Coastal Engineering 73 (2013) 7183

Contents lists available at SciVerse ScienceDirect

Coastal Engineering
journal homepage: www.elsevier.com/locate/coastaleng

Wave damping over articial Posidonia oceanica meadow: A large-scale

experimental study
Theoharris Koftis a,, Panayotis Prinos a, Vasiliki Stratigaki b
a
Hydraulics Laboratory, Department of Civil Engineering, Aristotle University of Thessaloniki, Thessaloniki, 54124, Greece
b
Department of Civil Engineering, Ghent University, Technologiepark 904, Zwijnaarde, B-9052, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: An experimental study, conducted in the large wave ume of CIEM in Barcelona, is presented to evaluate the
Received 15 September 2011 effects of Posidonia oceanica meadows on the wave height damping and on the wave induced velocities. The
Received in revised form 18 October 2012 experiments were performed for irregular waves from intermediate to shallow waters with the dispersion
Accepted 24 October 2012
parameter h/ ranging from 0.09 to 0.29. Various congurations of the articial P. oceanica meadow
Available online 16 November 2012
were tested for two stem density patterns (360 and 180 stems/m2) and for plant's height ranging from 1/3
Keywords:
to 1/2 of the water depth.
Posidonia oceanica The results for wave height attenuation are in good agreement with the analytical expressions found in
Wave damping literature, based on the assumption that the energy loss over the vegetated eld is due to the drag forces.
Drag coefcient Based on this hypothesis, an empirical relationship for the drag coefcient related to the Reynolds number,
Velocity attenuation Re, is proposed. The Reynolds number, calculated using the articial P. oceanica leaf width as the length
Articial sea grass scale and the maximum orbital velocity over the meadow edge as the characteristic velocity scale, ranges
Large scale experiment from 1000 to 3500 and the drag coefcient Cd ranges from 0.75 to 2.0.
The calculated wave heights, using the analytical expression from literature and the proposed relationship for
the estimation of Cd, are in satisfactory agreement with those measured. Wave orbital velocities are shown to
be signicantly attenuated inside the meadow and just above the ume bed as indicated by the calculation of
an attenuation parameter. Near the meadow edge, energy transfer is found in spectral wave velocities from
the longer to the shorter wave period components. From the analysis it is shown that the submerged vege-
tation attenuates mostly longer waves.
2012 Elsevier B.V. All rights reserved.

1. Introduction rocks being exposed to relatively high wave energy and wind driven
currents (Koch et al., 2006). The importance of seagrasses regarding bi-
Sustainable protection of coastal areas with respect to the marine ological and physical aspects has been well recognized; due to their ca-
ecosystem is of high importance and bioengineering can be a novel pacity to alter their environment, seagrasses have been referred to as
tool for such service. Bioengineering is the use of living materials ecosystem engineers. Seagrass meadows are of great importance for
such as plants or reef builders to alleviate the need of hard construc- maintaining biodiversity since they are highly productive and can
tion measures (de Oude et al., 2010). This need for mitigation of wave serve as important nursery grounds for numerous species of algae,
action and/or ooding and coastal erosion hazards with low environ- sh and invertebrates both above and below the seabed (Green and
mental impact on the coastal environment can be satised with the Short, 2003). Regarding the coastal protection aspect, a service com-
use of natural coastal defense structures such as seagrass meadows. monly listed for seagrasses is sediment and shoreline stabilization,
Seagrasses are marine plants that have roots, stems and leaves. From achieved by slowing down water motion and current ow and by re-
approximately 50 species worldwide (den Hartog, 1977) Posidonia ducing sediment suspension (Borum et al., 2004; Fonseca and
oceanica is the most common seagrass species in the Mediterranean Cahalan, 1992). In the Tigny et al. (2007) eld study, the same effects
Sea and is usually distributed from shallow subtidal waters to a depth are found; P. oceanica meadows signicantly affect the littoral geomor-
of 50 m in clear conditions (Borum et al., 2004). P. oceanica can colonize phology, providing biogenic sediments, controlling beach slope, and
soft substrates such as sand in wave-sheltered areas and also attach to acting as a brake on coastal water masses.
With regard to the wave and seagrass interaction, a complex
water ow system describes the situation, since not only water ow
Corresponding author. Tel.: +30 2310 995877; fax: +30 2310 995672. affects seagrasses and seagrasses affect water ow but seagrasses
E-mail address: [email protected] (T. Koftis). and water ow may interact in highly coupled, nonlinear ways

0378-3839/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.coastaleng.2012.10.007
72 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

