here. A marked eleva-tion loss was measured in the control area withoutthe fence; more than that reported by other au-thors. Boumans et al. (1997) reported similar re-sults with no change to slight elevation loss inunprotected areas. Low elevation changes havealso been reported by Childers et al. (1993) forseveral mid-Atlantic estuaries in the US.

The accretion values in the fenced area in theVenice lagoon are less than the high accretionrates on the order of 10 cm/year reported by someauthors for sediment fences and other types ofsediment trapping structures (Glopper, 1981;Stevenson et al., 1985; Bouwersma et al., 1986;Dijkema et al., 1990). An apparent equilibriumlevel, about 5 cm higher that at the beginning,was reached after 2 years at the study area. Bou-mans et al. (1997) also reported that an equi-librium level was reached behind the sedimentfences in the Mississippi delta.

Wright et al. (1997) reported that that large fishtraps, constructed of a network of interwovenmangrove sticks embedded in the sediment in theKosi estuary in South Africa, decreased currentvelocities and trapped sediments. These traps haveresulted in large areas being converted from inter-tidal flats to salt marsh or mangrove. These fishtraps thus act in the same manner as the sedimentfences constructed in Venice Lagoon.

With the measured increase in sediment surfaceelevation, vegetation establishment did not takeplace in the protected tidal flat. In order to deter-mine when vegetation might become established,the elevation of the site was compared to eleva-tions of different salt marsh plant associations inVenice lagoon. Pignatti (1966) described severalsuch vegetative associations correlated with eleva-tion for Venice Lagoon wetlands. For the eleva-tion range between the mud flat and the marsh atTessera (−0.05 to +0.38 m msl), there are fourvegetation associations. Pignatti described a Puc-cinellia-Arthrocnemum (now Sarcocornia) from0.25 to 0.40 m msl, a Limonium-Puccinellia asso-ciation from 0.15 to 0.30 m msl, a Limonium-Spartina association from 0.05 to 0.20 m msl, anda Salicornia spp. association from 0.05 to 0.10 cmmsl. Elevation data recorded by our group in1998 (unpubl. obs.) were in good agreement withthose from the literature; a sketch of a transectacross the Tessera salt marsh is shown in Fig. 6.Although there is considerable elevation overlapamong these associations, there is a general ten-dency for the vegetation to shift towards moreflood tolerant plants as marsh elevation decreases.Before the sediment fence was installed, the eleva-tion of the tidal flat in the vicinity of the sedimentfence ranged between −0.11 and −0.13 m msl.Before the fence was damaged, there was a netgain of about 6 cm, thus resulting in an elevationbetween −0.05 to −0.08 cm msl. Therefore, anadditional 10–13 cm must be added to achieve anelevation where S. 6eneta will begin to colonizethe flat. It seems that an equilibrium elevation ofthe mud flat has been reached at the present fenceheight. Observations at the site indicate that thefence elevation is slightly below the level of meanhigh tide. This information suggests that raisingthe fence by 10–20 cm (or about one vegetationbundle) would probably lead to a sediment eleva-tion increase sufficient to lead to vegetation estab-lishment immediately behind the fence within 2–3years. The tidal flat slopes up to an elevation ofabout −0.05–0 m msl at the base of the ero-sional face. If elevation increase is similar in thisarea to that at the SET stations, vegetation estab-lishment will likely occur in 2–3 years. Oncevegetation is established, elevation increase should

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233 231

Fig. 6. Vegetation transect across the Tessera salt marsh; horizontal and vertical axis are not on the same scale. For each of the mostcommon species data on elevation asl were obtained in 1998.

increase due to the role of the plant roots informing organic soil as well as a further reductionof wave energy due to the plants themselves(Knutson et al., 1981, 1990; Benner et al., 1982).These factors show the importance of taking intoconsideration both the rate of elevation increaseand initial elevation when estimating the time tovegetation establishment.

Since SET data integrate accretion and shallowsubsidence (Cahoon et al., 1995), accretion can becalculated if the rate of elevation change andshallow subsidence are known. Shallow subsi-dence is defined as the compaction of the first fewmeters of the substrate. The mean shallow subsi-dence was calculated for a number of marsh sitesin the lagoon to be 0.25 cm/year (Day et al.,1998a). If it was assumed that the tidal flat has asimilar value for shallow subsidence, the rate ofaccretion in the protected tidal flat ranges between1.1 and 4.2 cm/year (considering the minimumand maximum observed annual rates) while thatin the unprotected area is between −0.5 and−1.7 cm/year, which indicates it is experiencingnet erosion. Sedimentation rates over a long time(i.e. about a century) for bottoms or tidal flats areknown only for a few bottom cores in the Lagoonand have a maximum value of 0.7 cm/year(Pavoni et al., 1987).

