Erosion and accretion in the Ria de Aveiro inlet ( N Portugal )

Harbours are important man made economic structures that are usually constructed in hydrodynamic complex settings. Their life/ service cycle, integrity and maintenance are highly dependent on the knowledge of these settings and the capability to predict future changes and thus their economic impact. In this paper, data regarding bottom and suspended sediments of the Aveiro Harbour inlet were gathered and analysed. Measurements of suspended sediment concentrations (SSC) in the Aveiro inlet were taken during eight tidal cycles in 2007, includSediments were imported from the ocean sink, mostly near the south jetty and the western side of the Mira Channel, as textural, mineralogical, compositional, geochemical and microfaunal data of bottom sediments suggest. Higher stabilization of the Aveiro Harbour structures is achieved on these accretional areas. The high tidal-current velocities contribute to the erosive character which dominates the central area of the entrance channel, which is located between both breakwaters and close to the north jetty; thus some segments of the north jetty could become damaged through erosion. This is evidenced as much bottom sediments characteristics as patterns of SSC. SSC are in general higher during the winter, due to climatic conditions, current velocities, wind stress and greater availability of ISSN (print): 1698-6180. ISSN (online): 1886-7995 www.ucm.es /info/estratig/journal.htm Journal of Iberian Geology 37 (2) 2011: 215-230 doi: 10.5209/rev_JIGE.2011.v37.n2.9 216 < ' = > ' 9? ' !F (..# H . 5$.!#


Introduction
The Ria de Aveiro is the most extensive coastal lagoon of the Western coast of Portugal, roughly located between Espinho city and Cabo Mondego (40º38'N, 8º44'W), enclosed by a large sand barrier (Fig. 1).
The lagoon formation has evolved since the 10 th century, when a sandy spit started to proceed southwards from Espinho until the mouth of the Vouga River, isolating progressively the ancient bay from the sea (Abecasis, 1955).This sand spit grew southwards whereas another sand spit developed from Cabo Mondego northwards.The sand spits temporarily closed the communication between the lagoon and the sea in 1575.Since then, the natural inlet has migrated although it has recurrently suffered from obstructions, as was described by Abecasis (1955).From the middle of the 18 th century, EORFNDJHV LQ WKH ÀRZ EHWZHHQ WKH ODJRRQ DQG WKH VHD EHcame more frequent, with disastrous consequences not only for maritime activities but also for the public health of the local population (Amorim, 2008;Amaral, 1968).
&RQVHTXHQWO\ DQ DUWL¿FLDO LQOHW ZDV RSHQHG LQ which has since remained as the primary opening, serv-LFLQJ WKH $YHLUR +DUERXU DQG LWV DFWLYLWLHV ¿VKLQJ LQdustrial, commercial and recreational activities).The jetties that help maintain the lagoon mouth open have been the subject of several maintenance projects since being built.An extensive list of these interventions can be found in Plecha et al. (2010).Presently, the channel of the inlet is 1.3km long, 350m wide and 25/30m deep and connects the lagoonal body to the Atlantic Ocean (Fig. 1).
At present, the lagoon is 45km long and 10km wide, covering an area corresponding approximately to a minimum of 66km 2 , at low spring tides, and a maximum of 83km 2 , at high spring tides (Dias and Lopes, 2006).The main portion of the Aveiro lagoon is comprised of a complex dendritic pattern of channels, characterised E\ WKH H[LVWHQFH RI VLJQL¿FDQW LQWHUWLGDO ]RQHV QDPHO\ PXG ÀDWV DQG VDOW PDUVKHV ZLWK IRXU PDLQ EUDQFKHV connected to the entrance: Mira, S. Jacinto, Ílhavo and Espinheiro Channels (Dias et al., 2001).The Mira and S. Jacinto Channels are in the continuity of the Ria de Aveiro inlet.In the inner parts of the lagoon there is an extensive tributary area supplying this system with fresh water and sediments, but the only two important rivers ÀRZLQJ WR WKH ODJRRQ DUH WKH 9RXJD 5LYHU DQG WKH $Qtuã River with an annual mean discharge of 50m 3 s í and 5m 3 s í , respectively (Moreira et al., 1993).