(Koch et al., 2006). The degree of wave attenuation depends both on revealed a specic reduction in energy within the canopy. Granata et
the seagrasses' characteristics (the plant's density, the seagrass al. (2001) measured the particle and ow distribution within seagrass
height, the stiffness of the plant and the bending of the shoots) and meadows in a Northeast coast of Spain for both low and high wave and
the wave parameters (wave height, period and direction) so the current activities. The results revealed the 3-dimensionality of the
quantication of wave energy dissipation over seagrasses is difcult meadow, showing that it acts as a bluff body diverting ow over the
to express in a universal way (Mendez and Losada, 2004). meadow, while producing a secondary circulation cell at the
Various laboratory and eld studies on wave attenuation due to meadow's edge. Lowe et al. (2007) performed experiments with
seagrasses have been performed, with large variability of the results rigid cylinders representing a submerged canopy for the quantica-
for wave damping over seagrass meadows that conrm the complexity tion of the wave induced velocities by the introduction of an attenua-
of such ow system. Wave height reduction over vegetated seabed was tion parameter. The results showed that longer-period components in
studied by Fonseca and Cahalan (1992). Four common North American the wave spectrum are signicantly more attenuated than shorter-
seagrass species; Halodule wrihtii, Syringondium liforme, Thalassia period components. Enhancing these experimental results Lowe
testudinum and Zostera marina were harvested from several sites et al. (2008) presented a mathematical model based on porous ow
along the Florida Keys and were placed in a 6 m long wave ume. to account for the wave-driven ow within the canopy. The model
Wave height attenuation was found between 20% and 76% (~40% on unknown parameters such as the friction coefcient, Cf, dimensional
average) over 1 m length when the plants were occupying the entire drag parameter, , and inertial coefcient, CM, were calibrated by the
water depth. Articial plants were used in the experiments of Ota et experimental data. Experiments with a exible articial canopy,
al. (2004) in a 30 m long wave ume. Each stem was formed by four made of polyethylene, with = 920 kg/m 3 and modulus of elasticity
leaves made of polyester and all stems constituted a 6 m long vegetated E = 0.3 GPa were carried out in a 20 m long wave ume in Luhar et
seabed with stem density of 1000 stems/m 2, occupying half the water al. (2010). The experiments were performed for a range of parameters,
depth. The experimental results were compared with a simple numeri- stem density 3001800 stems/m 2 and plant leaves height ranging
cal model based on the linear wave theory and satisfactory agreement from 1/2 to 1/1 of the water depth. The results showed that a local
was found for values of the drag coefcient Cd = 1.3. Bradley and circulation pattern is revealed within the meadow.
Houser (2009) performed a eld study in a microtidal bay in northwest Numerical modeling for such waveseagrass interaction is a de-
Florida, where the main species of the meadows were T. testudinum manding task, since the parameters of the plant stiffness and movement
and H. wrightii. The study was performed for small Reynolds numbers with wave motion are difcult to model. Therefore in early theoretical
(200b Re b 800) and for these conditions they obtained large values and numerical studies, plants have been simulated as rigid cylinders
for the drag coefcient (1.5b Cd b 100) and suggested a relationship for with different values to the drag coefcient (Dalrymple et al., 1984;
the drag coefcient with either the Reynolds or the KeuleganCarpenter Kobayashi et al., 1993). Mendez and Losada (2004) developed an em-
number. The measured wave height decay for submerged vegetation pirical model for wave transformation on vegetation elds that includ-
was found to be described well with an exponential function proposed ed wave damping and wave breaking over vegetation elds on variable
by Kobayashi et al. (1993) and Mendez et al. (1999). The efcient depths. Based on a nonlinear formulation of the drag force, the model
scaling and reproduction of the physical plant's exibility by articial was calibrated for a specic type of plant (Laminaria hyperborea kelp)
meadows have been important features in most recent studies. In and the results were compared with available experimental data.
Elginoz et al.'s (2011) experimental study, the leaves of the articial Chen et al. (2007) developed a model to account for the interactions be-
models were made of Nickel-chrome stripe wire with a density of = tween waves, currents and sediment transport in seagrass systems
7757 kg/m3 and a modulus of elasticity E= 1.98 GPa. The model was within the nearshore circulation model SHORECIRC and the REF/DIF
in a 1:10 scale compared with prototype conditions and this analogy wave model. The drag coefcient Cd accounting for currents and
was selected to reproduce both the wave conditions and the plant's waves was estimated based on an average drag coefcient Cd ~ 1.17
characteristics. The model reproduced a natural meadow with and accounting for the seagrass density and submergence. Li and Yan
87 stems/m 2, the leaf length varied from 0.10 to 0.50 of the water (2007) developed a three-dimensional numerical model to simulate
depth approximately and was placed on a 1:5 slope beach at the end the wavecurrentvegetation interaction using the RANS equations
of the 24 m long wave ume. The results showed that the ratio of the where the extra drag force term due to vegetation takes different values
wave height on the vegetated side to the unvegetated side is between depending on the nature of the ow; unidirectional ow through vege-
0.78 and 0.94. In Snchez-Gonzlez et al. (2011) the experiments tation, wavevegetation interaction and wavecurrentvegetation in-
were conducted also in a 1:10 geometric scale in a 46.3 m long wave teraction. The drag coefcient Cd for the wavevegetation case was
ume. The articial plant leaves were made of polyethylene and calculated using the Mendez et al., 1999 empirical expression related
polypropylene with a density of = 900 20 kg/m3 and a modulus of to the Reynolds number. Suzuki and Dijkstra (2007) used a Volume of
elasticity E =1.6 GPa. The results regarding wave damping were Fluid model to simulate wave attenuation over strongly varying beds
found to follow the exponential decay law and steeper waves were and vegetation elds, both stiff and submerged exible articial vegeta-
found to be mostly attenuated. The experiments were performed in a tion on a sloped bed and a at bed. In Li and Zhang (2010) a 3D RANS
range of 100 b Re b 1500 and the obtained values for the drag coefcient model was employed for the study of the hydrodynamics and mixing
Cd were in the range 0.1b Cd b 1.0 approximately. They showed that the induced by random waves on vegetation. The vegetation was repre-
dependence of Cd on the KeuleganCarpenter number, ranging from 10 sented as an array of cylinders with the drag coefcient expressed
to 250 approximately, is stronger than that on the Reynolds number. with the empirical expression proposed by Mendez and Losada
The attenuation of wave induced velocities has been studied (2004), while the results of the model were in good agreement with ex-
experimentally to address the ability of the plants to slow the water perimental data. Recently Huang et al. (2011) presented a numerical
motion, increase sedimentation and provide efcient protection of model based on the Boussinesq equations, for the study of the interac-
the beaches against erosion. The interaction of ow and seagrass can- tion of solitary waves with emergent rigid vegetation. The model used
opies of Amphibolis antarctica species, which differ morphologically values for the drag coefcient calculated from the experimental part
from more commonly studied blade-like seagrasses such as Zostera of the study, which were in the range of 1.41b Cd b 2.45.
and Thalassia, was performed in the eld study of Verduin and From the above mentioned studies the following main issues regard-
Backhaus (2000). A series of velocity measurements were obtained ing the wave and submerged vegetation interaction arise: (i) the degree
within, above and adjacent to A. antarctica meadows for swell wave of wave height damping (ii) the efcient numerical modeling of such
conditions of the study area (T = 1316.5 s). The results showed an ows that strongly depend on the estimation of the meadow drag coef-
overall damping effect since the power spectra of the velocity data cient and (iii) the attenuation of wave-induced velocities. The
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 73

