Protective functions of coastal forests and

         trees against natural hazards


                   Eric Wolanski

             Australian Institute of Marine Science

        PMB No. 3, Townsville MC, Queensland 4810, Australia.

            E-mail: e.wolanski@aims.gov.au



Abstract
On the basis of field data as well as fluid and soil mechanics and structural engineering, it
is argued that mangroves and coastal forests need to be restored and even created to
enhance the protection of coastal areas from salt spray, wind damage, and erosion from
wind-driven waves and typhoon waves. These forests also save human lives during a
tsunami below a threshold level that depends on the tsunami wave height and the
characteristics of the forest. Mangroves also provide important ecohydrological services
such as providing self-scoured navigable channels, sheltering coastal seagrass beds and
coral reefs from excess sedimentation, and enhancing fisheries; these are all resources
that the human population living along tropical estuaries and coasts rely on for their
livelihood and quality of life. Mangroves are very susceptible to below-root level erosion
from boat wakes and shallow water wind waves near low tide.

  1. Introduction
    Tidal wetlands, including mangroves and salt marshes, are essential to the
maintenance of biodiversity of many estuaries by trapping sediment, converting nutrients
to plant biomass, trapping pollutants and serving as a habitat for fish and crustaceans
(Wolanski et al., 2004). They also enhance offshore fisheries as well as biodiversity in
seagrass and coral reefs offshore (Manson et al., 2005; Wolanski et al., 2003). They also
protect the coast from wave erosion (Mazda et al. 1997 and 2006; Prasetya 2006). They
have helped saved human lives in the 2004 Indian Ocean tsunami (Kathiresan and
Rajendran 2005; Vermaat and Thampanya 2006). Nevertheless mangroves are
increasingly destroyed and degraded by human activities (see the review papers in this
volume by Preuss 2006; Prasetya 2006; Latief and Ladi 2006). This paper argues that
coastal forests and mangroves need to be restored and even created, to enhance the
capacity of estuaries and coastal waters to provide ecological services to the human
population living on its shores, as well as to protect the coast from wind damage, salt
spray, coastal erosion, and even in saving human lives during a tsunami.


  3. Protection against coastal erosion by absorption of wave energy
Wind waves and typhoon waves
Mangroves are blunt bodies that can absorb water wave energy as a result of wave-
induced reversing and unsteady flows around the vegetation. The reduction of wave
energy can be estimated from fluid mechanics principles. The vector of the horizontal
force F is the sum of the inertial Fi and drag Fd forces (Figure 1; Milne-Thompson 1960):
        F = Fi + Fd                             (1)
where
        Fi = Cm ρw V a                            (2)
        Fd = 0.5 ρw Cd (D Z) |u| u                      (3)
where Cm (≈ 1.5 for fully turbulent flows) and Cd (≈0.45) are the inertia and drag
coefficients, u is the water velocity, V is the displaced volume of the body, ρw is the
density of water, a is the acceleration, D the diameter of the vegetation assumed
cylindrical, Z is the depth to which the vegetation is submerged. It is readily possible to
extend the theory for a tapered stem where D diminishes with elevation (Niklas 2000).
    For an unbroken wave of period T,
    u = U cos (kx – ωt)                             (4)
    a= 2 π U cos (kx – ωt) / T                          (5)
where
    U = π H cosh k(z+h) / T sinh(kh)                       (6)
where h is the water depth, H the wave height, k is the wave number defined from the
dispersion relation
    ω2 = g k tanh (kh)                              (7)
where g is the acceleration due to gravity.
    For tidal inundation and wind or typhoon waves, the vegetation is only partially
submerged; therefore Z=h. Mangroves thus protect the coast from wave erosion by
absorbing wave energy through the drag and inertial forces (Massel et al. 1999).
Probably the best data set on this process is that of Mazda et al. (1997) at the muddy
coast of Vietnam’s Thai Binh province where Kandelia candel trees have been planted at
1 m intervals in a strip 1.5 km wide (toward offshore) and 3 km long (along the coast;
Figure 2a). There was a wave swell of period 5-8 seconds entering the forest. They
measured the rate of wave reduction, r, per 100 m of mangroves in the direction of wave
propagation,
              r =(HS-HL) / HS                     (8)
where HS and HL are the wave heights at the offshore edge of the mangrove forest and
100 m inshore in the mangroves respectively. r varied between 0.2 for 5-6 year old
mangroves and 0.05 for 1 year old mangroves. Within six years after planting the trees
have grown sufficiently that the wave height of 1 m at the open sea was reduced to 0.05
m at the coast (Figure 2a), enabling aquaculture ponds behind a coastal levee. Without
the sheltering effect of mangroves the waves would arrive at the coast with wave height
of 0.75 m (Figure 2b) and the levees would have been eroded and breached. Mazda et al.
(2006) repeated that study for a Sonneratia plantation at sea protecting the Vinh Quang
coast, also in Northern Vietnam. They found (Table 1) that because of their
pneumatophores, the rate of wave reduction is much higher by up to a factor 3 for
Sonneratia forests than for Kandelia candel forests (Mazda et al., 2006); leaves are
important in absorbing wave energy.