Keywords: multiproxy approach, tidal currents, suspended and bottom sediments, Ria de Aveiro, N Portugal

Resumen
Los puertos son importantes estructuras económicas diseñadas por el hombre, que normalmente son construidas en lugares con una compleja hidrodinámica.Su ciclo de vida de servicio, integridad y mantenimiento dependen del conocimiento de dicha hidrodinámica y de la capacidad para predecir los cambios futuros y por tanto su impacto económico.En este trabajo, los datos sobre los sedimentos que están en el fondo y en suspensión de la entrada del puerto de Aveiro fueron recogidos y analizados.

%RWWRP VHGLPHQWV
The study area is located in the Ria de Aveiro (N Portugal) inlet and surrounding areas, i.e. the S. Jacinto and Mira channels (Fig. 1).A set of 113 surface sediment grab (Petit Ponar) samples were collected at the study area during different oceanographic campaigns throughout 2006 and 2007.At each sampling station the water depth was evaluated using the boat sonar.The metric coordinates and depth of the studied sites, as well as the dates on which samples were collected, have been included in Appendix 1.
In each station, two sub-samples of surface sediment ¿UVW §FP ZHUH FROOHFWHG 2QH EDWFK RI VXEVDPSOHV was preserved for textural, mineralogical and geochemical analyses.The other batch of sub-samples, at least 500ml of sediment, was preserved by adding a rose Bengal solution in 90% ethanol with the aim of studying foraminifera assemblages.Rose Bengal was used in order to discriminate between live (red-stained protoplasm) and dead foraminifers (empty tests).
Grain size analysis was performed by the classic sieving method.Sediment grain sizes were evaluated by the Folk and Ward method, using the scale in microns.Mineralogical and geochemical studies were carried out on a grounded split of the sediment fraction <2000μm of each sample.The samples were ground until they UHDFKHG VHGLPHQWDU\ SDUWLFOHV VPDOOHU WKDQ ȝP *HRchemical and mineralogical analyses were performed on the milled sediments.
Tides are semidiurnal and tidal range varies between 0.6m (neap tides) and 3.2m (spring tides), with an average range at the inlet of about 2m (Dias et al., 2000;Sousa and Dias, 2007).The tides are higher on ebb than RQ ÀRRG 'LDV et al., 2000), revealing an ebb dominated system (Picado et al., 2010).The ebb jet extends into the shelf as shown by numerical simulations of the ocean circulation in the vicinity of the inlet (Silva et al., 2009).
Understanding the erosion and accretion trends in the Ria de Aveiro inlet and surrounding areas is an important issue in order to predict the evolution of the lagoon, considering that the inlet channel depth determines the tidal prism, tidal velocities and sea surface elevation along the lagoon (Araújo et al., 2008;Picado et al, 2010).Regarding the adjacent coastal evolution, which KDV VKRZQ D VLJQL¿FDQW UHWUHDW DFFRUGLQJ WR 'LDV et al. (1994) and Oliveira et al. (1982), it's important to evaluate the lagoon sediment exchange with the littoral drift.
The sediment transport in the lagoon is essentially determined by the tidal currents' intensity and phase (Lopes et al. DQG E\ WKH PDULQH DQG ÀXYLDO VHGLment supply (Abrantes et al., 2006), whereas the sediments balance in the inlet and adjacent areas is also determined by the wave regime (Plecha et al., 2010).
This paper presents results obtained under the framework of the EMERA project (Morfodynamic Study of Ria de Aveiro Inlet) aiming: i) to perform a sediment characterization of the Ria de Aveiro inlet; ii) to identify the erosion/ deposition trends in the tidal inlet and surrounding area; iii) to identify areas where marine sediment contributions are important; iv) to analyse the process of sediment supply to the continental shelf.