investigation of these issues is the main objective of the present study. The geometry of the articial plant was chosen in order to reproduce
The large scale experiments, conducted in the CIEM wave ume at the that of the prototype one. Each stem of the articial plant was com-
Universitat Polytecnica de Catalunya, on irregular wave attenuation posed of four leaves with 1 cm width and 1 mm thickness and variable
and wave ow transformations induced by articial P. Oceanica mead- leaf length; one pair of 35 cm long leaves, and another pair of 55 cm
ow are analyzed. In Stratigaki et al. (2011) the results of the same length. The leaves were inserted in a stiff 10.0 cm long rod, made of
CIEM experiments for regular waves over articial P. oceanica meadows PVC, which was then placed in a metal board forming the articial
showed that the damping of wave height depends on seagrass density meadow (Fig. 2). Two different stem density patterns were reproduced;
and submergence ratio. The present study enhances insight regarding a high density conguration with N = 360 stems/m2, representative of
the spectral wave analysis and moreover the estimation of a drag coef- a dense Posidonia meadow patch (following Giraud, 1977 classication)
cient and the calculation of a velocity attenuation parameter. In partic- and an average one, with N = 180 stems/m 2.
ular, the study focuses on the effects of the submergence ratio = hs/h Regarding the mechanical properties for the model used, a dimen-
(hs =height of seagrass, h = water depth), the seagrass density N sional analysis was performed for determining the characteristic
(stems/m2) and the wave period on the wave height attenuation. An parameters using PVC as a material for constructing the articial
empirical relationship for the drag coefcient Cd related to the Reynolds P. oceanica meadow. The density and the modulus of elasticity
number, is obtained, characteristic for the P. oceanica, which can be use- of the PVC material used are 550700 kg/m 3 and 0.903 GPa, respec-
ful in the numerical analysis of such waveseagrass interactions. The ef- tively. Ghisalberti and Nepf (2002) proposed a non-dimensional pa-
fect of the above plant characteristics on the wave induced velocities is rameter 1, for the efcient modeling of movement of the leaves;
also examined. Following the analysis of Lowe et al. (2007) and Manca recently adopted by Snchez-Gonzlez et al. (2011). The parameter
et al. (2010) which showed the spectral wave energy dissipation within 1 is dened as:
submerged canopies is frequency-dependent, an attenuation parameter
of wave induced velocities is calculated for the frequency components W S hs 3
1 1
of the incident wave spectrum. Et2

2. Physical model where hs and t, are the length and thickness of leaves, respectively, and
w and s are the water and plant densities, respectively. Based on the
2.1. Experimental setup above relationship, the model parameter 1, m varies from 0.055 to
0.083 and is within the limits of the prototype parameter 1, p, since
The experiments were carried out in the CIEM wave ume (Canal 0.001 b 1,p b 0.091.
d'Investigaci i Experimentaci Martima) at the Universitat Polytecnica However such a relationship does not account for the effect of the
de Catalunya, Barcelona. The wave ume, 100 m long, 5 m deep and ow velocity. The use of the Cauchy number, dened as the ratio of
3 m wide, offers the ability to perform experiments in a prototype the dynamic pressure and the modulus of elasticity, which character-
scale. The waves are generated by a wave paddle at the left side of the izes the deformation of an elastic solid under the effect of ow is
ume, and a sandy slope beach of 1:15 was formed at the opposite more appropriate for problems of the owplant interaction. The
end, for the elimination of wave reection. A 20 m long horizontal Cauchy number is dened as:
and at sandy area was created in the central part of the ume and
the meadow of articial P. oceanica with length L = 10.70 m was placed w U2 3
Cy S 2
at this horizontal part, as shown in Fig. 1. E

where U is a characteristic velocity and S is the slenderness number, de-

2.2. Properties and scaling of articial P. oceanica
ned as L/l (L = maximum cross-sectional dimension of the plant leaf
and l = minimum cross-sectional dimension of the plant leaf). It should
The physical properties of the plant, such as the density and stiffness,
be noted that this denition includes the slenderness number S which
are important in order to study the wave interaction, the bending of the
needs to be taken into account in the deformability. It is well known
leaves and the resulting wave damping efciently. A typical physical
that the deformation of a slender beam in bending under a transverse
P. oceanica stem is composed of four to eight ribbon-like leaves
surface load is proportional to S3 (Nikklas, 1992). Hence, the ratio
(Cavallaro et al., 2010; Snchez-Gonzlez et al., 2011), each of them
Cy,m/Cy,p (m: model and p: prototype) given in Eq. (3) shows that the
1 cm wide, 1 mm thick and up to 1 m long. The density of the prototype
articial plant used is slightly stiffer compared to the prototype plant.
plant, S, ranges from 800 to 1200 kg/m3 and the modulus of elasticity,
E, ranges from 0.41 to 0.53 GPa as found in Folkard (2005). The plant's !2 !3  
Cy;m Um Sm Ep
stem density can vary from sparse (b150 stems/m 2) found in deeper 0:450:59: 3
Cy;p Up Sp Em
waters to dense (>700 stems/m 2).

Fig. 1. Sketch of the experimental setup of the CIEM ume.

74 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

shallow waters. The wave conditions were selected in order to repro-

duce mild wave conditions of the Mediterranean Sea where P. oceanica
is mainly found. Jonswap spectrum, with a parameter equal to 3.3,
was used for the generation of irregular waves with the signicant
wave height Hs ranging from 0.28 m to 0.40 m, the peak wave period
Tp from 2.0 s to 4.5 s, thus for intermediate to shallow waters,
0.09b h/ b 0.29 and for wave steepness 0.017 b H/ b 0.052. The water
depth in the ume, at the location of the meadow, h, ranged between
1.10 m and 1.70 m, resulting in a range of the submergence ratio
(=hs/h) from 0.32 to 0.50. The wave and meadow characteristics of
the experiments are shown in Table 1. Wave transformation was mon-
itored by 10 wave gauges distributed along the meadow, while veloci-
ties were measured at 12 locations as shown in Fig. 3.

3. Wave height attenuation analysis

3.1. Wave height attenuation

The effect of the seagrasses on the wave propagation over the mead-
ow and the resultant energy dissipation can be seen in Fig. 4, where the
wave energy spectrum is depicted for three characteristic experimental
tests. The spectrum is presented for the resistive wave gauges, WG4,
WG6, WG11, WG9 and WG13 at locations x/L = 0.05, 0.14, 0.61,
0.89 and 1.17, where x stands for the distance from the meadow front
and L = 10.70 m is the total meadow length. A gradual decrease of the
wave energy along the meadow is shown which is due to the drag
of the seagrass meadow. Wave energy dissipation is obvious for all com-
ponents of the wave spectra, especially at peak frequencies. It is also
noticed that the wave spectrum located onshore at x/L= 1.17 shows
a larger wave energy than the one located at x/L =0.89 inside the
meadow. Similar ndings are shown in the wave height variation
Fig. 2. Photo and schematic detail of the articial Posidonia oceanica meadow.
which is depicted in the following gures. Specically, as the wave
exits the submerged vegetation, the wave height slightly increases.
The differences between the wave celerity inside and downstream the
2.3. Test conditions meadow led to this increase, in similarity with the wave propagation
over a submerged porous step (Losada et al., 1997).
A series of experiments were performed for irregular waves propa- The resulting wave height variation over the meadow is shown in
gating over the articial P. oceanica meadow in intermediate and Figs. 5 and 6 for various experimental congurations and for stem

Table 1
Wave and Posidonia oceanica meadow characteristics.