Table 1. Wave reduction, r (in %) per 100 m of adult mangrove plantation (Data from
Mazda et al., 1997 and 2006). The value of r without mangroves was about 5% next to
the Kandelia candel site and 10% next to the Sonneratia site.

Mangrove     Water depth
species      (m)
         0.2        0.4        0.6        0.8
         20         20         18         17
Kandelia
candel
         60         40         30         15-40
Sonneratia

    Another quantified example of mangroves protecting the coast is that of India’s
Bhitarkanika mangroves following the October 1999 super cyclone with a wind speed of
around 260 km h−1 and a storm surge of about 9 m that hit the Orissa coast. The
economic impact of this cyclone was evaluated by Badola and Hussain (2005) for three
villages equidistant from the seashore and with similar aspects with different protection;
the damage included household damage by the wind, inundation of crops, loss of
fingerlings, and salt intrusion. The losses incurred per household were greatest (US$ 154)
in the village that was not sheltered by mangroves and had a dyke that failed, followed by
the village that was neither in the shadow of mangroves or the dyke (US$ 44), and the
losses were the least for the village that was protected by mangrove forests (US$ 33).
    Thus mangroves are efficient in protecting the coast.

Tsunamis
The tsunami propagating in a mangrove is initially a broken wave. The horizontal
velocity u is (Hedges and Kirkgoz 1981)
        u = 0.5 (g H)1/2                           (9)
The trees absorb some of this wave energy through energy dissipation by drag forces.
Because of the long duration of the tsunami wave, the trees cannot stop the wave. They
can however transform this broken wave into a flood; the shock wave effect is reduced
and thus human lives can be saved and damage to property is lessened, as long as the
trees survive flattening, trunk breaking or overturning as has been documented in the
worst affected areas in the December 2004 Indian Ocean tsunami (Figure 3). The
propagation of a 5 m tsunami at the shore over a flat terrain that was either bare ground or
heavily forested with mature trees of either Kandelia candel or Sonneratia was predicted
using a dam break model (Chanson, 2005). The predictions (Figure 4) suggest that at a
point 500 m from the shore the water depth rises 1 m in 77 sec for a bare ground, in 343
sec for Kandelia candel and 727 sec for Sonneratia.
    The empirical evidence that mangroves helped save human lives is very strong.
For instance Sri Lanka data suggest that in the Yan Oya River the 7 m tsunami in
December 2004 reduced to 0.5 m 3.5 km upstream (point b in Figure 5a), and in the
Mahaweli Ganga River the tsunami wave was 12 m at the cost, 4 m in the
Mudduchchenal Village (point b in Figure 5b) protected by a sand dune, and only 2 m at
a point a similar distance as the village from the shore but additionally protected by trees.
The saving in human lives and damage to property is substantial (Figure 6) though not
100% (Danielsen et al. 2005; see also
http://eqtap.edm.bosai.go.jp/useful_outputs/report/hiraishi/data/papers/greenbelt.pdf).
    Trees are thus effective as long as they are not uprooted or snapped at which case
they form debris that may destroy human lives and property. Trunk breaking occurs if the
horizontal breaking strength σs (≈ 185 ± 35 MN m-2) of the tree at the base is exceeded
by inertial and drag forces (Niklas 2000),
             4 (Fi + Fd) / 3 ≥ π (D/2)2 σs                 (10)
For typical values of σs of healthy trees, this mode of trunk breaking is unlikely. Breaking
instead seems to result from the overturning moment of forces Fi and Fi that act at en
elevation η (η = H/2 or L/2, whichever is the smallest, where L is the height of the
vegetation). The overturning moment lifts the upstream edge of the basal area and forces
down the downstream edge, creating tension at the upstream edge and compression at the
downstream edge. Trunk breaking occurs if tension exceeds the breaking strength in
tension σt of the tree,
         -G/ 2 π (D/2)2 + η (Fi + Fd) D/2 / 0.25 π (D/2)4 ≥ σs         (11)
or the breaking strength in compression σc of the tree,
         G/ 2 π (D/2)2 + η (Fi + Fd) D/2 / 0.25 π (D/2)4 ≥ σc          (12)
where G is the weight of the tree. The calculations suggest that, for typical material shear
strength of timber, for a 3 m tsunami wave in the mangroves, 3-m tall trees will snap and
for a 6 m tsunami wave, 8-m tall trees will snap.
    Trees overturn with their roots if the overturning moment exceeds the soil
resisting moment (Peck et al. 1973; Bowles 1988),
         (Fi + Fd) η/2 > π R (D1/2)2                      (13)
where (Figure 1) R is the force of resistance of the soil to shearing at the interface
between the root matrix and the soil and D1 (D1 ~4 D) is the width of the root system, and