7KH PLQHUDORJLFDO VWXGLHV ZHUH FDUULHG RXW XVLQJ ;±UD\ diffraction (XRD) techniques, according to the procedures described by Martins et al. (2007).Concentrations of Al, Ba, Ca, Ce, Cr, K, La, Li, Mg, Mn, Na, P, Rb, S, 6U 7K 7L 9 =Q DQG =U ZHUH GHWHUPLQHG E\ ,&3±06 ,Q-GXFWLYHO\ &RXSOHG 3ODVPD±0DVV 6SHFWURPHWHU DQDO\VLV DIWHU D ±DFLG GLJHVWLRQ +&O2 4 ±+12 3 ±+&O±+) DW DQ LQWHUQDWLRQDO DFFUHGLWHG ODERUDWRU\ ± $&0( $QDO\WLFDO /DERUDWRULHV /WG ± ,62 $FFUHGLWHG &R Each sample for foraminifera analysis was carefully ZDVKHG ZLWK IUHVK ZDWHU XVLQJ ±P DQG ±P mesh sieves, to remove mud-size sediments and the excess of rose Bengal solution, and dried in an oven, at ±& XQWLO WKH VDPSOH ZDV WKRURXJKO\ GU\ %HQWKLF foraminifera studies were performed using a light binocular microscope.Dead and living foraminifera densities, i.e. the number of tests/ individuals per gram of the VHGLPHQW IUDFWLRQ ±Pm, were determined in each sample.This sediment fraction was chosen because it includes the greater part of the specimens of the group studied in the Ria de Aveiro.Also, this method avoids VLJQL¿FDQWO\ UHGXFHG IRUDPLQLIHUD GHQVLW\ YDOXHV FDXVHG by coarser and denser terrigenous grains. The study of the dead and living assemblages of benthic IRUDPLQLIHUD ZDV PDGH RQ WKH GULHG UHVLGXH ±ȝP of each sample.The generic assignments followed Loe-EOLFK DQG 7DSSDQ 6SHFLHV LGHQWL¿FDWLRQ ZDV based on references cited by Martins and Gomes (2004), as well as on the Ellis and Messina Catalogue (1940 and subsequent).In order to characterize the composition of the assemblage by percentage of each species, about 300 empty and 300 pink coloured (living specimens) foraminifera tests was considered an ideal number.In samples where the number of specimens was lower than 100, the percentage of individuals per species and per sample in the dead and/ or in the living assemblages was not de-WHUPLQHG 7KH 6KDQQRQ LQGH[ 6KDQQRQ + íȈSL ln(pi), where pi is the proportion of each species) was used to identify the dead assemblage diversity.The use of this index has advantages because it incorporates the species richness and evenness and is applicable to a simple statistical analysis.
The proportion of imported shells from the ocean was determined, taking into account several aspects: the degree of degradation due to transport of foraminifera tests (the number of tests degraded by transport was recorded and duly noted in the initial data table) and the study of the patterns of distribution of living foraminifera in the Ria de Aveiro (according to the work of Martins et al., included LQ WKLV YROXPH ZKLFK SHUPLWWHG WKH LGHQWL¿FDWLRQ RI IDXnas more characteristic of oceanic or lagoonal environments.Previous studies analyzing foraminifera distribu-tion in other Iberian coastal and lagoonal areas were also taken in account (e.g.Alejo et al., 1999;Cearreta, 1988Cearreta, a, b, 1989;;Cearreta et al., 2000Cearreta et al., , 2002;;Diz et al., 2004, *RQ]iOH]±5HJDODGR et al., 1996;Moreno et al., 2005;Muñoz et al., 1996;Pascual, 1984Pascual, , 1992) ) and elsewhere (Murray, 1991).The proportion of empty tests imported from the ocean is hereafter referred to in this work as 'oceanic specimens'.The species list considered in this analysis is detailed in the following section (Results).
Areas affected by erosion or accretion have been identi-¿HG E\ XVLQJ ERWWRP VHGLPHQWV ¶ WH[WXUDO PLQHUDORJLFDO geochemical and microfaunal data (foraminifera density and percentage of 'oceanic specimens').A higher density of empty foraminifera tests indicates areas where sediments are being essentially deposited.The variation of the percentage of 'oceanic specimens' was used as a proxy to trace how important is the contribution of sedimentary particles originated in oceanic areas in the lagoon.The mineralogical and geochemical composition of the surface sediments was also used to help indicate DUHDV ZLWK KLJKHU PDULQH LQÀXHQFH LQ WKH 5LD GH $YHLUR inlet and surrounding areas.
