Marine Pollution Bulletin 61 (2010) 189–197

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier .com/locate /marpolbul

Hypertrophic lagoon management by sediment disturbance

Mauro Lenzi a,*, Francesca Birardi a, Roberto Calzolai b, Maria Grazia Finoia c, Francesco Marcone d,Stefano Nocciolini e, Rugiada Roffilli a, Sergio Sgroi f, Duccio Solari a

a Lagoon Ecology and Aquaculture Laboratory, LEALab – OPL srl, v. G. Leopardi 9, 58015 Orbetello, Italyb Tuscany Government, RT, v. di Novoli, Firenze, Italyc University of Siena, Pian dei Mantellini 44, Siena, Italyd World Wide Fund Italy onlus, via Po 25, Roma, Italye Environmental Protection Agency of Tuscany, ARPAT, via Fiume, Grosseto, Italyf Italian Ministry of Environment and Protection of the Territory and Sea, MATTM, v. C. Colombo 44, Roma, Italy

a r t i c l e i n f o

Keywords:Eutrophication managementLagoonOrganic matter pollutionSediment resuspensionSediment disturbanceSediment quality

0025-326X/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.marpolbul.2010.02.012

* Corresponding author. Fax: +39 0564 867572.E-mail address: lealab2@hotmail.com (M. Lenzi).

a b s t r a c t

Experimental control of eutrophication in a small coastal lagoon was attempted by means of sedimentdisturbance. A specially designed boat was used to resuspend the top 3 cm of sediment by a jets ofair–water directed towards the bottom. This disturbance was carried out for 3 months in each of twoareas with a surface area of 24 and 20 hectares respectively. In a total of 80 stations in these two areasand in two undisturbed areas of 16 and 20 ha, organic matter, porosity, density and redox potential weremonitored in sediment bimonthly and free sulphides were monitored in water close to the bottom.Before, during and after disturbance, the impact of daily sediment resuspension on the water columnwas monitored monthly, as ammonium nitrogen (N–NH4), soluble reactive phosphorus (SRP), dissolvedoxygen (DO) and pH. In the whole lagoon, sediment texture was determined at the start and at the end ofthe experiment in 120 stations, and seaweed (mainly Chaetomorpha linum and Lophosiphonia subadunca)and seagrass (Ruppia cirrhosa) biomasses were estimated in 42 stations every month. The results showeda stable organic matter content in disturbed areas and an increase in undisturbed areas, as well as anincrease in seaweed in areas distant from disturbed areas. No significant effect of sediment resuspensionon water column N–NH4, SRP, DO or pH was found.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Many coastal lagoons are subject to eutrophication and developlarge macroalgal biomasses (Morand and Briand, 1996), the break-down of which is a major source of particulate organic matter andat the same time constitutes a large proportion of net nutrient ex-port from these environments (Flindt et al., 2007). The conse-quences of eutrophication and the geochemistry of lagoon aretherefore often closely linked to the production and decompositionof settled organic matter. Bacterial sulphate reduction may trans-form more than 50% of organic matter in coastal lagoons (Jørgensen,1983) so that the increase in autochthonous organic detritus leadsto reduction of dissolved oxygen and production of hydrogensulphide. Sediment redox potential and pH decline under such con-ditions, leading to release of orthophosphates (Fenchel and Riedl,1970) and ammonia (Marty et al., 1990), which in turn promotenew algal blooms. Biogeochemical processes cause release of nutri-ents in surface sediments under low oxygen conditions, prolonging

ll rights reserved.

eutrophication even after cessation of direct man-made sources(Cartei et al., 1998). Dystrophic crises of lagoons may thereforeessentially be due to accumulation of organic matter in sediments,irrespective of the human and endogenous causes of eutrophicationand macroalgal bloom.

Resuspension of sediment accelerates mineralisation of organicmatter (Wainright and Hopkinson, 1997; Stemborg Larsson, 2005)and frequent resuspension can lead to organic matter reduction(Stahlberg et al. 2006). The effect of resuspension on nutrient re-lease is a controversial subject with conflicting aspects, probablyrelated to differences between field and laboratory situations(Tengberg et al., 2003). However, resuspension of sediment maylead to adsorption of phosphorus. Inorganic matter is also oxidisedin sediment dispersed in the water column and ferric oxides areformed if reduced iron is present. Ferric oxides, together with car-bonates and resuspended clays, may adsorb orthophosphates, sub-tracting them from the water column (De Jonge and Villerius,1989; Koski-Vähälä, 2001). The result may be a state of phosphoruslimitation for phytoplankton and the macrophytobenthos (Lenziet al., 2003, 2005). Moreover, the anoxia and bacteria associatedwith the sulphur cycle could mobilise mineral contaminants, such

Fig. 1. Lagoon Burano. Traditional division of the lagoon into western (W), central (C) and eastern (E) parts. WDA, west disturbed area; WUA, west undisturbed area; CDA,central disturbed area; CUA, central undisturbed area; EUA, east undisturbed area.

190 M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197

as mercury in cinnabar producing methylmercury (CH3Hg+) (Woodand Wang, 1983; Kim et al., 2006). Thus, on one hand, resuspen-sion may favour the diffusion of contaminants in the water col-umn; on the other, it may interrupt the activities of sulphurcycle bacteria (Kim et al., 2006). These mechanisms are not yetwell known and call for further research.