         R = c + P tan θ                            (14)
                     -2
where c is the cohesion (c ~ 75 kg m for compacted clay, c=0 for pure sand), P is the
normal stress, and θ is the angle of internal friction (θ ~ 40o for sand, θ ~ 0 for clay). The
calculations suggest that a 3 m tsunami is thus able to uproot an isolated 3 m mangrove
tree. However if the trees grow close together, then R1 can become much larger as a
result of interlocking of roots from adjoining trees. Thus densely vegetated trees of a
height > 6 m may be able to sustain a 6 m tsunami wave. There is some evidence for that
in the observations of the 2004 Indian Ocean tsunami.
    Mangroves can thus survive and attenuate tsunami waves by transforming them in
a flood instead of a shock wave. Thus they can save human lives and decrease property
damage (Figure 6). However if the wave exceeds a threshold level, a catastrophic failure
of the mangrove ecosystem occurs whereby snapped or uprooted trees are carried by the
currents as debris to destroy tress in its path and create debris that harms people and
property. From models this threshold can be determined by the size and the density of the
mangrove vegetation, the root and soil structure, and the size of the tsunami.
    All those results agree qualitatively with observations of the impact on mangroves
of the 2004 Indian Ocean tsunami (Figure 3 a-d).
    Non-mangrove trees such as casuarnia and palm trees also exerted a similar
impact on absorbing tsunami wave energy, though their threshold level was smaller,
about 2-3 m even for fully grown trees, at which the trees were snapped or uprooted
(Figures 3e; http://river.ceri.go.jp/rpt/asiantsunami/en/survey.html).

Erosion from wind waves and boat wakes
Mangroves and other trees fringing rivers, estuaries and coasts are however not efficient
at absorbing small water waves from boat wakes and shallow-water wind waves at low
tide when the wave erodes the soil below the root level. When the erosion starts, the
erosion can be swift, taking sometimes only a few weeks, until the undercut trees fall
down in the water (Figure 3f).

  4. Protection of the coast against wind and salt spray
In the same way that the water current through a mangrove forest is decreased by the
vegetation that exerts a drag force on the water, the vegetation also exerts a drag force on
the air flow through the canopy. The wind through a tree canopy (Figure 7a) is decreased
throughout the height of the tree, H, particularly at leaves’ height. This diminishes the
momentum of the air exiting the canopy and thus shelters the area downstream by a wake
effect to a distance L rarely longer than 5 H (Figure 7b), as dictated by classical fluid
dynamics. Because the flow around vegetation is three-dimensional, the sheltering
distance L (i.e. the length of the wake) can be significantly increased if the trees are
grown in a wide shelterbelt (one or two rows of trees whose canopy touches each other)
and forest belts (multiple rows); then L ≈ 20H if the wind blows perpendicular to the
shelterbelt and L < 5H if the wind blows parallel to the shelterbelt (Takle et al. 2006). A
wide shelterbelt forms a long turbulent wake in its lee (Figure 7c).
     This sheltering effect diminishes if the tree is partially or partially defoliated
(Figure 7 d-e) and disappears if the tree is overturned (Figure 7f). Large trees are
overturned more readily than small trees (Figure 6f) because they are exposed to stronger
winds as a reason of the wind shear near the ground (Figure 7a). Experience with
cyclones in Australia and hurricanes in the USA shows that full defoliations of fully
developed mangrove mature trees happens only during super cyclones winds, and is
typically restricted to a strip less than 50 m wide (M. Williams, pers. comm.).
     The foliage of some non-mangrove trees such as casuarina is less hardy and the
tree can be defoliated for severe but non-cyclonic winds. These trees often also are much
more readily snapped or overturned by the wind; they thus constitute an unreliable wind
bioshield during a typhoon.
     Trees also offer significant protection against salt spray, i.e. fine salt particles
carried in suspension in air (Figure 7b; Takle et al., 2006). Just like the water flow
through mangrove forests creates less turbulent wakes behind trees where the suspended
mud deposits, the air flow downwind of trees creates less turbulent zones where the
suspended salt particles deposit. This effect extends to a distance ≈ 10 H downwind of a
windbreak. Thus a wide shelterbelt projects a wide turbulent wake that is very efficient
at capturing salt particles in suspension (Figure 7 c-d). This process traps the salt spray
near the coast, facilitating its return to the sea rather and preventing the salt to pollute inland
soils.