There was a trend of reduced distance betweenthe edge of the marsh platform and vegetation inthe fenced area, although the difference in vegeta-

tion distance from the edge was still not signifi-cant after 3 years, due to high variability.Boumans et al. (1997) reported that vegetationinvaded a fenced area in Louisiana. It has beenshown that exposure of intertidal mud flats andmarshes to wave action leads to reduced seedlinggermination and to physical damage to thoseyoung plants which do germinate (Van Eerdt1985; Foote and Kadlec, 1988).

The information developed in this study showsthe importance of sediment fences in both increas-ing elevation and in protecting the salt marsh edgefrom storm caused erosion. The study alsodemonstrates the need to maintain the fence. As amanagement tool, sediment fences in tidal flatscan be very effective if placed in areas with mod-erate levels of suspended sediment and shallowelevation. These two factors will ensure rapidelevation gain and vegetation establishmentwithin a few years. Fences refurbishment andrepair is necessary every 2–3 years but given thelow cost and relative ease of doing this, the effortinvolved in maintaining the fences is not exces-sive. This also shows the importance of site selec-tion. A site with sufficient suspended sedimentsand a shallow depth can become vegetated afterone maintenance cycle. Nowadays, five additionalfences for a total length of about 1100 m havebeen built in the lagoon of Venice which wouldeventually lead to significant reduction of edgeerosion as well as vegetation establishment.

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233232

Acknowledgements

This work was supported by the the ConsorzioVenezia Nuova, on behalf of the State WaterAuthority of Venice, the Magistrato alle Acque diVenezia. We thank D. Smania for help with fieldoperations and D. Tronchin for drawing the vege-tation transect.

References

Anderson, F.E., Black, L., Watling, L.E., Mook, W., Mayer,L.M., 1981. A temporal and spatial study of mud flaterosion and deposition. J. Sed. Petrol. 51, 729–736.

Benner, C.S., Krutson, P., Brochu, R., Hurme, A., 1982.Vegetative erosion control in an oligohaline environment.Currituck Sound, North Carolina. Wetlands 2, 105–117.

Bettinetti, A., Pypaert, P., Sweerts, J.-P., 1996. Application ofan integrated management approach to the restorationproject of the lagoon of Venice. J. Environ. Manage. 46,207–227.

Boumans, R.M., Day, J.W., Jr, 1993. High precision measure-ments of sediment elevation in shallow coastal areas usinga sedimentation-erosion table. Estuaries 16, 375–380.

Boumans, R.M.J., Day, J.W., Kemp, G.P., Kilgen, K., 1997.The effect of intertidal sediment fences on wetland surfaceelevation, wave energy and vegetation establishment in twoLouisiana coastal marshes. Ecol. Eng. 9, 37–50.

Bouwersma, P. Bossinade J.H., Dijkema K.S., Meegen,J.W.T.M.V., Reenders, R., 1986. De ontwikkeling van dehoogte en van de omvang van de kwelders in de landaan-winnigswerken in Friesland en Groningen (The progressionin height and area of the saltwater marshes associated withthe reclamation projects in Friesland en Groningen).R.I.N., rapport 86/3, The Netherlands.

Cahoon, D., Reed, D., Day, J., 1995. Estimating shallowsubsidence in microtidal salt marshes of the southeasternUnited States: Kaye and Barghoorn revisited. Mar. Geol.128, 1–9.

Caniglia, G., Contin, G., Fusco, M., Anoe, N., Zanaboni, A.,1997. Confronto su base vegetazionale tra due barene dellalaguna di Venezia. Fitosociologia 34, 111–119.

Childers, D., Sklar, F., Drake, B., Jordan, T., 1993. Seasonalmeasurements of sediment elevation in three Mid-Atlanticestuaries. J. Coastal Res. 9, 986–1003.

Coalition to Restore Coastal Louisiana, 1989. Coastal Louisi-ana, here today and gone tomorrow? Baton Rouge, LA,USA. Unpublished Report.

Day, J.W., Jr, Martin, J., Cardoch, L., Templet, P., 1997.System functioning as a basis of sustainable managementof deltaic ecosystems. Coastal Manage. 25, 115–153.

Day, J.W., Jr, Rismondo, A., Scarton, F., Are, D., Cecconi,G., 1998a. Relative sea level rise and Venice lagoon wet-lands. J. Coastal Conserv. 4, 27–34.

Day, J.W., Jr, Scarton, F., Rismondo, A., Are, D., 1998b.Rapid deterioration of a salt marsh in Venice Lagoon,Italy. J. Coastal Res. 14, 583–590.

Day, J.W., Jr., Rybczyk J., Scarton, F., Rismondo, A., Cec-coni, G., 1999. Accretionary dynamics, sea level rise andthe survival of wetlands in venice lagoon: a field andmodeling approach. Est. Coast Shelf Sci. 49, 607–628.

Dijkema, K.S., Bergs, J.V.D., Ossinade, J., Kroeze, T.A.G.,1990. Experiment natuurbeheer rijkslandaanwinningen: sa-menvatting van de onderzoeks resultaten en jaarverslagover 1989 en 1990. Rijkswaterstaat Directie Groningen,RIN Estuariene Ecologie. Jaarverslag. Nota GRAN 1990-2006 Dec. 1990.