Water depth Signicant Peak wave Wavelength H/ h/ Meadow density Meadow

at meadow wave height period (m) submergence ratio
h (m) Hs (m) Tp (s) (stems/m2) = hs/h

1.70 0.28 2.00 5.92 0.047 0.29 360 0.32

1.70 0.40 3.00 10.69 0.037 0.16 360 0.32
1.70 0.30 4.50 17.34 0.017 0.10 360 0.32
1.50 0.28 2.00 5.78 0.048 0.26 360 0.37
1.50 0.40 3.00 10.21 0.039 0.15 360 0.37
1.50 0.31 4.00 14.37 0.022 0.10 360 0.37
1.30 0.28 2.00 5.60 0.050 0.23 360 0.42
1.30 0.40 3.00 9.67 0.041 0.13 360 0.42
1.30 0.31 4.00 13.50 0.023 0.10 360 0.42
1.10 0.28 2.00 5.36 0.052 0.21 360 0.50
1.10 0.40 3.00 9.04 0.044 0.12 360 0.50
1.10 0.35 3.50 10.81 0.032 0.10 360 0.50
1.10 0.31 4.00 12.53 0.025 0.09 360 0.50
1.70 0.28 2.00 5.92 0.047 0.29 180 0.32
1.70 0.40 3.00 10.69 0.037 0.16 180 0.32
1.70 0.31 4.00 15.17 0.020 0.11 180 0.32
1.50 0.28 2.00 5.78 0.048 0.26 180 0.37
1.50 0.40 3.00 10.21 0.039 0.15 180 0.37
1.50 0.31 4.00 14.37 0.022 0.10 180 0.37
1.30 0.28 2.00 5.60 0.050 0.23 180 0.42
1.30 0.40 3.00 9.67 0.041 0.13 180 0.42
1.30 0.31 4.00 13.50 0.023 0.10 180 0.42
1.10 0.28 2.00 5.36 0.052 0.21 180 0.50
1.10 0.40 3.00 9.04 0.044 0.12 180 0.50
1.10 0.31 4.00 12.53 0.025 0.09 180 0.50
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 75

Fig. 3. Location of resistive wave gauges and current-meters along the meadow.

densities of 180 and 360 stems/m 2 respectively. The wave height

decay is shown as the ratio of H/Ho, where H is the root-mean-
square wave height along the meadow, Hrms, and Ho is the root-
mean-square wave height before the meadow at WG4, Hrms,o. It is
evident that the wave damping depends strongly on the meadow
properties such as the stem density, N and the submergence ratio,
and on the peak wave period, Tp.
This strong relationship between these parameters can be found in
the relevant literature, where analytical solutions have been developed
in order to describe the wave damping over submerged vegetation.
Usually the wave attenuation over seagrasses is described either by an
exponential function shown (Eq. (4)) (Kobayashi et al., 1993) or
by the expression given in Eq. (5) (Dalrymple et al., 1984; Mendez
and Losada, 2004), based on the conservation of energy equation; the
energy loss of waves propagating through vegetation is due to the
work carried out on the vegetation:

H
Kv expki x 4
Ho

H 1
Kv 5
Ho 1 x
Fig. 4. Wave spectrum transformation along the meadow.

where, Kv is a damping coefcient and ki and are parameters related

to the plant and wave characteristics and x is the distance along the can be seen if the dense meadow is considered as a rigid submerged
meadow. Both equations are shown (Dalrymple et al., 1984) to be iden- barrier. This is not the case for the sparse one where the wave height
tical for small values of the arguments ki and respectively, which is the is gradually attenuated. The effect of the submergence ratio on wave at-
case for such wave and seagrass interaction. tenuation is also evident. The degree of wave attenuation is related to
Based on the experimental results, an average value of is obtained the fraction of the water column occupied by the seagrass as expressed
for each test and then the theoretical variation for Kv is calculated from by the submergence ratio and the stem density. This was expected,
Eq. (5) which is also shown in Figs. 5 and 6. The proposed analytical since the wave energy loss over the meadow is due to the drag force in-
expression ts well the experimental results regarding the wave height duced by the vegetation, which is higher when a larger fraction of water
exiting the meadow while in the seaward side small differences are is within the meadow.
observed which are discussed below. The effect of the dispersion parameter h/ on wave attenuation is
The effect of the submergence ratio on wave attenuation is shown in shown in Fig. 8 where Kv is plotted for the same plant congurations
Fig. 7 where Kv is plotted for the same wave conditions (Hs = 0.28 m, (N = 360 stems/m 2 and 180 stems/m 2 respectively, hs/h = 0.50) and
Tp = 2 s) and meadow density N = 360 stems/m2 and 180 stems/m2. variable wave conditions. The longer waves are mostly attenuated
For both densities, the incident wave height (measured at x/L = with a maximum wave attenuation up to 35% observed for h/ =
0.05) is reduced by ~1525%, with a larger attenuation observed for 0.09 and the dense meadow (N = 360 stems/m2). This can be easily
the dense meadow. In the seaward side of the meadow (x/L = 0.05) interpreted based on wave kinematics, since for longer waves the
the wave height increases for the dense meadow (360 stems/m2). highest portion of the water column beneath the free surface is
This is due to the sudden local decrease of the water depth, as under wave motion, so the plant resistance to the ow is stronger.
76 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

Fig. 5. Kv (=H/Ho) variation over the articial P. oceanica meadow length, x, for different experiments (N = 180 stems/m2). (+) experimental results, (line) theory as Eq. (5).