Management operations to improve the oxidation status of sur-face sediment are essential to combat eutrophication of lagoonsand coastal wetlands and to control dystrophic conditions (Del-l’Anno et al., 2002; Lenzi et al., 2003). In Mediterranean lagoons,especially those along the Tyrrhenian coast, water exchange andoxidation of surface sediment is made difficult by the low tiderange. Sediment oxidation normally only occurs up to the depthof a few millimetres, unless strong wind weather disturbs andresuspends sediment. Such events are normally too infrequent tohave any significant effect on yearly accumulation of organic mat-ter in sediment. The effects of occasional resuspension may vary inrelation to sediment conditions, accumulation of organic matter,the quality of the latter and water column conditions. The effectsof disturbance repeated with a relatively high frequency may bequite different. In laboratory studies, Stahlberg et al. (2006)showed that frequent resuspension of sediment increased the min-eralisation rate with respect to undisturbed sediment by a factorbetween 2 and 5. In a study conducted in a limited area of a shal-low lagoon, Lenzi et al. (2005) found that repeated passages ofboats that resuspended soft surface sediment increased the oxida-tive status of sediments (increase in Eh), reducing the organic con-tent, without any significant increase in nutrients or oxygenconsumption in the water column. The idea of using resuspensionof sediment to reduce eutrophication of lagoons was conceived as amanagement strategy in the first half of the nineties, after observa-tion of the effect of harvesting and disposal of algal masses in theeutrophic lagoon of Orbetello (Italian west coast). Although the al-gae harvested were only 3–6% of the maximum spring standingcrop, a sharp change in vegetation quality and quantity was ob-tained after only two years (Lenzi et al., 2003). Resuspension ofsediment was an evident effect of the macroalgal harvesting oper-

ations carried out by large harvesting boats in water 1–1.5 m deep.A series of studies was conducted to assess the effects of this phe-nomenon (Lenzi et al., 2004, 2005). They showed that nutrientswere not released into the water column as a result of disturbance,and sediment oxidative status improved. The results were suffi-ciently convincing to warrant a full-scale study to evaluatewhether disturbance of sediment may be a tool to prevent theeutrophication effects in a lagoon ecosystem.

A small eutrophic coastal lagoon subject to periodic dystrophiccrises, which had never undergone sediment operations, was cho-sen for a full-scale study in the framework of a 3-year project of theMinistry for the Environment and Territory (MATTM). The workinghypothesis was that sufficiently prolonged disturbance of surfacesediment may improve environmental quality through oxidationof sediment, reduction of its organic load, and selective effects onsubmerged vegetation, favouring seagrass (that obtains nutrientsdirectly from sediment through its roots) rather than seaweed(caused by P limitation due to oxidation of sediment). The aimswere to determine the effects of large-scale disturbance in the fieldon parameters of environmental quality and dystrophic risk, and totest a boat designed to disturb the surface layer of sediment.

2. Materials and methods

2.1. Description of study area

The lagoon of Burano, situated in the southernmost part of Tus-cany (Italy west coast), is separated from the sea by a sandy spitthat has long been buried in soil. The lagoon has a water surfacearea of 120 ha surrounded by about 30 ha of marshes with Frag-mites. It is elongated in form, extending parallel to the coast in aNW–SE direction and is traditionally distinguished as western,central and eastern parts (Fig. 1). Mean depth is about 0.8 m. It re-ceives water pumped from agricultural areas through a network ofcanals that discharge into the lagoon through two main affluent ca-nals, an eastern pump ditch and Melone ditch to the west (Fig. 1).

Fig. 2. Boat designed to disturb and resuspend the top layer of sediment.

M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197 191

The lagoon receives rainwater from a relatively large agriculturalcatchment (about 13.7 km2). To prevent flooding of the surround-ing land, an underground outlet to the sea is periodically opened.Salinity ranges from 5 to 27 psu, in an annual cycle due to rain,evaporation and subdunal infiltration of seawater. The lagoon isa relict of a greater lagoon and has not been subject to humanintervention since the early 1900s when the Tyrrhenian marsheswere drained. Because of its small size and the fact that it hasnot been interfered with for a long time, it was considered idealfor the present experiment.

2.2. Boat

A specially designed boat was used to disturb and resuspendthe top few centimetres of sediment (Fig. 2). The boat measured4.70 m � 1.60 m (Carolina Skift J 16, 417 kg) and was equiped withan outboard motor (Johnson 15 CV, 48 kg). A generator (GreenPower GP 6000 ST/DE, 6 kW, 250 kg) was installed to power two0.75 kW motors having shafts terminating in propellers, usuallyused to oxygenate aquaculture ponds. The motors were mountedon steel frames on either side of the hull so that the shafts couldbe inclined at different angles. When the motors were in actionthey produced a jet of water and air directed towards the bottom,and could be inclined by varying the angle of the motor shaft inrelation to the speed of the vessel and depth.

2.3. Resuspension and monitoring

About 120 stations (one per hectare) were identified by GPScoordinates. The sediment parameters considered were redox po-tential (Eh, mV) which was measured in the field with a two-elec-trode redox HYDRONAUT (a platinum measuring electrode and acalomel reference electrode), density (Dsed), porosity (g), organicmatter percentage content (OM) and grain size (GS). In 24 of the120 stations we collected water samples to determine free sulp-hides (FS = H2S + HS� + S2�).