  5. Ecohydrological services provided by mangroves
Mangrove provide important ecological services (Badola and Hussain 2005; Blaber 1997;
Prasetya 2006; Preuss 2006). Three key ecohydrological services provided by mangroves
are summarised below.

a.   Enhancement of estuarine and coastal fisheries
Mangroves are an important component of tropical estuarine and coastal ecosystems.
Mangroves provide for marine animals a refuge from predators, high levels of nutrients,
and shelter from physical disturbance (Manson et al. 2005). Typically 1 ha of mangrove
forests support 100-1000 kg yr-1 of marine fish and shrimp catch; for the Mekong this
catch is about 450 kg yr-1 (de Graaf and Xuan 1998).
    Mangroves are also important in supporting fish. An ecohydrology model was
applied to Darwin Harbour, Australia, and calibrated against field data. Darwin Harbour
covers 3227 km2. The land drainage area is 2417 km2, about 10% of which is mangrove.
The harbour is a macro-tidal estuary with three arms, each of which drains a seasonal
river with negligible flow in the dry season. During that time, Darwin Harbour is an
inverse estuary. Seawater enters the estuary as a result of evapotranspiration. In the dry
season the flushing time of water is about 20 days. This period of time for is sufficient
for the biology to become important. The model reveals that the mangroves are essential
for the fisheries, because if they were destroyed the carnivorous fish would decrease by
70% and the detritivores by 50%, in the mangrove-fringed reaches where they are
presently abundant (Wolanski et al., 2006).

b. Trapping of fine sediment
    Muddy waters enter the mangroves at rising tide, deposit some of the suspended
sediment in quiet zones near slack high tide in the mangroves, and return to the estuary
with less sediment; the difference between the mud that it comes in and what goes out is
the sediment trapping (Figure 8). As a rule of thumb, as sketched in Figure 9a, a
mangrove that covers 3.8% of the river drainage area traps about 40% of the riverine mud
inflow; of the remaining 60%, 20% contributes to estuarine siltation and 40% is exported
to coastal waters. This relationship is independent of land use in the catchment, holding
true for catchment sediment yield in the range 1.5 to > 150 tons km-2 year-1 (Victor et al.
2004 and 2006).
    Mangroves fringing muddy open waters trap large amounts of mud from coastal
waters (e.g. Figure 9b).
    Thus mangroves also can shelter coastal seagrass and corals from excessive
sedimentation.

  c. Maintaining navigable channels
Mangrove creeks are self-scouring (Wolanski et al., 2001). During spring tides there
exists a marked tidal asymmetry of the currents in the channel or estuary that drains the
mangroves, the peak ebb tidal currents at the mouth of the creek being measurably larger
than the peak flood tidal currents (Figure 10). If the mangrove area decreases from
mangrove land reclamation, the creek silts. Examples of that abound in areas where
developers, such as prawn farmers in Thailand, have reclaimed mangrove land. In the
case of the Klong Ngao estuary in Thailand where half of the mangrove land was
reclaimed, the tidal creek has been observed to silt within 5-10 years so that it now dries
up completely at low tide. In its natural state it was navigable even at low tide
(Wattayakorn et al. 1990). This sedimentation can be even more rapid when the sediment
inflow has increased in the river catchment as a result of poor land use (Wolanski and
Spagnol 2000).