Discrimination of spatial variations between dead and living assemblages, in all samples, was compared through a presence/ absence matrix by applying the Bray Curtis FRHI¿FLHQW DQG XVLQJ WKH SURJUDP 3ULPHU ,Q RUGHU WR discriminate the relationships of textural, mineralogical, geochemical and microfaunal data within samples and between the sampled stations, cluster analysis was applied, using Statistica V.07 software.
Suspended sediment concentrations were determined according to the methodology applied by Abrantes et al.

%RWWRP 6HGLPHQWV
Sediment median grain sizes (D50) were found to vary between 0.041mm and 3.7mm.Most of the surface sediment of the studied area is composed of sand.The coarse VHGLPHQW IUDFWLRQ !ȝP YDULHV IURP DOPRVW WR 99%, in individual samples.'Poorly sorted' coarse sand and gravel are found mostly along the S. Jacinto Channel, in the central area of the Ria de Aveiro inlet and close to the north jetty.In more sheltered areas, such as at the western side of the Mira Channel, near the south jetty and at the head of north jetty, 'moderately well sorted' and µZHOO VRUWHG ¶ ¿QH VDQG DQG FRDUVH VLOW SUHGRPLQDWH Sediments in the area are composed mainly of quartz ± SK\OORVLOLFDWHV SODJLRFODVH ± .±IHOGVSDUV± FDOFLWH DQG GRORPLWH (<7%).Other minerals such as opal C/T, anhydrite, siderite, anatase, zeolites, alunite, magnetite, hematite and pyrite are vestigial.Concentrations of Al, Ba, Ca, Ce, Cr, K, La, Li, Mg, Mn, Na, P, Rb, S, Sr, Th, Ti, V, Zn and Zr are included in Appendix 1.The distribution patterns of these chemical elements are subsequently detailed in the statistical analyses section.
Appendix 1 also contains the stations' coordinates, as well as other sedimentological data (grain size, and mineralogical composition) and selected microfaunal results (Shannon index values, foraminifera density, percentage of the most abundant dead and living benthic foraminifera species and groups of those species, as well as the percentage of 'oceanic specimens').Maximum, minimum and median values of those variables are also listed.
The density of dead foraminifera is less than around 2300 tests per gram of sediment.However the abundance of foraminifera is in general very low in many of the studied stations.The percentage of dead and living species was not determined in 20% and 38% of the samples, respectively, due to the reduced number of specimens.
Dead benthic foraminifera assemblages were composed of around 175 species and, among these, around 112 species were found to be alive.The number and the percentage of individuals per species, in each sample, in dead and living benthic foraminifera assemblages are presented in Appendix 2.
The results from discrimination analysis of the spatial variations between dead and living assemblages (considering the samples with dead and living specimens), ap-SO\LQJ WKH %UD\ &XUWLV FRHI¿FLHQW WR D SUHVHQFHDEVHQFH matrix, are reported in Fig. 2.This analysis separates the dead from the living assemblage of most of the samples.
Many of the individuals of these benthic foraminifera VSHFLHV UHÀHFW WKH HIIHFW RI WUDQVSRUW RULJLQDWLQJ IURP hydrodynamic processes.Some species that live on the oceanic bottom are resuspended, by wave action, and introduced in the lagoon, swept up by tidal currents.Bolivina/Brizalina spp.are typical oceanic species, however B. ordinaria and B. pseudoplicata can be found alive in muddy sediments of inner lagoonal areas.Q. seminulum can also be found alive in inner lagoonal areas, away from WKH ULYHUV GLUHFW LQÀXHQFH 6SHFLHV VXFK DV L. ochracea, T. concava, T. haynesi, R. helgolandica and S. gonzalezi can be found in the continental shelf environment.This group of species is an important member of the living assemblages under active currents in the S. Jacinto Channel and in the inlet.However, they are poorly represented in the dead assemblage, since their agglutinated tests are fragile and easily destroyed after the death of the organism.In fact several species that comprise the dead assemblage are not found in the living assemblage of the study area (as can be seen in Appendix 2).
Some of the species found alive in the Ria de Aveiro entrance, such as L. lobatula, P. In accordance with the above, the percentage of damaged empty tests of benthic foraminifera belonging to these species, were considered as having been imported from the adjacent shelf environments and introduced into the Aveiro lagoon.This group was referred to as 'oceanic specimens', as was previously noted above.