The first sampling was carried out in August 2006 to define thestarting condition of sediment in all 120 stations. On the basis ofthese first results and logistic requirements, we chose the areasin which to resuspend and not to resuspend sediment. Two areaswere situated in the western part, west disturbed area (WDA)and west undisturbed area (WUA), 24 and 16 hectares, respec-tively, and two in central part, central disturbed area (CDA) andcentral undisturbed area (CUA), 20 and 20 hectares (Fig. 1). Distur-bance was created by trolling the boat daily over WDA and CDA,with oxygenators in operation; this was done during alternatemonths in the two areas, from October to April. The maximumspeed that the vessel could troll and still maintain the efficacy ofthe air–water jet on the bottom, was 1.9 m s�1. In a month of activ-ity (5 h per day for 24 working days) the area of WDA was there-fore completely disturbed 3.4 times and that of CDA 4.1 times.To aid orientation of the operator, the areas were divided into unitsof 3–4 ha by means of markers.

In November 2006 and January, March and May 2007 monitor-ing was limited to the areas in which resuspension operations wereconducted and to neighbouring undisturbed areas, making a totalof 80 stations (WDA = 24; WUA = 16; CDA = 20; CUA = 20). GSand Eh measurement were repeated in all 120 stations in May2007, at the end of the cycle of experimentation. We also consid-ered a further undisturbed area for GS and Eh, very distant fromthe resuspension operations: east undisturbed area (EUA; 31 sta-tions; Fig. 1).

To determine g, Dsed and OM, sediment was sampled in theupper 3 cm using a horizontal sampler with 60 ml syringe. Sampleswere frozen until analysis. Dsed and g were determined from a1 cm3 subsample (two replicates) of each sample, using the follow-

ing equations: Dsed = Msp � Vws�1; g = Vp � Vws�1, where Msp wasmass of solids and pore-water, Vws wet sediment volume, and Vppore-water volume, determined as loss of weight after drying at90 �C for 24 h. OM percentage was evaluated as loss on ignitionat 400 �C for 3 h after drying samples at 90 �C for 24 h. For GSdetermination (Shepard, 1954), sediment samples were collectedwith a vertical corer (three replicates) taking the top 6 cm of sed-iment. About 60 ml water (three replicates) was obtained 5 cmfrom the bottom with a syringe and silicon tube mounted on apole, in four stations in each of WDA, WUA, CDA and CUA. Avoidingcontact with air, part of the water sample was transferred into agraduated tube containing 2 ml of 20% zinc acetate solution up toa volume of 20 ml. Free sulphides were then determined by spec-trophotometry (Cline, 1969).

Monitoring of the effect of resuspension operations on thewater column was conducted one day per month in the monthsof disturbance operations. Water samples and field measurementswere obtained in one of the disturbed areas (DA), before the boatwent out (7 am, B), during operations (11 am, D) and 2 h after ter-mination of disturbance (5 pm, A). The same monitoring was car-ried out in an undisturbed area (UA) far from the boat’sactivities. The following parameters were measured in five stationsalong a 100 m transect: temperature (t �C) and pH with a DeltaOHM HD8705 digital electronic pH meter; dissolved oxygen (DO,saturation percentage), with an OxyGuard Handy Mk III oxymeter;salinity (s, psu) by Atago S/Mill refractometer. In the end stationsof the transect, three water samples were collected at a depth of20 cm for determination of ammonia nitrogen (N–NH4) and solu-ble reactive phosphorus (SRP). Samples were filtered in the field(0.45 lm mesh), refrigerated and frozen at �18 �C shortly after-wards. N–NH4 was determined by Bran Luebbe AA3 autoanalyzer.SRP was determined by spectrophotometry (Strichland and Par-sons, 1972).

Finally, starting in June 2006, determinations of submergedplant biomass were conducted once a month in 42 stations along14 transects throughout the lagoon. In each station we collectedplant mass in a 50 � 50 cm square (three replicates, 126 quantita-tive samples). The harvested material was divided according tospecies, and each fraction was weighed wet and after desiccationat 90 �C for 24 h. Data processing was carried out separately foreach area (WDA, WUA, CDA, CUA, EUA), dividing the plant biomasssampling stations according to the area in which they weresituated.

2.4. Statistical analysis

Analysis of physical parameters considered: (1) principal com-ponents analysis (PCA) to find correlations between measured vari-ables in time and space; (2) discriminant analysis on PCA factors in

% OM

6

7

8

9

10

11

12

13

14

WDA WUA CDA CUA

Aug 06

Nov

Jan 07

March

May

Eh (mV)

-350

-300

-250

-200

-150

-100

-50

0WDA WUA CDA CUA EUA

Aug 06

Nov

Jan 07

March

May

a b

Fig. 3. Means and standard deviations of estimated percentage organic matter (OM) and redox potential (Eh) in surveys conducted in August and November 2006 andJanuary, March and May 2007, in disturbed (WDA, CDA) and undisturbed (WUA, CUA, EUA) areas.

Table 1Results of paired t-test for the organic matter (OM) values of the sampling areas(WDA, WUA, CDA, CUA) between couples of months. Only significant results arereported.

