DSpace at VNU: Sedimentation in an estuarine mangrove system

Journal of Asian Earth Sciences 29 (2007) 566–575 www.elsevier.com/locate/jaes Sedimentation in an estuarine mangrove system P Van Santen a,Ô, P.G.E.F Augustinus a, B.M Janssen-Stelder b, S Quartel a, N.H Tri c a Utrecht University, Department of Physical Geography, P.O Box 80115, 3508 TC, Utrecht, The Netherlands b Rijkswaterstaat, Rijksdienst IJsselmeerpolders, Lelystad, The Netherlands c Mangrove Ecosystem Research Division, Vietnam National University, Hanoi, Viet Nam Received 10 April 2003; received in revised form 25 January 2005; accepted 26 May 2006 Abstract Sediment dynamics were studied in a mangrove system in the Ba Lat estuary of the Red River, Vietnam The study area was situated on a riverbank, which consisted of a bare mudXat containing a vegetation gradient from pioneering mangroves to 5- to 7-year-old mangroves Measurements were conducted in the dry season (February/March 2000) and in the wet season (July/August 2000) Actual sedimentation rates were measured using sediment traps and the elevation of the area was surveyed in both periods Two undisturbed sediment cores were taken in the densely vegetated area for 210Pb-dating analysis and showed long-term sedimentation rates of 0.22 g cm¡2 yr¡1 and 0.36 g cm¡2 yr¡1 Sedimentation rates recorded by the sediment traps in the vegetated area were 2.94 g cm¡2 yr¡1 in the dry season and 3.46 g cm¡2 yr¡1 in the wet season In both seasons, the sedimentation rates on the bare mudXat were Wve to ten times higher than in the densely vegetated area However, the elevation of the mudXat was slightly lower in the wet season than in the dry season This implies that erosion must have taken place between March and July This study demonstrates that the bare mud bank of an estuary is highly dynamic until mangroves cover it Sediment delivery to the vegetated zones is low but the protective eVect of vegetation against erosion by waves and currents is strong This results in small but steady actual sedimentation rates under mangroves and therefore sustained long term accretion The study further identiWes the need for similar measurements to be made in diVerent seasons covering the eVects of periodic storm events © 2006 Elsevier Ltd All rights reserved Keywords: Mangroves; Sedimentation rate; Estuary; Red River; Vietnam Introduction Mangroves are tidal forest ecosystems in sheltered saline to brackish environments They are generally known to favour deposition The dense and rigid network of stems, branches and aerial roots causes an increased bottom roughness that results in reduction of currents and attenuation of waves (see also Quartel et al (2006)) However, sedimentation rates under mangroves, based on reliable measurements, are scarcely published and show large variation (Augustinus, 1995) This is especially true for estuarine mangroves Estuaries are characterized by a large spatial and temporal variation in current velocity, current direction, suspension load and related processes of sedimentation and erosion due to the action of river outXow and/or tides Wave * Corresponding author E-mail address: pim.vansanten@gmail.com (P Van Santen) 1367-9120/$ - see front matter © 2006 Elsevier Ltd All rights reserved doi:10.1016/j.jseaes.2006.05.011 activity in estuaries is generally of minor importance Large waves will only occur in the mangrove area during storm or typhoon events with onshore directed winds Estuarine mangroves give natural support to the protection of the riverbanks, due to their accumulative nature To study this eVect, it is necessary to measure sedimentation rates under estuarine mangroves For this reason, a research study has been conducted on a riverbank of the Ba Lat estuary in the Red River delta (Vietnam) The aim of the measurements was: • to demonstrate the current velocity reduction by estuarine mangroves; • to establish the actual accumulation or erosion on an estuary bank with coverage, ranging from bare to dense coverage by mangroves, in the dry as well as in the wet season; • to establish long-term sedimentation rates under estuarine mangroves by 210Pb-dating analysis P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 The Ba Lat estuary The delta of the Red River extends from Viet Tri, a town some 50 km upstream from Hanoi, to the south-east (Fig 1A) It is dissected by several distributaries, which discharge into the Gulf of Tonkin The main stream of the Red River Xows into the sea through the Ba Lat estuary 2.1 Geomorphology The southern part of the Red River delta, including the Ba Lat estuary, has been classiWed as a wave-domi- 567 nated system (Mathers and Zalasiewicz, 1999), according to the classiWcation of Galloway (1975) Sandy bars and cheniers are parallel oriented to the coast (Fig 1B) and develop due to wave action (Van Maren, 2004) The waves rework the initially deposited riverine sediments The sandy ridges are separated by Xat areas which are covered with mangrove swamps or marshes These are the former tidal Xats and lagoons, which have been Wlled up with silt, clay and peat The rivers crossing the delta show relatively straight to irregularly oriented segments, and are characterized as anastomosing streams (Mathers and Zalasiewicz, 1999) Fig (A) Location of the Red River delta in Vietnam; (B) geomorphology of the Ba Lat Estuary, the main branch of the Red River; (C) study area with tripod locations (triangles) and vegetation zones (bordered by thin lines) 568 