Favero, V., 1992. Evoluzione morfologica e trasformazioniambientali dalla conterminazione lagunare al nostro sec-olo. In: Conterminazione lagunare: storia, ingegneria, po-litica e diritto nella laguna di Venezia. Proceedings of theConference ‘Convegno di studio nel bicentenario dellaconterminazione lagunare’. Venice, 14–16 March, 1991.

Foote, A.L., Kadlec, J.A., 1988. Effects of wave energy onplant establishment in shallow lacustrine wetlands. J.Freshwater Ecol. 4, 523–532.

Gatto, P., Carbognin, L., 1981. The lagoon of Venice: naturalenvironmental trend and man-induced modification. Hy-drol. Sci. Bull. 26, 379–391.

Glopper, R.J.D., 1981. De shnelheid van de opslibbing en vande terugschrijdende erosie op de kwelders langs de noord-kust van Friesland en Groningen. R.I.J.P. Ministerie vanVerkeer en Waterstaat.

Ibanez, C., Day, J.W., Jr, Pont, D., 1999. Primary productionand decomposition of wetlands of the Rhone Delta,France: interactive impacts of human modifications andrelative sea level rise. J. Coastal Res. 15, 717–731.

Kamps, L.F., 1962. Mud distribution and land reclamation inthe eastern Wadden shallows. Rijkswaterstaat, Communi-cations Baflo, 4.

Knutson, P.L., Ford, J.C., Inskeep, M.R., Oyler, J., 1981.National survey of planted salt marshes (vegetative stabi-lization and wave stress). Wetlands 1, 129–157.

Knutson, P., Allen, H., Webb, J.W., 1990. Guidelines forvegetation erosion control on wave-impacted coastaldredge material sites (Final rept). Army Engineer Water-ways Exp. Station. Vicksburg, MS. Environmental Lab.Dept., cWES/TR/D-90-13.

Lieberman, N., Matheja, A., Zimmermann, C., 1997. Forelandstabilisation under waves in shallow tidal waters. 2ndIndian National Conference on Harbour and Ocean Engi-neering (Inchoe-97), Thiruvananthapuram, December 7–10, 1997, 1236–1245.

McKee, K., Patrick, W., 1988. The relationship of smoothcordgrass (Spartina alterniflora) to tidal datums: a review.Estuaries 11, 143–151.

Mendelsshon, I., McKee, K., 1988. Spartina alterniflora die-back in Louisiana: time-course investigation of soil water-logging effects. J. Ecol. 76, 509–521.

Nagai, T., Yamamoto, T., Figueroa, L., 1984. A laboratoryexperimentation on the interactions between water wavesand soft clays beds. Coastal Eng. Jpn. 27, 279–291.

F. Scarton et al. / Ecological Engineering 16 (2000) 223–233 233

Pavoni, B., Donazzolo, R., Marcomini, A., Degobbis, D.,Orio, A.A., 1987. Historical development of the Venicelagoon contamination as recorded in radiodated sedimentcores. Mar. Pollut. Bull. 18, 18–24.

Pignatti, S., 1966. La vegetazione alofila della laguna veneta.Mem. Ist. Ven. Sc. Lett. Arti, Classe di Scienze MM. FF.e NN., Vol. 33.

Scarton, F., Rismondo, A., Day, J., 1998. Above- and below-ground production of Arthrocnemum fruticosum on aVenice lagoon salt marsh. Boll. Museo Civ. St. Nat.Venezia 48, 237–245.

Scarton, F., Rismondo, A., Nascimbeni, P., 2000. Primi datisu biomassa e produzione di Spartina maritima, Limoniumserotinum e Juncus maritimus in laguna di Venezia. LavoriSocieta Veneziana di Scienze Naturali. 25, 16–21

Shybayama, T., Takikawa, H., Horikawa, K., 1986. Mudtransport due to waves. Coastal Eng. Jpn. 29, 151–161.

Stevenson, J.C., Kearney, M., Pendleton, E., 1985. Sedimenta-tion and erosion in a Chesapeake Bay brackish marshsystem. Mar. Geol. 67, 213–235.

Van Eerdt, M.M., 1985. The influence of vegetation on ero-sion and accretion in salt marshes of the Oosterschelde,The Netherlands. Vegetatio 62, 367–373.

Ward, L.G., Kemp, W.M., Boynton, W.R., 1984. The influ-ence of waves and seagrass communities on suspendedparticulate in an estuarine embayment. Mar. Geol. 59,85–103.

Wright, C., Lindsey, P., Cooper, J., 1997. The effect of sedi-mentary processes on the ecology of the mangrove-fringedKosi estuary/lake system, South Africa. Mangroves SaltMarshes 1, 79–94.

Zar, J.H., 1996. Biostatistical Analysis, Third edition. PrenticeHall, Englewood Cliffs, NJ, p. 662.

.