August–May WDA df = 12, t = 0.62, P > 0.05WUA df = 12, t = �3.38, P < 0.01CDA df = 19, t = 0.67, P > 0.05CUA df = 18, t = �2.37, P < 0.05

August–March WUA df = 14, t = �5.00, P < 0.001August–January CDA df = 19, t = 2.10, P < 0.05

192 M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197

order to: (a) work with linearly independent variables; (b) verifythe existence of any discrimination of data between disturbedand undisturbed areas caused by disturbance activities; (c) identifyvariables that contributed to any discrimination. The validity of thelatter analysis was verified by the Montecarlo test. To analyse pat-terns of sediment and water variables in disturbed and undis-turbed areas, the data was processed with Student’s paired t-testand one-way ANOVA after checking for normal distribution of vari-ables by the Kolmogorov–Smirnov test. The paired t-test was per-formed for each sampling area to detect significant differences invalues of variables between August 2006 (starting conditions)and subsequent sampling months (January, March and May ‘07)during which disturbance was carried out. One-way ANOVA wasperformed for each sampling month to detect significant differ-ences between sediment variables in disturbed and undisturbedareas, and to detect significant differences in the water columnphysical and chemical variables between times of day B, D andA; the post hoc Tukey test was used to isolate any significant differ-ence detected by ANOVA. When the distribution of a variable wasnot normal, non parametric tests were used (Wilcoxon and ANOVAKruskal Wallis tests). The Pearson correlation coefficient was usedto determine correlations between pairs of variables. The analysiswas carried out using SPSS 13.0 and R 2.2.0 software, packageADE4.

3. Results and discussion

3.1. Monitoring sediment parameters and free sulphides near thebottom

Organic matter percentages are reported in Fig. 3a. Student’s t-test showed that mean OM did not vary significantly in WDA be-tween August 2006 and May 2007 (Table 1). In WUA, OM increasedsignificantly by 24.9% between August 2006 and March 2007 andby 21.1% between August and May. In CDA, OM decreased by10.0% between August 2006 and January 2007, but did not showany significant variation between August and May (Table 1). InCUA, OM showed an increasing trend with a significant incrementof 11.0% between August 2006 and May 2007 (Table 1). One-wayANOVA showed significant differences between mean values of

OM in the four areas in August, January, March and May(P < 0.05; Table 2). The post hoc Tukey test showed that OM wassignificantly lower in WDA than in CUA (P < 0.05) in August andJanuary, and than in all other sampling areas (P < 0.001) in Marchand May. In WDA and CDA, disturbance probably counteractedan increase in organic detritus from autumn breakdown of macro-algae, whereas this increase was evident in the two undisturbedareas.

The degree of OM mineralisation depends on its quality (Arnostand Holmer, 2003), but the present research did not consider OMquality. In the case of this small lagoon, often dominated by sea-grass, sedimentary OM could contain a large proportion of materialrefractory to bacterial degradation, and this component may not bedistributed homogeneously throughout the basin. Moreover, thedegree of mineralisation also depends on the frequency of resus-pension of detritus, increasing when the frequency of disturbanceenables a stable bacterial community to become established(Stahlberg et al. 2006).

Porosity did not show significant variations between Augustand May in the four areas (Wilcoxon test, P > 0.05). One-way ANO-VA of monthly sampling data did not show significant differencesbetween the four areas (August, November, January, March, MayP > 0.05). The effects of disturbance were insufficient to signifi-cantly lower porosity values and to differentiate treated from un-treated areas.

Bulk density showed a significant increase in WDA (4.36%) andCDA (1.74%), and a substantial stability in WUA and CUA, betweenthe end-months of the experiment (Table 3). The effects of distur-bance on this variable were evident: both disturbed areas showed

Table 2Monthly means and standard deviation of organic matter (OM,%), bulk density (Dsed, g cm�3) and grain size fraction < 63 lm (GS,%), obtained for sampling areas (WDA, WUA,CDA, CUA), and significant one-way ANOVA results.

WDA WUA CDA CUA EUA One-way ANOVA

OM August 06 9.52 ± 2.29 9.40 ± 1.66 11.09 ± 1.78 10.53 ± 2.60 F3,75 = 3.22, P < 0.05January 07 8.92 ± 1.40 9.72 ± 1.04 9.98 ± 1.53 10.89 ± 1.97 F3,76 = 6.26, P < 0.05March 07 9.24 ± 1.62 11.74 ± 1.18 11.25 ± 1.15 10.99 ± 1.46 F3,76 = 12.92, P < 0.05May 07 9.10 ± 0.61 11.39 ± 1.41 10.76 ± 1.28 11.69 ± 1.57 F3,64 = 12.06, P < 0.05

Dsed January 07 1.15 ± 0.04 1.16 ± 0.04 1.13 ± 0.05 1.19 ± 0.11 F3,66 = 2.76, P < 0.05GS August 06 87.8 ± 4.2 84.3 ± 7.7 80.2 ± 10.2 78.5 ± 14.0 73.2 ± 9.5 F4,107 = 8.17, P < 0.001

May 07 84.5 ± 9.4 86.1 ± 3.3 78.1 ± 9.5 77.1 ± 7.3 71.9 ± 9.5 F4,106 = 16.84, P < 0.001

Table 3Results of Wilcoxon test for bulk density (Dsed), redox (Eh), grain size (GS), freesulphide (FS) values of the sampling areas (WDA, WUA, CDA, EUA) between couples ofmonths. Only significant results are reported.

Dsed August–May WDA Z = �2.39, P < 0.05

CDA Z = �2.22, P < 0.05Eh March–May WDA Z = �2.60, P < 0.01

August–March WDA Z = �2.94, P < 0.01August–May EUA Z = �3.23, P < 0.001

GS August–May WDA Z = �2.03, P < 0.05FS August–May WDA Z = �2.00, P < 0.05

WUA Z = �2.20, P < 0.05CDA Z = �2.20, P < 0.05

M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197 193

an increase in this variable, while values remained stable in undis-turbed areas. The increase may be attributed to a decrease in OMand replacement of pore-water with heavier mineral fractions,such as clay, which was abundant in sediments, rendering the sub-strate more compact. One-way ANOVA of monthly sampling dataonly showed a significant difference among the areas in January(Table 2) (post hoc Tukey test: CDA vs. CUA, P < 0.05).