6.   Discussions

Bioshields
Mangroves and coastal forests provide important ecological services that provide a
livelihood for the local population and help improve their quality of life. They help
protect the coast from wave erosion from wind-driven waves and typhoon waves.
Mangroves and coastal forests can save human lives and property in a tsunami, by
transforming a shock wave into a flood, at least up to a threshold level. They also shelter
the inland from salt spray and wind damage up to a threshold level. These findings are
observed in the field and they are science-based because they result from well-known
principles of fluid and soil mechanics, structural engineering and ecology.
     Thus the benefits of mangroves and coastal forests against natural hazards, as well
as to improve the quality of life of the human population, are numerous. Where possible
thus it is necessary to conserve and even create mangroves and coastal forests and this is
compatible with economic and social development (Wolanski 2006).
     Creating these bioshields is very much soft technology that involves tree
plantations on land as well as restoring former mangrove areas or even planting over
coastal areas.

Restoring mangroves
Recreating mangroves from abandoned shrimp ponds is the ‘easiest’ option; it has the
additional benefit to enable some parallel commercial activities. For instance fish and
crab ponds in the mangrove-fringed tidal creeks can be successful. However planting
mangroves in shallow shrimp ponds as is tried in Vietnam largely fails (B.Clough, pers.
comm.). Planting seedlings in reclaimed shrimp ponds fails if the natural drainage
pattern is not properly restored and the seedlings rot in stagnant water (Figure 11a). This
is not always successful. The degree of success depends on local expertise in finding and
selecting seedlings, growing seedlings in nurseries, local soil conditions, the choice of
locations, the elevation of the substrate and frequency of tidal inundation, the acidity and
salinity of the degraded soils, and on water currents, tides and storms (see a review of the
technology by Prasetya 2006). Soil preparation is needed for former mangrove soils that
have turned acidic when used as shrimp ponds.

Mangroves at sea
There have even been mangrove forests created at sea by planting seedlings over a wide,
intertidal area above mean sea level; this was done successfully over a muddy substrate in
Vietnam (Mazda et al. 1997 and 2006), with moderate success over a muddy sand
substrate in Thailand (e.g. in Eritrea, see Figure 11b; in Thailand’s Chumphon Bay;
Brown and Limpshaichol, 2000), and has somewhat failed in the upper Gulf of Thailand
due to persistent wave erosion (Figure 11c). Where waves dominate, attempts have been
made, with mixed success, to protect the young trees until it can survive the waves.
Some techniques include (1) to let the seedling grow much longer in a nursery until it
becomes a small tree that can better resists waves, (2) to plant the seedlings in hollows
within solid structures such as tyres or hollows in a concrete structure, (3) to plant the
seedlings in transparent PVC tubes (Figure 11d), and (4) to protect the seedlings behind
temporary structures such as bamboo walls. Mixed success is achieved, because in this
case people try to plant mangroves in wave-dominated areas where they would not occur
naturally.

Sacrificial bioshields
The seaward edge of the forest (mangrove or non-mangrove) is the sacrificial zone (Takle
et al., 2006).; it takes the brunt of the damage due to “flagging” by wind, high salt
impaction, and occasionally total defoliation in typhoons, and the destruction of the tree
in typhoons and tsunami. It must be managed as such, by accepting that this zone is
unsteady and unstable and thus not allowing human settlements, and by allowing for
natural regeneration, replanting of destroyed trees, and possibly even building artificial
fences to better protect the first line of trees from lethal salt and wind damage in high
impact zones.