The dependence of the parameter on the above characteristics is since a 100% increase of Tp (from 2 s to 4 s) results in an average
shown in Fig. 9, where values of range from 0.005 to 0.035. For 115% increase of .
these small values for , the above mentioned similarity of Eqs. (4)
and (5) is veried. It is shown that a 50% increase of the submergence 3.2. Estimation of drag coefcient Cd
ratio, (from 0.32 to 0.50), results in an average 117% increase of
. Also a 100% increase of the stem density, N, results in an average To obtain a general expression for estimating the attenuation coef-
80% increase of . The effect of the wave period is also signicant cient Kv which can be considered characteristic for the P. oceanica

Fig. 6. Kv (=H/Ho) variation over the articial P. oceanica meadow length, x, for different experiments (N = 360 stems/m2). (+) experimental results, (line) theory as Eq. (5).
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 77

Fig. 7. Kv (=H/Ho) variation over the meadow for different submergence ratios hs/h (Hs = 0.28 m, Tp = 2 s).

plant, the analysis of Mendez and Losada (2004) is adopted, which is efcient modeling of the hydrodynamic behavior of P. oceanica
an extension of the Dalrymple et al. (1984) approach. The same anal- under wave motion. In literature Cd is usually related to the Reynolds
ysis has been conducted also by Bradley and Houser (2009) number, Re, in the form of Eq. (8):
and Cavallaro et al. (2010) among others. They showed, based on  
the conservation of wave energy equation that the parameter
Cd 8
used in Eq. (5) for random waves, can be expressed as: Re

1 sin h3 kh 3 sin hkh where the coefcients , , depend on the plant characteristics
p Cd bv NHo k 6 (shape of leaves, length and thickness of leaves, density, modulus of
3 sin h2kh 2kh sin hkh
elasticity, and stiffness) with values found in literature shown in
where, Cd is the average drag coefcient of the meadow, bv is the Table 2. The Reynolds number is given as:
plant area per unit height of each vegetation stand normal to the hor-
uc;0 bv
izontal velocity (0.01 m for the present experiments) and k is the Re 9
wave number. In fact Cd is the most important parameter for the cal-
culation of the damping coefcient and is characteristic for each plant where v is the kinematic uid viscosity of water (10 6 m 2/s) and uc,0
related to its biomechanical properties. Based on the results for , the is a characteristic velocity, selected as the maximum orbital velocity
average Cd can be obtained, by rewriting Eq. (6): upward the meadow, at the meadow edge (z = h + hs) and calcu-
p lated by applying the linear wave theory, using Hrms,0 and Tp as the
3 sin h2kh 2kh sin hkh
Cd : 7 wave height and period corresponding to a monochromatic wave
bv NHo k sin h3 kh 3 sin hkh
train:

For such wave seagrass interaction a relationship between Cd and 0 cos hkh
uc;0 : 10
a characteristic nondimensional ow parameter is desired for the sin hkh

Fig. 8. Kv (=H/Ho) variation over the meadow for different h/ (hs/h = 0.50).
78 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

Fig. 9. parameter variation for various experimental congurations.

Fig. 10 shows the measured Cd values against the Reynolds num- Fig. 10. Variation of Cd with a Reynolds number.
ber for both stem densities, together with the plot of Eq. (8) for the
coefcients , and taken from the Mendez et al.'s (1999) study
on rigid plants and Cavallaro et al. (2010) on exible P. oceanica arti- (1000 b Re b 3200). Looking closely to the form of Eqs. (8) and (11)
cial meadow. The Reynolds number considered in the present study and the experimental results as depicted in Fig. 10, the high decrease
varies from 1000 to 3200, while in the above mentioned studies it of the Cd with the increase of Re is shown, for such range of Re. This is
varies from 200 to 15,500. Based on the present experimental results not the case for the storm conditions that correspond to Re in the
for a meadow density of 180 stems/m 2, a new relationship that corre- range of 10,00015,000. For this region Cd has a small decrease with
lates the drag coefcient and the Reynolds number is proposed for the increase of Re and is asymptotic to a specic value, which is calcu-
the plant of P. oceanica with a correlation coefcient R 2 = 0.91: lated as Cd ~ 0.200.30 by applying Eq. (11), Cd ~ 0.40 for the rigid
  plants and Cd b 0.10 for the exible plants, by applying Eq. (8) with
2400 0:77 , and taken from the Mendez et al.'s (1999) and Cavallaro
Cd : 11
Re et al.'s (2010) studies respectively. This comparison demonstrates
the effect of the stiffness of the plants to the estimated Cd and sug-
The results from other experimental studies on wave and sub- gests that the extraction of the outcome of the present study to
merged vegetation interaction (Augustin et al., 2009) show that storm conditions give reasonable results. What is also shown is that
such a correlation can be poor. It should be mentioned that for the the validity of such a relationship in Eq. (8) is subject to the efcient
wave conditions tested, poor correlation was found for the Cd with modeling of the mechanical properties of the articial plant used.
the KeuleganCarpenter number K, in the range of 20 b K b 100. The The validity of Eq. (11) can be checked, through the comparison of
proposed Cd formulation is useful in numerical modeling of such the measured Hrms along the meadow of all the experiments with the
wavesea grass interactions, where the extra terms in the momen- predicted Hrms based on the use of Eqs. (5), (6) and (11). The results
tum equations, due to the drag force, are expressed in terms of Cd. show a satisfactory agreement, with the predicted wave heights
Also in eld applications, the estimation of Cd is simple and based being slightly overestimated at all positions along the meadow,
on measurements of the wave conditions of the study area. except at x/L = 0.79, suggesting that such a relationship can be useful
The results indicate that the meadow stem's density is important for estimating the wave height damping over a P. oceanica meadow
for the evaluation of the proper Cd, as has been shown also in (Fig. 11).
Huang et al. (2011), while the submergence ratio has a minor effect
for the tested conditions (0.32 b b 0.50). It is shown that for the
dense meadow (360 stems/m 2) the estimated Cd is greater than the
sparse case (180 stems/m 2) and the experimental data for Re b 2500
tend to follow the curve for the rigid plants proposed by Mendez
et al. (1999). Actually for mild wave conditions Re b 2500 the denser
meadows are shown not to be inserted into motion by wave action
and the drag force is shown to be the same with the rigid plants' as-
sumption. Moreover it is shown that for Re > 2500 the results for
both densities follow the same trend, showing a possible threshold
for indication of dense (to be modeled as porous ow) or sparse
vegetation.
It should be mentioned that the wave conditions examined
within the present study refer to relatively mild wave conditions

Table 2
Values of coefcients , , found in literature.