In August, the areas with lowest Eh values were CDA, CUA andEUA (Fig. 3b). In WDA, Eh increased significantly (P < 0.01) in win-ter (by as much as 47% between August and March) but decreasedwhen warmth returned (between March and May; P < 0.01), mak-ing the difference between the two end points non-significant (Ta-ble 3). A similar situation was found for WUA, CDA and CUA(Fig. 3b), with similar significance, which suggests that seasonal ef-fects, with a rise in redox potential between August and March fol-lowed by a decline, cancelled out the effects of disturbance in thetwo areas. EUA was the only area in which Eh decreased signifi-cantly between August and May (Table 3).

PCA of Dsed, g, Eh and OM did not show any true pattern inWDA, WUA, CDA, CUA, but in the plane of months and stations itshowed the dependence of August from Eh. Discriminant analysisof PCA (Fig. 4) showed: (1) separation of WDA from WUA, CDAand CUA due largely to OM; (2) distancing of Dsed and Eh valuesof CDA from those of CUA. The first two discriminant functions ex-plained 74.5% and 17.8% of overall variability. The Montecarlo test

Table 4Means and standard deviations of total data of dissolved oxygen (DO, %), pH, nitrogen ammoundisturbed (UA) areas, before (B), during (D) and after (A) sediment resuspension operationtime of day (before, B, during, D, and after, A, disturbance activity).

B D

DO DA 86.1 ± 20.6 96.9 ±UA 92.6 ± 25.8 102.8

pH DA 7.79 ± 0.44 8.01 ±UA 7.92 ± 0.53 8.02 ±

N–NHþ4 DA 23.04 ± 19.55 18.26UA 23.52 ± 17.61 26.33

SRP DA 0.88 ± 1.47 1.29 ±UA 1.39 ± 1.85 1.16 ±

indicated that the results were significant (P = 0.001). These resultsprovide further evidence that disturbed areas (WDA more thanCDA) underwent a change in sedimentary parameters with respectto undisturbed areas.

Regarding GS (Fig. 5), the fraction <63 lm decreased signifi-cantly from west to east in August 2006 and May 2007 (Table 2).The outlet of the Melone ditch was evidently the main source ofclay (Fig. 1). The percentage of the fraction <63 lm decreased sig-nificantly in WDA between August (87.8 ± 4.2%) and May(84.5 ± 9.4%) (P < 0.05; Table 3), whereas in the other areas therewere no significant variations. The reduction in values of this frac-tion in WDA could be due to dispersal of resuspended clay to moredistant areas by disturbance. Since WDA received alluvial clay to agreater extent than other areas, transport from this area could havebeen greater than in the other disturbed area and thereforedetectable.

As expected, free sulphides (FS, Fig. 6) had higher levels in thehottest month, indicating WDA, WUA and especially CDA as criticalareas; CUA had lower levels. Between August 2006 and May 2007,FS significantly decreased in WDA, WUA and CDA (P < 0.05),whereas no significant variation was recorded in CUA (Table 3).Although the relatively small number of readings precluded validstatistical analysis, the two undisturbed areas WUA and CUAshowed rises in FS in March and January, respectively. These anom-alies could be attributed to resumption of sulphate reduction inavailable organic substrate due to a particularly mild winter. Thefact that this did not occur in both disturbed areas, where therewas a constant reduction in FS between August and May, may bedue to interruption of sulphate reduction processes by disturbanceoperations.

3.2. Evaluation of impact of disturbance on the water column

The impact of disturbance on DO was assessed by ANOVA for re-peated measures. Specifically, within-subject analysis showed: (a)an effect due to DO repeated measures, irrespective of the effect ofthe other experimental factors (F1,56 = 704.5, P < 0.001); (b) nointeractive effect between area and DO (F1,56 = 0.005, P > 0.05);(c) a joint effect of time and DO (DO-month; F6,56 = 58, P < 0.001);

nium (N–NHþ4 , lM) and soluble reactive phosphorus (SRP, lM), in disturbed (DA) ands, and results of the ANOVA (one way or non parametric) for the means comparison of

A One-way ANOVA

13.2 113.3 ± 19.8 F2,102 = 19.36, P < 0.001± 28.6 120.4 ± 34.7 F2,96 = 8.46, P < 0.0010.48 8.01 ± 0.51 F2,39 = 0.25, P > 0.050.56 8.07 ± 0.61 F2,39 = 0.77, P > 0.05± 11.61 22.47 ± 14.66 F2,34 = 0.73, P > 0.05± 21.47 23.79 ± 31.04 F2,30 = 0.14, P > 0.05

ANOVA Kruskal Wallis1.47 0.85 ± 0.74 v2 = 0.375 P > 0.050.91 0.84 ± 0.83 v2 = 0.006 P > 0.05

Fig. 4. Discriminant analysis of principal components showing: (a) canonic weights; (b) correlation between discriminant variables and functions; (c) histogram ofeigenvalues; (d) correlation between PCA factors and discriminant functions; (e) group centroids; (f) discriminant scores grouped according to area. Ellipsoids represent 95%confidence intervals calculated for each group.

GS <63 µm fraction (%)

60

65

70

75

80

85

90

95

100

WDA WUA CDA CUA EUA

Aug 06M ay 07

Fig. 5. Means and standard deviations of grain size fraction < 63 lm for disturbed(WDA, CDA) and undisturbed areas (WUA, CUA, EUA), at the start (August 2006)and end (May 2007) of the experimental cycle.