Limitations and constraints
     In practice the level of applicability of these science-based principles depends on
the local climate, hydrology, drainage patterns, meteorology, oceanography, soil types,
waves, storm patterns, and socio-economic drivers. Climate change may complicate the
situation. There is no unique recipe; nevertheless the failure to recognize these science-
based principles will always result in environmental and socio-economic degradation.
The applicability of these principles to various sites is more to do thus with the
willingness and the capacity of individual countries to adopt mangroves and coastal
forests as bioshields, on a case-to-case basis. The level of adoption will depend on socio-
economic imperatives. Where the coast has been totally urbanized, by planned
developments or slums, that solution is then in practice impossible to implement. In other
cases the practical use of coastal forests and mangroves is feasible to protect people and
property and enhance their livelihood and quality of life. The protection that bioshields
offer is far from total, it involves risks because the usefulness of bioshields has limits in
extreme events. The bioshield solution has the advantage that it is often practical,
inexpensive by comparison with pure engineering solutions such as the Dutch solution of
dyking the coast, relatively low technology, and it protects and enhances the
environmental services provided by estuaries and coastal waters on which the population
often depends for their livelihood. It is thus a matter of living within accepted risks.
The level of risk will vary from site to site with details of the bathymetry of the coastal
waters and the topography of the coastal areas, the geology, the meteorology, and the
oceanography. Judging from typhoon statistics alone, no two sites are alike in Asia,
suggesting that the relevance and use of bioshields will vary spatially just as much (Takle
et al. 2006).
     What science needs to provide now is a synthesis of all this physical and
biological knowledge so that the risk can be quantified for individual sites for various
choices of bioshields’ characteristics, size and location. This is very much a new science;
unfortunately at present the risk can be qualified from basic science processes and
empirical evidence, but it cannot be quantified yet for predictive purposes for specific
applications of bioshields at individual sites.
7.   References