Study

Mendez et al. (1999) rigid plant 0.40 2200 2.2

Cavallaro et al. (2010) swaying plant 0 2100 1.7
Present study 0 2400 0.77
Fig. 11. Comparison between measured and predicted Hrms.
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 79

4. Wave induced velocities attenuation analysis are made dimensionless with the maximum free surface velocity at
x/L = 0.07 and are plotted together with the velocity proles
In this section, the data of the orbital velocities attenuation by the derived from the linear wave theory at x/L = 0.07. The velocity at-
articial P. oceanica meadow are discussed and a subsequent spectral tenuation is evident inside the lower canopy position (z/hs = 0.36)
analysis of the data is performed with the calculation of an attenua- as the wave exits the vegetated ume bed (x/L = 0.80), while much
tion parameter. The purpose of such analysis is to obtain insight on above the meadow (z/hs = 1.45) the velocities are slightly attenuated.
the ow structure above and inside the meadow. The complex ow pattern is revealed at the position just above the
meadow edge (z/hs = 1.09) where horizontal velocities are increased
and vertical velocities are slightly decreased due to the higher inter-
4.1. Wave induced velocity attenuation action between waves and moving leaves. The excess of the positive
wave induced horizontal velocities above the meadow (z/hs > 1)
The velocities were measured at three vertical locations, upstream and the negative ones below the meadow (z/hs b 1), which is more
the meadow at x/L = 0.07, inside the meadow at x/L = 0.17 and 2 m obvious upstream the meadow (x/L = 0.07), suggests the existence
from the meadow end, at x/L = 0.80. At each vertical location ADVs of a local circulation pattern.
were installed at four different levels; at 20 cm, 40 cm, 60 cm and In order to analyze the effect of the plant properties on the veloc-
80 cm above the seabed, or at z/hs = 0.36, 0.72, 1.09 and 1.45 respec- ity structure, an analysis based on the power spectrum of the horizon-
tively, where z denotes the distance from the ume bed and hs = tal velocities follows.
0.55 m is the seagrass height (Fig. 3). Fig. 12 shows the power spectrum
of the horizontal velocity at the four levels. Within the meadow, z/hs = 4.2. Velocity attenuation parameter
0.36, the velocities are decreased signicantly for all wave components,
and the maximum value of the spectrum for the peak period is reduced The wave induced velocity spectral densities is calculated for all
by 85% when wave exits the meadow (x/L = 0.80). A small shift of the tests and x/L locations along the meadow. An attenuation parameter
energy from lower to higher frequency components is observed as the (j) is calculated to evaluate the changes in the wave-induced spec-
wave enters the meadow, from x/L = 0.07 to x/L = 0.17, evident at tral ows caused by the canopy and to compare them under different
all four levels. This energy transfer is not clearly shown in the wave test conditions. It was calculated for the lower canopy at each fre-
height spectrum shown in Fig. 4 for this test. The reduction of velocities quency component of the spectrum, following a method similar to
is shown also above the meadow at z/hs = 1.45 for the wave exiting the that developed for coral reefs (Lowe et al., 2007):
meadow (x/L = 0.80). However for the position just above the top of the
seagrass (z/hs = 1.09) the energy at the peak frequencies of the spec- !0:5 !0:5
2 SU;low 2 SU;z20 cm
trum is amplied even at x/L = 0.80, revealing the effects attributed to j j j 12
SIN
U ;high SIN
the movement of the plant leaves. Regarding the vertical velocities, a U z80 cm

gradual decay of the velocities along the meadow is observed at all

levels (z/hs = 0.36, 0.72, 1.09 and 1.45) as the wave exits the meadow where SU,low is the wave-induced velocity spectra obtained from the
(Fig. 13). Moreover the vertical velocities are much lower than the hor- lower ADV (zlow = 0.20 m) at each of the three x/L locations and
izontal ones for these wave conditions, h/ = 0.12. SU IN,high is the wave induced ow spectra obtained from the higher
These ndings are shown more clearly in Fig. 14 where the exper- ADV (zhigh = 0.80 m) at the location upstream the meadow, at
imental vertical distributions of minimum and maximum horizontal x/L = 0.07. The term j is a correction factor to take into account
and vertical velocities are shown along the meadow. The velocities the predicted vertical increase of the ow magnitude with distance

Fig. 12. Power spectrum of measured horizontal velocity for Hs = 0.4 m, Tp = 3 s, h/ = 0.12, hs/h = 0.50 and N = 360 stems/m2.
80 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

Fig. 13. Power spectrum of measured vertical velocity for Hs = 0.4 m, Tp = 3 s, h/ = 0.12, hs/h = 0.50 and N = 360 stems/m2.

from the ume bed. It is derived from the linear wave theory (Dean where zhigh and zlow are the elevations of the highest current-meter
and Dalrymple, 1991) and calculated as: (zhigh = 0.80 m) and lowest current-meter (zlow = 0.20 m). Kj is the
  wave number evaluated for each component of the spectrum. For
cos h kj zhigh each test, the attenuation parameter was calculated at the three x/L
j   13 locations (x/L = 0.07, 0.17 and 0.80) and for spectral components
cos h kj zlow
containing most of the wave energy.

Fig. 14. Vertical distribution of the dimensionless horizontal (above) and vertical (below) velocities for Hs = 0.4 m, Tp = 3 s, h/ = 0.12, hs/h= 0.50 and N= 360 stems/m2. The
velocities are made dimensionless with umax and vmax, maximum horizontal and vertical free surface velocities (z/hs = 2) from the linear wave theory. (+) experimental results,
(line) linear wave theory.
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 81

The spectrum of the of wave induced velocities measured in the

lower canopy (z/hs = 0.36) at the three locations along the ume
(ADV0, ECM0, ECM2 as in Fig. 3) is shown in Fig. 15, for Hs = 0.4 m,
Tp = 3 s, hs/h = 0.50 and N = 360 stems/m2, together with the atten-
uation parameter. Upward the meadow (x/L = 0.07) the attenua-
tion parameter is about equal to one, showing that the reection
due to the meadow is insignicant. As the wave enters the vegetated
bed (x/L = 0.17), it is shown that the velocities of the longer wave
components (Tj > Tp = 3 s) are mostly attenuated, while the short
wave components obtain values j > 1. This reveals a ow eld with
nonlinear interactions, with a partly energy transfer from the longer
to the shorter period wave components. This is more evident from
Fig. 16 where the attenuation parameter is depicted as calculated
for Hs = 0.4 m, Tp = 3 s and N = 360 stems/m 2, at the three locations
along the meadow (x/L = 0.07, 0.17 and 0.80). The small values of
j ~ 0.4 at the leeward side of the meadow (x/L = 0.80) indicate the
efcient velocity reduction for all the submergence ratios at the
lower canopy (z/hs = 0.36). The wide reduction on the velocities
above the seabed indicates the efciency of the meadows in minimiz-
ing sediment suspension under wave action and eventually erosion.
Minor attenuation in wave energy is observed from the lower sub-
Fig. 15. Spectra of wave induced velocities measured in the lower canopy (z/hs = 0.36)
at the three locations along the meadow (above). Attenuation parameter for the spec- mergence ratio, particularly for the shorter period wave components.
tral components Tj (below). The inuence of the stem density on the attenuation parameter is not

Fig. 16. Attenuation parameter (j) calculated for each wave period (Tj) of the wave-induced velocity spectra component. Hs = 0.4 m, Tp = 3 s, N= 360 stems/m2.