Free sulphide (µM)

0

24

68

10

1214

16

WDA WUA CDA CUA

Aug 06

Nov

Jan 07

March

May

Fig. 6. Means and standard deviations of free sulphide for disturbed (WDA, CDA)and undisturbed areas (WUA, CUA), in the surveys of August 2006, November,January 2007, March and May.

194 M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197

(d) an effect due to DO repeated values detected in various areas atvarious times (DO-month-area: F6,56 = 556.5, P < 0.001). Therefore,disturbed and undisturbed areas always showed similar patterns.With regard to the effect of time of day (before B, during D andafter A disturbance), one-way ANOVA detected significant differ-ences between means at the three times (P < 0.001; Table 4). Inthe disturbed area (DA), there were always significant increasesamong B, D and A (post hoc Tukey test, P < 0.05), while in the

undisturbed area (UA) a significant increase occurred only betweenB and A (Tukey test P < 0.001). Increases in values from morning toafternoon can be ascribed to photosynthesis, which is typical ofshallow lagoons with poor water turnover. Paradoxically, the dis-turbed area showed a greater rise in dissolved oxygen in the courseof the day than the undisturbed area, establishing that the effect ofdisturbance on this parameter was undetectable. pH did not differbetween B, D and A in UA or DA (P > 0.05; Table 4).

There are contrasting opinions about the effect of resuspensionon water column SRP concentrations: according to de Groot (1981)

seaweed gWW m-2

0

200

400

600

800

1000

1200

1400

1600

1800

Jun06 July Aug Sept Oct Nov Dec Jen07 Feb March Apr May June July

WDA

WUA

CDA

CUA

EUA

Fig. 7. Monthly means and standard deviations of seaweed biomass (g wet weight m�2) in disturbed (WDA, CDA) and undisturbed (WUA, CUA, EUA) areas between June 2006and July 2007.

seaweed standing crops (tonnes WW)

0

500

1000

1500

2000

2500

2004 2005 2006 2007

June

July

Fig. 8. Standing crops of seaweed in Lagoon Burano, in tonnes wet weight,estimated in June and July in the period 2004–2007.

M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197 195

and S�øndergaard et al. (1992), resuspension increases SRP in thewater column, according to others it reduces it (Reddy et al.,1996), and according to others it does not have any significant ef-fect (Peters and Cattaneo, 1984; Koski-Vähälä, 2001). In our study,SRP did not show significant variations between B, D and A in DA orin UA (P > 0.05; Table 4). The data collected so far did not enableany correlation to be found between SRP and pH. This is not in linewith the hypothesis that the two main mechanisms of phosphorusdesorption, low redox potential and high pH, increase release of Pin water when anoxic sediment is mixed in the water column athigh pH (Koski-Vähälä, 2001). Release of P from solids into solutionnaturally depends on many other factors, such as temperature, anincrease in which increases sorption (Koski-Vähälä, 2001), salinity,an increase in which favours desorption (Gardolinski et al., 2004),and suspended solids, at high concentrations of which more P re-mains linked to particles (Deborde et al., 2007). More data is there-fore needed to investigate SRP and specifically the possibility of acorrelation with pH due to the effects of sorption/desorption.

There are also contrasting opinions about the effects of resus-pension on the water column with regard to ammonia. Blackburn(1997) minimised release, considering that it would occur by diffu-sion, whereas Morin and Morse (1999) emphasised release ofammonia by resuspended sediment, largely by desorption. In ourstudy, we did not observe significant variations in ammonia be-tween B, D and A in DA or in UA (P > 0.05; Table 4). Available datasuggests that disturbance did not significantly affect the physicalchemistry of the water column.

3.3. Vegetation

The dominant species were the Chlorophyta Chaetomorpha li-num (Mull.) Kütz., the Rhodophyta Lophosiphonia subadunca(Kütz.) Falkenberg and the seagrass Ruppia cirrhosa (Petagna)

Grande. Minor species included the Chlorophyta Cladophora spp.,Ulva laetevirens Areschoug, Ulva prolifera Müller and Ulva intestinal-is L., the Rhodophyta Ceramium codii (Richards.) Feld. et Mazoyerand the seagrass Potamogeton pectinatus L. In previous studies, R.cirrhosa was found throughout the lagoon with higher densitiesin the central and eastern areas, while C. linum was more abundantin the the west and L. subadunca in the central and eastern areas(Lenzi, WWF technical reports).

In WDA, seaweed biomass (Fig. 7) was high until November in2006, declining rapidly in winter. Biomass did not increase againthe following spring, remaining very low in 2007. A similar patternwas observed in WUA, although with an early peak in June 2006and higher biomass than in WDA in 2007. In CDA, macroalgal bio-mass was slightly higher in 2007 than in the corresponding periodsof 2006. Compared with the previous two areas, biomass in CDAshowed a typical increase between winter and spring, with a max-imum in May. CUA and EUA showed an increase in biomass inOctober 2006, and stable, relatively high values for the rest ofthe study period.

In WDA and WUA, macroalgal development therefore turnedout to be limited with respect to the past. It is not clear whyWUA showed a similar pattern to the disturbed area. However,since WUA was relatively small (16 ha) and situated betweentwo large disturbed areas, it may have received fall-out of oxidisedmatter that may have altered the biogeochemistry of the surfacelayer of sediment. It is known, in fact, that bacterial activity in-creases in sediment receiving fall-out of resuspended organic sed-imentary material (Wainright, 1987; Logan and Kirchman, 1991).