Badola R., Hussain S.H. 2005. Valuing ecosystem functions: an empirical study on the
   storm protection function of Bhitarkanika mangrove ecosystem, India.
   Environmental Conservation 32: 1-8.
Blaber S.J.M. 1997. Fish and Fisheries of Tropical Estuaries. Fish and Fisheries Series
   no. 22, Chapman and Hall, London. 367 pp.
Bowles J.E. 1988. Foundation analysis and design. 4th ed., McGraw Hill, New York.
Brown B., Limpsaichol P. 2000. Carbon cycling in a tropical coastal ecosystem, Sawi Bay,
   southern Thailand. Phuket Marine Biological Center Special Publication 22, 86 pp.
Chanson H. 2005. Analytical solution of dam break wave with flow resistance. Application to
   tsunami surges. In B.H. Jun, S.I. Lee, I.W. Seo, G.W. Choi (eds.), 31st IAHR Congress, 11-
   17 September 2005, Seoul, Korea. pp. 3341-3353.
Danielsen, F., Sorensen, M.K., Olwig, M.F., Selvam, V., Parish, F., Burgess, N.D.,
   Hiralshi, T., Karunagaran, V.M., Rasmussen, M.S., Hansen, L.B., Quarto, A.,
   Suryadiputra, N., 2005. The Asian Tsunami: a protective role for coastal vegetation.
   Science 310, 643.
de Graaf G.J., Xuan, T.T. 1998. Extensive shrimp farming, mangrove clearance and
   marine fisheries in the southern provinces of Vietnam. Mangroves and Salt
   Marshes 2: 159-166.
Duke N.C., Wolanski, E. 2001. Muddy coastal waters and depleted coastlines – depleted
   seagrass and coral reefs. In Wolanski E (ed.), Oceanographic Processes of Coral
   Reefs: Physical and Biological Links in the Great Barrier Reef. CRC Press, Boca
   Raton, Florida, 356 pp.
Hedges T.S. and Kirkgoz N.S. 1981. An experimental study of the transformation zone
   of plunging breakers. Coastal Engineering 5: 353-370.
Kathiseran K., Rajendran N. 2005. Coastal mangrove forests mitigated tsunami.
   Estuarine, Coastal and Shelf Science 65: 601-606.
Latief, H., Hadi S. 2006. The role of forests and trees in protecting coastal areas against
   tsunamis. This volume
Manson F.J., Loneragan N.R., Skilleter G.A., Phinn, S.R. 2005. An evaluation of the
   evidence for linkages between mangroves and fisheries: A synthesis of the literature
   and identification of research directions. Oceanography and Marine Biology: An
   Annual Review 43: 483-513.
Massel S.R., Furukawa K. and Brinkman R.M. 1999. Surface wave propagation in
   mangrove forests. Fluid Dynamics Research 24: 219-249.
Mazda Y., Magi M., Kogo M. and Hong P.N. 1997. Mangroves as a costal protection
   from waves in the Tong King Delta, Vietnam. Mangroves and Salt Marshes 1, 127-
   135.
Mazda Y., Magi M., Ikeda Y., Kurokawa T., Asano, T. 2006. Wave reduction in a
   mangrove forest dominated by Sonneratia sp. Wetlands Ecology and Management
   14: 365-378.
Milne-Thomson L.M. 1960. Theoretical hydrodynamics. Macmillan, London.
Niklas K.J. 2000. Computing factors of safety against wind-induced tree stem damage.
   Journal of Experimental Biology 51: 797-806.
Peck R.B., Hanson W.E. and Thornburn T.H. 1973. Foundation Engineering, 2nd ed.. J.
   Wiley and sons, New York.
Prasetya G.S. 2006. The role of coastal forest and trees in combating coastal erosion. This
   volume.
Preuss, J. 2006. Coastal are planning and management using forests and trees as
   protection from tsunamis. This volume.
Riley R.W., Kent C.P.S. 1999. Riley encased methodology: principles and processes of
   mangrove habitat creation and restoration. Mangroves and Salt Marshes 3: 207-
   213.
Takle E.S., Chen , T.-C., Wu X. 2006. Protective functions of coastal forests and trees
   against wind and salt spray. This conference.
Vermaat J. and Thampanya U. 2006. Mangroves mitigate tsunami damage: A further
   response. Estuarine, Coastal and Shelf Science 69: 1-3.
Victor S., Golbuu Y., Wolanski E. and Richmond R. 2004. Fine sediment trapping in two
   mangrove-fringed estuaries exposed to contrasting land-use intensity, Palau, Micronesia.
   Wetlands Ecology and Management 12: 277-283.
Victor S., Neth N., Golbuu Y., Wolanski E. and Richmond R.H. 2006. Sedimentation in
   mangroves and coral reefs in a wet tropical island, Pohnpei, Micronesia. Estuarine, Coastal
   and Shelf Science 66: 409-416.
Wattayakorn G., Wolanski E., Kferfve B. 1990 Mixing, trapping and outwelling in the
   Klong Ngao mangrove swamp, Thailand. Estuarine, Coastal and Shelf Science
   31: 667-688.
Wolanski, E. 1992 Hydrodynamics of mangrove swamps and their coastal water.
   Hydrobiologia 247, 141-146.
Wolanski E., Spagnol S., Ayukai T. 1998. Field and model studies of the fate of
   particulate carbon in mangrove-fringed Hinchinbrook Channel, Australia.
   Mangroves and Salt Marshes 2: 205-221.
Wolanski E., Spagnol S. 2000. Environmental degradation by mud in tropical estuaries. Regional
   Environmental Change 1: 152-162.
Wolanski E., Mazda Y., Furukawa K. et al. 2001. Water circulation through mangroves and its
   implications for biodiversity. In Wolanski E (ed.), Oceanographic Processes of Coral
   Reefs: Physical and Biological Links in the Great Barrier Reef. CRC Press, Boca Raton,
   Florida, 356 pp.
Wolanski E., Richmond R., McCook L., Sweatman H. 2003. Mud, marine snow and coral reefs.
   American Scientist 91: 44-51.
Wolanski E., Boorman L. A., Chicharo L., Langlois-Saliou E., Lara R., Plater A.J., Uncles R.J., Zalewski M.
   2004. Ecohydrology as a new tool for sustainable management of estuaries and coastal waters.
   Wetlands Ecology and Management 12: 235-276.
Wolanski E. 2006. The Environment in Asia Pacific Harbours. Dordrecht, The
  Netherlands: Springer, 497 pp.
Wolanski E., McKinnon D., Alongi D., Spagnol S. and Williams D. 2006. An
  ecohydrology model of Darwin Harbour. pp. 477-488 in Wolanski E. (ed.), The
  Environment in Asia Pacific Harbours. Springer, Dordrecht, 497 pp.
Figure 1. Idealised wind and typhoon waves (left) and a tsunami wave (right) through
trees showing terms used in text.
                                         b



                    a

Figure 2. (a) A map of the mangrove-fringed Thuy Hai coast in the Thai Binh Province,
Vietnam. Groups A, B and C are mangrove plantations comprising respectively 0.5 year-
old trees, 2-3 year-old trees and 5-6 year-old trees. The symbols • indicate the field
measurements sites of tides, waves and currents of Mazda et al. (1997a). (b) A sketch of
the wave field at that site (top) with and (bottom) without mangroves. Adapted from
Mazda et al. (1997).
a
                       b