Statistical analyses
In general, the spatial distribution of sediment grain size, and its composition in coastal lagoons, may essentially be related to tidal energy, storm-dominated sorting and sediment sources (Anthony and Héquette, 2007).In order to better understand how these forcing factors control the sedimentary processes in the study area, and considering the large number of variables to be analysed, it was considered advantageous to group selected variables with similar occurrence.For this purpose, cluster analysis was performed using Pearson correlation.Several methods to carry out the data agglutination were used; the results obtained with Ward's method were considered coherent, since the variables included in each group have maps with identical distribution patterns.
The results of this analysis, represented in the dendrogram shown in Fig. 3, establish three main groups of variables: *URXS is composed of the sedimentary fraction !ȝP .±IHOGVSDUTXDUW] DQG PHGLDQ JUDLQ VL]H This group is associated with coarser sediments, enriched in quartz and feldspars.This kind of sediment can be found mostly in S. Jacinto Channel and in the lagoon inlet (Fig. 4).*URXS LQFOXGHV WKH VHGLPHQWDU\ IUDFWLRQ ±ȝP plagioclase, carbonates, 'oceanic specimens', Shannon index values and concentrations of K, Rb, Ba, P, Ca, and 6U &RQWHQWV RI ¿QH VDQG IUDFWLRQV ±ȝP KDYH higher positive correlations (p>0.05) with K (PC 0.81), Rb (PC 0.76) and Ba (PC 0.75); its correlation with P is lower but also positive (PC 0.45, p>0.05).7KH PDS RI 6U LV QRW VKRZQ LQ WKLV ¿JXUH LWV GLVWULEXWLRQ pattern is similar to that of Ca since they have high and positive correlation (PC 0.89, p>0.05).'Oceanic specimens' reach higher percentages in the western side of Mira channel, close to the south jetty, and near the head of the north jetty.
Group 3 contains the phyllosilicates, the sedimentary IUDFWLRQV ȝP DQG ±ȝP IRUDPLQLIHUD GHQVLW\ and Cr, Ti, Mn, La, Zr, Ce, Th, V, Na, Al, Li, S, Mg and Zn.The variables of this group are represented in Fig. 7 by the distribution maps of a) phyllosilicates, b) Ti, c) YHU\ ¿QH VDQG IUDFWLRQ ±ȝP G /D H IRUDPLQLIera density, and f) Al.In general, these variables reach higher values in the western side of the Mira channel, close to the south jetty and near the north jetty head, ZKHUH VHGLPHQWV DUH ¿QHU Cluster analysis in mode Q, using Euclidean distances and Ward's method for data agglutination, was applied to a normalised data matrix, with the aim of grouping stations with bottom sediments of similar characteristics.The variables used in the dendrogram of Fig. 3 (cluster analysis in mode R) were also considered in this analysis.Fig. 8 presents the results of this analysis, identifying each group of stations with similar bottom sediment characteristics by a different shade of grey.

Suspended sediment concentrations
Suspended sediment concentrations (SSC) varied between 7.7 mg.l ± and 105.6 mg.l ± .Despite SSC generally increasing from the surface to the bottom, the differences between surface, middle-depth and bottom water samples were very small, during the sampling period in the studied sites (Fig. 1).
Therefore, for each studied depth (surface, middle and bottom), averages of the SSC in each station were determined during neap and spring tides of winter and summer RFHDQRJUDSKLF UHJLPHV IRU GLIIHUHQW WLGDO SKDVHV ÀRRG high tide, ebb and low tide).The results are shown in )LJ 9DOXHV SORWWHG LQ WKLV ¿JXUH VKRZ WKDW WKH KLJKHVW SSC were recorded during winter in the inlet and in the S. Jacinto Channel during different tidal phases.In the inlet, WKH KLJKHVW 66& RFFXU PRUH IUHTXHQWO\ GXULQJ WKH ÀRRG phase, whereas in the S. Jacinto and Mira Channels these occur normally during ebb and low tides.