Fig. 17. Attenuation parameter (j) calculated for each wave period (Tj) of the wave-induced velocity spectra component. Hs = 0.4 m, Tp = 3 s, hs/h = 0.42.
82 T. Koftis et al. / Coastal Engineering 73 (2013) 7183

Fig. 18. Attenuation parameter (j) calculated for each wave period (Tj) of the wave-induced velocity spectra component. Hs = 0.4 m, Tp = 4 s, hs/h = 0.50.

signicant as the wave exits the meadow (x/L = 0.80), as shown from 5. Conclusions
Fig. 17 for the same wave conditions (Hs = 0.4 m, Tp = 3 s) and sub-
mergence ratio (hs/h = 0.42). However at x/L = 0.17, the velocities The main objective of this study was to evaluate the effects of the
are reduced for the dense meadow (j ~ 0.8) while for the sparse P. oceanica meadow on the wave height formulation and on the wave
one they are found unaltered. This is in accordance with the ndings orbital velocities. Full scale experiments were conducted in the CIEM
from the wave height analysis, where it was shown that the denser ume in Barcelona and several wave conditions and plant congura-
meadow acts as a submerged barrier as the wave enters the vege- tions were tested. The following conclusions can be derived:
tated seabed. Similar results are obtained for the longer wave period
test (Hs = 0.4 m, Tp = 4 s) and for hs/h = 0.50 shown in Fig. 18; longer 1. The wave height distribution follows well the proposed analytical
wave components are mostly attenuated from the canopy, which is in expression found in literature, of the form 1/1 + x, where x is
accordance with the ndings of the wave height analysis. Similar re- the distance from the meadow boundary and parameter depends
sults were found also in the experimental study of Lowe et al. (2007). on the plant and wave characteristics. Parameter is shown to de-
If the velocities measured at the higher located ADVs, at z = 40 cm pend strongly on the meadow submergence ratio and stem density
(z/hs = 0.72) and z = 60 cm (z/hs = 1.09), are used for the calculation and on the wave conditions. Experimental values for indicate the
of j in Eq. (11), the results for the attenuation parameter are shown similarity of this expression to another commonly used, in such
in Fig. 19. Regarding the location upstream the meadow (x/L = wave and vegetation interaction, exponential decay law.
0.07) the attenuation parameter shows that it is unaffected by the 2. For the dense meadow, in its upward side, the wave height slightly
meadow (j ~ 1) but for the location inside the meadow the variation increases due to the sudden local decrease of the water depth.
for the results seems to be affected by the leaves' motion. The atten- The dense meadow acts as a rigid submerged barrier.
uation parameter uctuates arbitrarily between 0.8 and 1.4 at both 3. The degree of the wave attenuation observed is related to the frac-
locations (x/L = 0.17 and 0.80), with no specic dependence on the tion of the water column occupied by the seagrass as expressed by
wave period. This reveals that near the meadow edge the motion of the submergence ratio and the stem density. The wave damping
the plant leaves highly interacts with the water as indicated by increases with an increasing submergence ratio and stem density,
Koch et al. (2006). as expected, since the wave energy loss over the meadow is due

Fig. 19. Attenuation parameter (j) for the cases as in Fig. 17 as calculated for the ADVs at z = 40 cm and z = 60 cm (SU,low = SU,40 and SU,low = SU,60 respectively).
 T. Koftis et al. / Coastal Engineering 73 (2013) 7183 83