c                                            d




                                          f
                    e


Figure 3. (a-d) photographs of mangroves damaged by the Indian Ocean 2004 tsunami.
(a-b) mangroves snapped. Note in (a) that the mangroves in the background were
sheltered by the 20-m wide coastal strip of mangroves that were flattened by the tsunami.
(c) mangrove tree uprooted with its roots. (d) the catastrophic failure of mangroves
occurs when large trees break and are carried as debris destroying smaller trees, property
and human lives. (e) Other trees beside mangroves, such as this palm tree, were also
uprooted or snapped, by this tsunami. (f) Mangroves destroyed by small wind-driven
waves undercutting the banks at low tide on the Daly Estuary, Australia. A similar effect
results from boat wakes. Sources: Muntadhar et al., V. Selvam, the Internet, the author.
Figure 4. Model predictions of the rise of water level at a point 500 m from the shore for
a flat terrain following a 5 m tsunami at the shore, for three scenarios, namely a bare
ground, mature Kandelia candel forest and mature Sonneratia forest. The trees are
assumed not to be destroyed by the wave. Time starts when the tsunami arrives at that
point.
a




                  b

Figure 5. Maps of the mouths of the (a) Yan Oya River and (b) the Mahaweli Ganga
River, together with estimate of maximum water level during the Indian Ocean 2004
tsunami. Source: http://river.ceri.go.jp/rpt/asiantsunami/en/survey.html
Figure 6. For the December 2004 Indian Ocean tsunami, plot of property loss (per capita
2005US$, left) and mortality (lives lost per 1000, right) in 18 hamlets on the coast of
Tamil Nadu against distance from the shore (km) and elevation above mean sea level
(m), respectively. Different symbols depict different coastal types. The fitted curve is to
the pooled data set of 18 hamlets. Reproduced from Vermaat and Thampanya (2006).
                                             b


             a




                     c
                                          d




                                        f
                  e

Figure 7. (a) Vertical profile of wind speed u through a 13.1 m high wild cherry tree, 0.4
m in diameter at 0.5 m from ground level, growing in an open site. The velocity is
normalised by the maximum wind speed U measured at the top of the canopy Silhouette
of tree shows all branches for which d >5 mm. Reproduced from Niklas et al. (2000).
Sketch of the air flow around, over and through (a) a single tree and (b) a shelterbelt,
showing that that the wind accelerates over and around the vegetation, and a turbulent
wake forms behind the tree; SSC = suspended salt concentration. Down arrows =
deposition of the salt particles in the wake. Mangroves in Florida defoliated (d) partially
and (e) totally by the wind in a hurricane, and (f) overturned by the hurricane wind; large
trees are overturned more readily than small trees. Source: USGS Florida.
Figure 8: Composite diagram illustrating the net fluxes of suspended sediment in the
mangrove swamp originating from the tidal creek. About 80% of the sediment entering
the swamp at rising tide remained trapped in the swamp and was not exported back into
the creek at falling tide. The sediment traps reveal that the bulk of the suspended
sediment deposited within 50 m from the edge of the creek. Reproduced from Wolanski
et al. (2001).
                   a


                        b

Figure 9. (a) Sketch of the fate of riverine mud in tropical, muddy, estuaries fringed by a
mangrove that covers 3.8% of the river catchment area. (b) Sketch of the fate of riverine
mud in the mangrove-fringed open waters of Hinchinbrook Channel, Australia (Adapted
from Wolanski et al., 1998).
Figure 10. Time series plot over one tidal cycle of the observed currents at the mouth of
a 5km long mangrove creek in Missionary Bay, tropical Australia, at different depths
from near the surface (O), in mid-water (h), to near the bottom (o), during a spring tidal
cycle when the mangroves were fully inundated at high tide. The creek receives no
freshwater except direct rainfall. The thin line is the sea level at the mouth. Note that the
peak currents are larger at ebb tide than at flood tide. Reproduced from Wolanski (1992).
                                           b
                        a




                                d
                  c

Figure 11. (a) In this mangrove plantation in a reclaimed shrimp pond near Surat Thani in
Thailand, mangrove seedlings died in an area where water ponded. (b) Planting of
mangrove seedlings in Eritrea where mangroves did not exist
(http://www.ramsar.org/wwd/3/wwd2003_rpt_eritrea1.htm). (c) Efforts to plant
mangroves along the north coast of the Inner Gulf of Thailand are somewhat
unsuccessful due to accelerated wave erosion since the natural mangroves were destroyed
and replaced by shrimp ponds; hard structures are necessary to combat wave erosion. (d)
Protection of mangrove seedlings in the Riley encased method (Riley and Kent 1999;
http://mangrove.org/video/rem.html).