Suspended sediment grain size and mineralogy
Suspended particulate matter (SPM) in bottom water GXULQJ WKH ÀRRG SKDVH PRVWO\ FRQVLVWV RI ¿QH IUDFWLRQ (<63μm; >70%) in all the study area along the year, independently of winter or summer, spring or neap tides, (Fig. 10a).The contents of silt are similar in bottom water ZLQWHU VSULQJ WLGHV GXULQJ ÀRRG DQG HEE )LJ E In the analysed water samples, the sediment median JUDLQ VL]HV ' RI 630 YDULHV EHWZHHQ ±P DOVR showing the predominance of silt-sized materials.SPM median grain size, in bottom water samples, during win-WHU VSULQJ WLGH UHDFKHG KLJKHU YDOXHV GXULQJ WKH ÀRRG DQG high tide, in the S. Jacinto Channel, and during the ebb phase, in the Mira Channel (Fig. 10c).

Discussion
In all types of coastal environments it is common to ¿QG VHYHUDO FODVVHV RI EHG VHGLPHQWV GXH WR WKH WUDQVSRUW processes operating on regional and local scales, and to different sediment sources, either coming from the continent or from the sea (Waeles et al., 2007).Surface sediments from the Ria de Aveiro inlet and surrounding areas show considerable differences in terms of grain size and geochemical, mineralogical and microfaunal composi-WLRQ )LJV ± %RWWRP VHGLPHQWV RI WKH ODJRRQ LQOHW S. Jacinto Channel and eastern side of the Mira Channel are composed mainly of 'poorly sorted' coarse sand and gravel.However, in the western side of the Mira Channel and at the head of north jetty, which are more sheltered DUHDV µPRGHUDWHO\ ZHOO VRUWHG ¶ DQG µZHOO VRUWHG ¶ ¿QH VDQG and coarse silt can be collected.Thus, several grain popu-ODWLRQV FDQ EH LGHQWL¿HG LQ WKH ERWWRP VHGLPHQWV RI WKH studied area, which are related to different processes of sediment transport.

,GHQWL¿FDWLRQ RI DUHDV RI QHW HURVLRQ
Different suspension patterns for SSC and median grain size were observed in the whole of the studied area (Figs. 9 and 10c, respectively).The highest SSC have been re-FRUGHG LQ WKH 6 -DFLQWR &KDQQHO GXULQJ WKH ÀRRG RI WKH winter neap tide (Fig. 9).During this period SSC in the lagoon inlet were lower.However, an opposite pattern was expected, this is, higher concentrations of suspended sediments in the lagoon entrance, due to the supply of sediments from the nearby oceanic areas and the progressive reduction of concentrations towards inner lagoonal areas (as a result of the progressive deposition of sediments).The fact that the highest recorded SSC values came from the S. Jacinto Channel may be a consequence of the local resuspension of sediments.This means that erosion is occurring on the bottom of the lagoon inlet channel and along S. Jacinto Channel.
Further evidence of the possible resuspension of sediments from the bottom of the lagoon entrance area is the higher values of SSC during low winter spring tide when compared with other tidal phases, in the lagoon inlet and the S. Jacinto Channel (Fig. 9).This may be also related to erosion of the bottom of these channels or of the edges of the nearby channels.The coarser sediments in bottom water SPM (due to the relatively higher content of the VDQG IUDFWLRQ )LJ F E GXULQJ ÀRRG DQG KLJK ZLQWHU spring tide, when compared with the inlet and the Mira channel, may also be indicative of erosion of bottom sediments in this area (Fig. 10c).In fact this erosional nature shows up in the characteristics of the bottom sediments found in central areas of the S. Jacinto Channel, the lagoon inlet and close to the north jetty.Here, very strong tidal currents prevail, where, for example, the magnitude can be in excess of 2 ms í at the centre of the inlet (Vaz et al., 2009).In these areas, coarser bottom sediments composed mostly of quartz and feldspars (Fig. 4b, a), with low concentrations of geogenic chemical elements (such as those included in cluster 3 of Fig. 3 and represented in the distribution maps of samples when looking at the composition of the living and dead assemblages (Fig. 2).Some species found in the area, such as P. mediterranensis, L. ochracea, T. concava, T. haynesi, R. helgolandica and S. gonzalezi, live attached to sand grains.When the organism dies, it is easily removed and transported to other areas by the strong tidal currents.Moreover the agglutinated specimens can easily be destroyed by desegregation.Thus the increased density of foraminifera in bottom sediments indicates the occurrence of zones of sediment deposition (Fig. 7e).