Comparing seaweed standing crops in the whole lagoon for Juneand July of the years 2004 and 2007, we found lower values in2007 (Fig. 8), attributable to the lack of production in the westernarea. Moreover, the usual past distribution with high biomass of C.linum in the west and L. subadunca in the east did not occur: bothtended to develop in the central and eastern areas.

Biomass of seagrass was higher in WDA and WUA in 2007 thanin 2006, whereas in all the other areas (CDA, CUA, EUA) biomassdeclined in 2007 (Fig. 9). Between March and April 2007, we ob-served considerable development of seedlings on bottoms free ofalgae, especially in WDA and much less in WUA and CDA.

The distribution of biomasses of seagrass and seaweed accord-ing to these first results did not fully clarify the effect that distur-bance could have on these populations. Indeed, in WDAdisturbance clearly seemed to favour seagrass and limit seaweed,as hypothesised, even if the same occurred in WUA. On the otherhand, in CDA the expected reduction in seaweed was not observedbut rather a reduction in seagrass. Finally, the undisturbed areas

seagrass gWW m-2

0

500

1000

1500

2000

2500

Jun06 July Aug Sept Oct Nov Dec Jen07 Feb March Apr May June July

WDA

WUA

CDA

CUA

EUA

Fig. 9. Monthly means and standard deviations of seagrass biomass (g wet weight m�2) in disturbed (WDA, CDA) and undisturbed (WUA, CUA, EUA) areas between June 2006and July 2007.

196 M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197

CUA and EUA showed a reduction in the biomass of phanerogamsand an increase in macroalgal biomass. The emerging complex pic-ture and the anomalies of WUA and CDA with respect to thehypothesis call for further research.

4. Conclusions

These results indicate that: (1) resuspension of sediment did nothave a significant direct effect on concentrations of N–NH4, SRP, DOor pH in the water column; (2) disturbance of sediments main-tained constant values of OM in the sediments against the increas-ing trend in OM content observed in sediment of undisturbed areasdue to production of macroalgal detritus; (3) discriminant analysisof PCA on sedimentary parameters showed differentiation betweendisturbed and undisturbed areas; (4) macroalgal vegetation de-creased considerably in one of the disturbed areas and also in theundisturbed area situated between the two disturbed ones,whereas it increased in areas far from the disturbed areas; (5) sea-grass meadows increased greatly in the western disturbed area,where seedlings of R. cirrhosa were observed for the first time inmany years; 6) the increase in seaweed and decrease in seagrassin undisturbed area EUA led to a significant fall in Eh.

These results largely verify our working hypothesis that suffi-ciently prolonged disturbance of surface sediment may improveenvironmental quality through reduction (or limitation of furtherincrease) of its organic load, and promotion of seagrass rather thanseaweed. Indeed, the results from WDA were almost completely inline with the working hypothesis as far as the quality of sedimentsand submerged vegetation were concerned. On the other hand, theresults from CDA showed some contradictions: sediment qualityimproved but not vegetation quality. Further research is thereforeneeded to clarify these partial contradictions and the interpretativedifficulties of data obtained in this large-scale field experimentwith inherently complex dynamics. This study was a first attemptat managing a eutrophic ecosystem by direct intervention on sed-iments, and further work will be needed to tackle the aspects notyet considered. In line with the results of Stahlberg et al. (2006),it will also be necessary to establish the intensity of the minimumdisturbance required to obtain useful effects, which probably var-ies in relation to the type of sediments, height of the water columnand degree of eutrophication. Finally, due to the approach used inthe present study, certain factors, such as presence of contami-nants and risk of their resuspension and effects of disturbance

and sedimentary fall-out on zoobenthic populations, could not beconsidered. These factors need to be accounted in future experi-ments, and should of course be evaluated in relation to the fre-quency of natural phenomena of resuspension and in the light ofrecent studies (Kim et al., 2006; Widdicombe and Austen, 2001).

References

Arnost, C., Holmer, M., 2003. Carbon cycling in a continental margin sediment:contrasts between organic matter characteristics and remineralization ratesand pathways. Estuarine, Coastal and Shelf Science 58, 197–208.

Blackburn, T.H., 1997. Release of nitrogen compound following resuspension ofsediment: model predictions. Journal of Marine Systems 11, 343–352.

Cartei, P., Innamorati, M., Melillo, C., 1998. Trophic Homeostasis Sea modulated andlagoon autochthonous hypertrophy. Biologia Marina Mediterranea 5 (1), 41–46.

Cline, J.D., 1969. Spectrophotometric determination of hydrogen sulphide in naturalwaters. Limnology and Oceanography 14, 454–459.

de Groot, W.T., 1981. Phosphate and wind in a shallow lagoon. Archiv FurHydrobiologie 91, 475–489.

Deborde, J., Anschutz, P., Chaullou, G., Etcheber, H., Cammarieu, M.V., Lecroart, P.,Abril, G., 2007. The dynamics od phosphorus in turbid estuarine systems:example of the Gironde estuary (France). Limnology and Oceanography 52 (2),862–872.

De Jonge, V.N., Villerius, L.A., 1989. Possible role of carbonate dissolution inestuarine phosphate dynamics. Limnology and Oceanography 34, 332–340.

Dell’Anno, A., Mei, M.L., Pusceddu, A., Danovaro, R., 2002. Assessing the trophic stateand eutrophication of coastal marine system: a new approach based on thebiochemical composition of sediment organic matter. Marine Pollution Bulletin44, 611–622.