to the drag force induced by the vegetation, which is higher when Dalrymple, R.A., Kirby, J.T., Hwang, P.A., 1984. Wave diffraction due to areas of energy
dissipation. Journal of Waterway, Port, Coastal, and Ocean Engineering 110, 6779.
a larger fraction of water is within the meadow. Dean, R.G., Dalrymple, R.A., 1991. Water Wave Mechanics for Engineers and Scientists.
4. The longer waves are mostly attenuated from the meadow with a World Scientic.
maximum wave height damping ~ 35% observed for h/ = 0.09 den Hartog, C., 1977. Structure, function and classication in seagrass communities. In:
McRoy, C.P., Helfferich, C. (Eds.), Seagrass Ecosystems, A Scientic Perspective.
and the dense plant. This was expected since for longer waves, Marcel Dekker, New York, Basel, pp. 89121.
the highest portion of the water column beneath the free surface Elginoz, N., Kabdasli, M.S., Tanik, A., 2011. Effects of Posidonia oceanica seagrass
is under the wave motion, so the plant resistance to the ow is meadows on storm waves. Journal of Coastal Research SI 64, 373377.
Folkard, A.M., 2005. Hydrodynamics of model Posidonia oceanica patches in shallow
stronger. water. Limnology and Oceanography 50 (5), 15921600.
5. An empirical relationship for the drag coefcient Cd related to Fonseca, M.S., Cahalan, J.H., 1992. A preliminary evaluation of wave attenuation by four
the Reynolds number is obtained. For mild wave conditions, species of seagrass. Estuarine, Coastal and Shelf Science 35 (6), 565576.
Ghisalberti, M., Nepf, H.M., 2002. Mixing layers and coherent structures in vegetated
Re b 2500, Cd is shown to be the same with the rigid plants' as-
aquatic ows. Journal of Geophysical Research 107 (C2), 111 (3).
sumption for the dense meadow. The results for the calculated Giraud, G., 1977. Essai de classement des herbiers de Posidonia oceanica (L.) Delile.
wave heights using the proposed relationship are in satisfactory Botanica Marina 487491.
agreement with that of the measured ones, suggesting that such Granata, T.C., Serra, T., Colomer, J., Casamitjana, X., Duarte, C.M., Gacia, E., 2001. Flow
and particle distributions in a nearshore seagrass meadow before and after a
a relationship can be useful for estimating the wave height storm. Marine Ecology Progress Series 218, 95106.
damping over a P. oceanica meadow. Green, E.P., Short, F.T., 2003. World Atlas of Seagrasses. UNEP-WCMC.
6. Regarding the velocity structure, it is found that inside the mead- Huang, Z., Yao, Y., Sim, S.Y., Yao, Y., 2011. Interaction of solitary waves with emergent,
rigid vegetation. Ocean Engineering 38, 10801088.
ow and just above the ume bed, the wave orbital horizontal and Kobayashi, N., Raichlen, A.W., Asano, T., 1993. Wave attenuation by vegetation. Journal
vertical velocities are signicantly decreased. The velocity attenua- of Waterway, Port, Coastal, and Ocean Engineering 119, 3048.
tion parameter was found to be j ~ 0.4 for all frequency compo- Koch, E.W., Sanford, L.P., Chen, S.-N., Shafer, D.J., Smith, J. Mckee, 2006. Waves in
seagrass systems: review and technical recommendations. Technical Report,
nents of the wave spectrum. This shows explicitly, the ability of ERDC TR-06-15. US Army Corps of Engineers.
the meadows to diminish the sediment suspension under wave ac- Li, C.W., Yan, K., 2007. Numerical investigation of wavecurrentvegetation interaction.
tion and provide protection against erosion. The velocities at the Journal of Hydraulic Engineering 133 (7), 794803.
Li, C.W., Zhang, M.L., 2010. 3D modelling of hydrodynamics and mixing in a vegetation
meadow edge are increased and the wave kinetic energy at the eld under waves. Computers & Fluids 39 (4), 604614.
peak frequencies of the spectrum is amplied, while it is reduced Losada, I.J., Patterson, M.D., Losada, M.A., 1997. Harmonic generation past a submerged
mainly at the low frequencies. porous step. Coastal Engineering 31, 281304.
Lowe, R.J., Falter, J.L., Koseff, J.R., Monismith, S.G., Atkinson, M.J., 2007. Spectral wave ow
7. The longer wave components of the kinetic energy spectrum are
attenuation within submerged canopies: implications for wave energy dissipation.
mostly attenuated compared to the short wave components, which Journal of Geophysical Research 112, C05018.
is in accordance with the ndings of the wave height analysis. The Lowe, R.J., Shavit, U., Falter, J.L., Koseff, J.R., Monismith, S.G., 2008. Modeling ow in
partial energy transfer from the longer period to the shorter period coral communities with and without waves: a synthesis of porous media and
canopy ow approaches. Limnology and Oceanography 53 (6), 26682680.
wave components, reveals the ow pattern inside the meadow due Luhar, M., Coutu, S., Infantes, E., Fox, S., Nepf, H., 2010. Wave-induced velocities inside a
to the effects introduced in the ow by the plant leaves. model seagrass bed. Journal of Geophysical Research 115, C12005.
Manca, E., Stratigaki, V., Prinos, P., 2010. Large scale experiments on spectral wave
propagation over Posidonia oceanica seagrass. Proceedings of the 6th International
Acknowledgments Symposium on Environmental Hydraulics, Athens, Greece, vol. 1, pp. 463468.
Mendez, F.J., Losada, I.J., 2004. An empirical model to estimate the propagation of random
The experiments were conducted within the frame of the Hydralab breaking and nonbreaking waves over vegetation elds. Coastal Engineering 51,
103118.
III EU project 022441 (RII3). The authors gratefully acknowledge the Mendez, F.J., Losada, I.J., Losada, M.A., 1999. Hydrodynamics induced by wind waves in
assistance of the CIEM laboratory staff. a vegetation eld. Journal of Geophysical Research 104 (C8), 1838318396.
The support of the European Commission through FP7.2009-1, Nikklas, K., 1992. Plant Biomechanics: An Engineering Approach to Plant Form and
Function. Univ. Chicago Press, Chicago.
Contract 244104 THESEUS (Innovative Technologies for Safer Ota, T., Kobayashi, N., Kirby, J.T., 2004. Wave and current interactions with vegetation.
European Coasts in a Changing Climate), is also acknowledged. Proceedings of 29th International Coastal Engineering Conference, World Scientic,
Singapore, pp. 508520.
de Oude, R., Augustijn, D.C.M., Wijnberg, K.M., Dekker, F., de Vries, M.B., Suzuki, T.,
References
2010. Bioengineering in front of a River dike: wave attenuation by vegetation.
Proceedings of the 6th International Symposium on Environmental Hydraulics,
Augustin, L.N., Irish, J.L., Lynett, P., 2009. Laboratory and numerical studies of wave
Athens, Greece, vol. 1, pp. 253258.
damping by emergent and near emergent wetland vegetation. Coastal Engineering
Snchez-Gonzlez, J.F., Snchez-Rojas, V., Memos, C.D., 2011. Wave attenuation due to
56, 332340.
Posidonia oceanica meadows. Journal of Hydraulic Research 49, 503514.
Borum, J., Duarte, C.M., Krause-Jensen, D., Greve, T.M., 2004. European Seagrasses:
Stratigaki, V., Manca, E., Prinos, P., Losada, I.J., Lara, J.L., Sclavo, M., Amos, C.L., Cceres, I.,
An Introduction to Monitoring and Management. Monitoring and Managing of
Snchez-Arcilla, A., 2011. Large-scale experiments on wave propagation over
European Seagrasses Project.
Posidonia oceanica. Journal of Hydraulic Research 49 (Suppl.1), 3143.
Bradley, K., Houser, C., 2009. Relative velocity of seagrass blades: implications for wave
Suzuki, T., Dijkstra, J., 2007. Wave propagation over strongly varying topography: cliffs
attenuation in low-energy environments. Journal of Geophysical Research 114,
and vegetation. Proceedings of 32nd IAHR Congress, Venice, Italy (CD-ROM).
F01004.
Tigny, V., Ozer, A., De Falco, A.G., Baroli, M., Djenidi, S., 2007. Relationship between the
Cavallaro, L., Lo Re, C., Paratore, G., Viviano, A., Foti, E., 2010. Response of Posidonia
evolution of the shoreline and the Posidonia oceanica meadow limit in a Sardinian
oceanica to wave motion in shallow-waters preliminary experimental results.
coastal zone. Journal of Coastal Research 23, 787793.
Proceedings of the International Conference on Coastal Engineering, 1(32)
Verduin, J.J., Backhaus, J.O., 2000. Dynamics of plantow interactions for the seagrass
(Shanghai, China).
Amphibolis antarctica: eld observations and model simulations. Estuarine, Coastal
Chen, S.N., Sanford, L.P., Koch, E.W., Shi, F., North, E.W., 2007. A nearshore model to
and Shelf Science 50, 185204.
investigate the effects of seagrass bed geometry on wave attenuation and
suspended sediment transport. Estuaries and Coasts 30, 296310.