Hydrodynamic processes operating on areas of erosion IXUWKHU OHDG WR VHJUHJDWLRQ RI VLOW DQG YHU\ ¿QH VDQG IUDFtion which are rapidly evacuated from the most hydrodynamic areas located in the inlet and S. Jacinto Channel, and may be trapped in tidal embayments, such as the zone located in the western side of Mira Channel or in more protected areas, such as the one close to the south jetty.The lower-energy bottom currents, in these areas, allow WKH GHSRVLWLRQ RI ¿QHU VHGLPHQWV ZLWK D KLJKHU FRQWHQW of phyllosilicates and carbonates, including foraminifera tests, and higher concentrations of geogenic chemical elements (see Fig. 7).Both these areas are also marked by higher concentrations of K, Ba, P (Figs. 5b, 5c, 5d) and Rb (Cluster 2, Fig. 3), which may indicate a larger oce-DQLF LQÀXHQFH 2WKHU SUR[LHV IRU WKLV NLQG RI VRXUFH DUH Ti, La and Al, in Figs.7b, 7d and 7f, respectively) and foraminifera tests (Fig. 7e), can be found, all a sign of such strong erosive character.This erosional character is represented by white and lighter shaded areas in Fig. 8, which map an integrated synthesis of the results, comparing the sedimentary characteristics of the studied stations.

,GHQWL¿FDWLRQ RI DUHDV RI QHW GHSRVLWLRQ
According to Waeles et al. (2007), grain size and compositional distributions can be related to bottom shear stress, and can also vary depending on the relative energy of the waves and currents.Tidal currents, waves, and FRPSHQVDWRU\ ÀRZV LQ UHVSRQVH WR VWRUP ZLQG DUH WKH most important forces driving the transport of sediments in the lagoon (according to Lopes et al., 2006).
Because of the relatively high settling velocities of sand grains compared to clay-sized particles, their transport adjusts very quickly to dynamic variations.The response of foraminifera tests to the stress of currents can however be much faster than that of quartz grains of equivalent sizes.This is because the former have the dimension of sand particles but have lower density than quartz grains.The hydraulic behaviour of foraminifera tests explains the dissimilarity between the majority of the studied

6HGLPHQW LPSRUW IURP WKH FRQWLQHQWDO VKHOI
Suspended sediments transported landward, through WKH ODJRRQ HQWUDQFH GXULQJ WKH ÀRRG PD\ HVVHQWLDOO\ the percentage of 'oceanic specimens' in the sediments (Fig. 6a).That foraminifera tests that are provided by the oceanic environments is one of the reasons for the increase in diversity (higher Shannon index values; Figs.6a, b).Evidence of sediments originating from oceanic sources is also found in the SSC measured, for instance, GXULQJ WKH ÀRRG WLGDO SKDVHV )LJ VXJJHVWLQJ HQWU\ into the lagoon of materials provided from the adjacent continental shelf.
The deposition processes may be more important on the western side of the Mira Channel when originate from the nearby continental shelf.The values of 66& WKURXJK WKH LQOHW ODJRRQ FKDQQHO GXULQJ WKH ÀRRG are similar in winter and summer, spring and neap tides and are not higher than other sites (Fig. 9).It was expected that higher SSC values would be recorded in the inlet channel during the winter, since, as according to Vitorino et al. (2002a, b), the hydrodynamic conditions are stronger during this season, and storms linked with southerly winds are more frequent.In addition, rainfall is also XVXDOO\ PRUH LQWHQVH OHDGLQJ WR PRUH IUHTXHQW ÀRRGLQJ RI rivers in this season.During winter, the currents on the VKHOI DUH GLUHFWHG WRZDUG WKH QRUWK SROHZDUG ÀRZ GXH WR the weakening of the northerly winds or the prevalence of southerly winds (Vitorino et al., 2002a, b).Otherwise the main sources of sediments for this area are located to the north (Dias et al., 2002).Sediments are being transported to this area of continental shelf by the littoral drift and are mainly supplied by the river Douro, at north of this region (Abrantes and Rocha, 2007 and references herein).However, during the winter the sediments launched onto the continental shelf by this river are transported by the oceanic currents northward, providing little contribution to the study area.