Fenchel, T.M., Riedl, R.J., 1970. The sulfide system: a new biotic communityunderneath the oxidized layer of marine sand bottoms. Marine Biology 7, 255–268.

Flindt, M.R., Pedersen, C.B., Amos, C.L., Levy, A., Bergamasco, A., Friend, P.L., 2007.Transport, sloughing and settling rates of estuarine macrophytes: mechanismsand ecological implications. Continental Shelf Research 27, 1096–1103.

Gardolinski, P.C.F.C., Worsfold, P.J., McKelvie, I.D., 2004. Seawater induced releaseand transformation of organic and inorganic phosphorus from river sediments.Water Research 38, 688–692.

Jørgensen, B.B., 1983. The microbial sulphur cycle. In: Krumbein, W. (Ed.), MicrobialGeochemistry. Blackwell Scientific Publications, Oxford, pp. 91–124.

Kim, E.H., Mason, R.P., Porter, E.T., Soulen, H.L., 2006. The impact of resuspension onsediment mercury dynamics, and methylmercury production and fate: amesocosm study. Marine Chemistry 102, 300–315.

Koski-Vähälä, J., 2001. Role of Resuspension and Silicate in Internal PhosphorusLoading. Academic Dissertation in Limnology. Department of Limnology andEnvironmental Protection, University of Helsinki, p. 45 <http://www.diva-portal.org/liu/searchresult.xsql>.

Lenzi, M., Finoia, M.G., Persia, E., Comandi, S., Gargiulo, V., Solari, D., Gennaro, P.,Porrello, S., 2005. Biogeochemical effects of disturbance in shallow watersediment by macroalgae harvesting boats. Marine Pollution Bulletin 50 (5),512–519.

Lenzi, M., Palmieri, R., Porrello, S., 2003. Restoration of the eutrophic Orbetellolagoon (Tyrrhenian Sea, Italy): water quality management. Marine PollutionBulletin 46, 1540–1548.

M. Lenzi et al. / Marine Pollution Bulletin 61 (2010) 189–197 197

Lenzi, M., Persia, E., Sangermano, G., Gennaro, P., Mercatali, I., Solari, D., Porello, S.,2004. Ipotesi di gestione di un ambiente lagunare ipertrofico mediante disturbosedimentario: un approccio preliminare. Atti del Convegno SItE, Siena.

Logan, B.E., Kirchman, D.L., 1991. Uptake of dissolved organics by main bacteria as afunction of fluid motion. Marine Biology 111, 175–181.

Marty, D., Esnault, G., Caumette, P., Ranaivoson-Rambeloarisoa, E., Bertrand, J.C.,1990. Denitrification, sulfato-reduction et methanogenese dans les sedimentssuperficiels d’un étang saumatre méditerranéen. Oceanologica Acta 13, 199–210.

Morand, P., Briand, X., 1996. Excessive growth of macroalgae: a symptom ofenvironmental disturbance. Botanica Marina 39, 491–516.

Morin, J., Morse, J.W., 1999. Ammonium release from resuspended sediments in theLaguna Madre estuary. Marine Chemistry 65, 97–110.

Peters, R.H., Cattaneo, A., 1984. The effects of turbulence on phosphorus supply in ashallow bay of Lagoon Memphremagog. Verhandlungen des InternationalenVerein Limnologie 22, 185–189.

Reddy, K.R., Fisher, M.M., Ivanoff, D., 1996. Resuspension and diffusive flux ofnitrogen and phosphorus in a hypereutrophic lagoon. Journal of EnvironmentalQuality 25, 363–371.

Shepard, F.P., 1954. Nomenclature based on sand, silt, clay ratio. Journal ofSedimentary Petrology 25, 151–158.

S�øndergaard, M., Kristensen, P., Jeppesen, E., 1992. Phosphorus release fromresuspended sediment in the shallow and wind-exposed Lagoon Arroese,Denmark. Hydrobiologia 228, 91–99.

Stahlberg, C., Bastviken, D., Svensson, B.H., Rahm, L., 2006. Mineralisation of organicmatter in coastal sediment at different frequency and duration of resuspension.Estuarine, Coastal and Shelf Science 70, 317–325.

Stemborg Larsson, C., 2005. Effect of Resuspension on Mineralisation of OrganicMaterial. Master Thesis, Linköpings universitet, Sweden, p. 17 <http://www.diva-portal.org/liu/searchresult.xsql>.

Strichland, J.D.H., Parsons, T.R., 1972. A practical handbook of seawater analysis.Bulletin of the Fisheries Research Board of Canada 167, 1–312.

Tengberg, A., Almroth, E., Hall, P., 2003. Resuspension and its effects on organiccarbon recycling and nutrient exchange in coastal sediments: in situmeasurements using new experimental technology. Journal of ExperimentalMarine Biology and Ecology 285–286, 119–142.

Wainright, S.C., 1987. Stimulation of heterotrophic microplankton production byresuspended marine sediment. Science 238, 1710–1712.

Wainright, S.C., Hopkinson Jr., C.S., 1997. Effects of sediment resuspension onorganic matter processing in coastal environments: a simulation model. Journalof Marine Systems 11, 353–368.

Widdicombe, S., Austen, M.C., 2001. The interaction between physical disturbanceand organic enrichment: an important element in structuring benthiccommunities. Limnology and Oceanography 46 (7), 1720–1733.

Wood, J.M., Wang, H.K., 1983. Microbial resistance to heavy metals. EnvironmentalScience and Technololy 17, 82–90.