During the spring-summer oceanographic regime, the VWUHQJWKHQLQJ RI QRUWKHUO\ ZLQGV LQGXFHV FXUUHQWV ÀRZing towards the south, onto the continental shelf, favouring the southward littoral drift (Dias et al., 2002).These VHDVRQV DUH LQ JHQHUDO GU\HU DQG WKH ULYHU ÀRZ LV ORZHU however.On the other hand, the layout of the Ria de $YHLUR EUHDNZDWHUV GRHV QRW IDYRXU D ODUJH LQÀRZ RI VHGiments into the lagoon.Thus the contribution of sediment from the ocean to the lagoon should be not very high.

([SRUW RI ¿QHJUDLQHG PDWHULDO WR WKH VKHOI
6XVSHQGHG VHGLPHQW FRQFHQWUDWLRQV 66& ÀRZLQJ seaward, during the ebb (Fig. 9), are higher during the winter season.Picado et al. (2010) and Dias et al. (2000) UHFRUGHG KLJKHU FXUUHQW YHORFLWLHV RQ HEE WKDQ RQ ÀRRG During ebb, as expected, the sediment concentration may QRW EH DV LQÀXHQFHG E\ ZDYH FOLPDWH VLQFH WKH WLGDO ÀRZ is transporting sediments from the marsh located in inner lagoonal areas to the ocean.Thus, the higher SSC YHUL¿HG GXULQJ ZLQWHU IRU WKH SHULRGV RI HEE DQG HYHQ the low tide, can be explained by the availability of sediments during this season which, in turn, can be related to increased rainfall and river runoff.
Most of the sediment introduced into the lagoon by the rivers is silt and clay, which is deposited in low hydro-G\QDPLF DUHDV VXFK DV LQWHUWLGDO ÀDWV DQG VDOW PDUVKHV (Teixeira, 1994).During the winter season stronger tidal currents and strong wind stress, may also contribute to the higher SSC along the main channels, due to higher bottom shear stresses that favour sediment remobilization on channels and marsh surfaces and the advective transport of sediments (Lopes et al., 2006).The ebb jet geographically stable.The most protected zones, with a tendency for sediment accretion, are located, for in-tendency for sediment accretion, are located, for instance, at the western side of the Mira Channel, close to the south jetty and near the head of north jetty.Areas with an erosive character are located along the centre of the inlet and the S. Jacinto Channels.The analysis of SSC also provides evidence that bottom sediments in some zones of these channels are being resuspended.Thus it is possible to assume that the erosive action of tidal currents may have its origin in the natural sinking of these channels.Furthermore the obtained results in this study also raise a concern: some segments of the north jetty may be more vulnerable to erosion.Damage due to strong stresses caused by intense tidal currents in the area is, as such, a real possibility.extends into the shelf as shown by numerical simulations of the ocean circulation in the vicinity of the inlet (Silva et al., 2006), transporting mostly muddy sediments, composed mainly of quartz and phyllosilicates with a lower FRQWHQW RI .±IHOGVSDUVSODJLRFODVH DQG FDOFLWH DV GHPRQstrated in this work (Fig. 10).
Fig. 5 repre-VHQWV PDSV RI GLVWULEXWLRQ RI D ¿QH VDQG DQG E .F %D and, d) P. The distribution pattern of Rb is not presented LQ WKLV ¿JXUH EHFDXVH LW LV VLPLODU WR WKDW RI %D 3& and K (PC 0.96).'Oceanic specimens' have positive correlations (PC 0.78) with the Shannon index.Both these variables also have positive correlations (p>0.05) with carbonates, Ca and Sr. Fig. 6 includes the distribution maps of the following variables: a) 'oceanic specimens', b) Shannon index values (H), c) carbonates and, d) Ca.
7KHVH LQIHUHQFHV DUH VXSSRUWHG E\ WKH ¿QGLQJV RI Plecha et al. 7KHVH DXWKRUV LGHQWL¿HG ]RQHV RI scouring and erosion by comparing the bathymetric evolution, between 2002 and 2005, in the lagoon inlet and the S. Jacinto Channel.Their results show intense erosion between the two breakwaters.