P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 2.2 Vegetation The coastal vegetation mainly consists of mixed species mangroves Mangroves thrive in the tropics and even extend into the temperate-zone if frost is rare and of short duration The Red River delta is situated in the tropics Nevertheless, the climatic conditions appear to be marginal for the growth of mangroves, because mangroves of the species Kandelia candel appear in dwarf form and other species, e.g., Sonneratia, lose their leaves during wintertime 2.3 Study area The study area is situated at the southern bank of the Ba Lat estuary, approximately km from the open sea (Fig 1B) The Ba Lat estuary is the eastern boundary of the study area Tidal creeks form all other boundaries A small dike along the north-west side protects the inland lying Wshponds (Fig 1C) The study area consists of a bare (intertidal) mudXat in the east, which grades into an area with 5- to 7-year-old mangroves in the west through a zone with pioneering mangroves (Fig 1C) The pioneer plants are mostly Aegiceras corniculatum, 0.7 m high in average Bulrush (Scirpus spec.) comes up in large quantities in this zone in summer The pioneer vegetation towards the west changes into a narrow zone, some 10 m wide, with moderately dense vegetation Here, 0.9 to 1.3 m high mangroves grow in groups, which stand about m from each other Each group is about 2.5 m wide and consists of three to Wve individual shrubs West of this moderately vegetated zone lies the zone with dense vegetation The vegetation is mainly composed of Aegiceras corniculatum (75%) that have a maximum height of 1.85 m in the study area, and Acanthus ilicifolia (20%) The species Kandelia candel, Sonneratia and Avicennia grow in the study area in small numbers The separate shrubs of Aegiceras had a width of 0.55 m at the bottom and 1.20 m at their top Methods This study comprised two measurement campaigns The Wrst lasted from 28 February 2000 to 28 March 2000 and was indicative for the dry season The wet season was represented by measurements conducted in the period 25 July 2000 to August 2000 In the Wrst period, instrumented tripods were used to measure the hydrodynamics Three tripods were installed in the study area (Fig 1) Tripod A was placed in the river at a bed level 0.5 m below the lowest water level (Fig 2) Tripod B is situated in the front zone of the densely vegetated part of the study area, where the vegetation had a moderately dense character Tripod C was installed in the densely vegetated part of the study area Hydrodynamics were measured with an electro-magnetic current meter (EMC) and a pressure sensor The suspended sediment concentration (ssc) was measured with an optical back-scatterance sensor Measurements were done in bursts of 1024 s (17 and s) with a frequency of Hz Burst interval was h The measurements were started on 28 February 2000, at 15:00 h (burst 1408) and ended on 28 March 2000 at 8:17 h (burst 2097) The OBS sensor was calibrated in situ based on sediment concentrations of water samples taken during the measurements Sedimentation rates were measured in diVerent ways • Bed level heights were obtained with a levelling instrument in both measuring campaigns A continuous surface was created by spatial interpolation of the measured bed level The interpolated surfaces of the two campaigns were compared to estimate bed level changes The measurement error associated with levelling is generally 0.5 cm The overall error is estimated to be in the order of 2–3 cm, due to the lack of a Wxed reference point and the muddy bed • Thin (0.4 mm) pieces of canvas of 0.4 by 0.4 m, with a rough surface, were placed on the bed and Wxed with 0.5 cm thick wooden sticks into the bed These canvas pieces acted as sediment traps Most traps were fully covered by sediment after one Xood period and were completely assimilated with its surroundings The traps with the sediment were removed after weeks in February/March and after weeks in July The amount of sediment was measured and for the March traps the fractions of lutum ( 50 m were determined The bed level of the trap locations was measured with a levelling instrument The total time of inundation per trap was estimated with the water level records of the Ba Lat river Accumulation rates of dry sediment were expressed in g cm¡2 yr¡1 • A marker (pure kaolinite) was smoothly dispersed over the bed surface in two locations in the dense mangrove vegetation Two undisturbed samples of the top layers (10 cm) were taken at these two locations at the end of the campaign Thin sections were prepared according to the method described by (Jongerius and Heitzberger (1975)) The thin sections were studied in plain transmitted light and under crossed polarisers, with magniWcations up to ten times • Two sediment cores of the top soil were taken using a 0.09 m diameter PVC tube in the dry season campaign The sediment was dated with the 210Pb method by means of -spectrometry as described by Van Weering et al (1998) and results in a mass accumulation rate expressed in g cm¡2 yr¡1 Dividing this value by the mean dry bulk density over the proWle with decreasing 210Pb values, results in a sedimentation rate in mm yr¡1 Results 4.1 Hydrodynamics The tidal water level oscillations were diurnal with a neap and spring tidal range of 0.7 and 2.5 m, respectively During neap tide, a semidiurnal oscillation of approximately 0.3 m P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 569 A h (m) B 0.2 on (m/s) 0.15 U 0.05 1/3 0.1 Ur (m/s) C 0.5 –0.5 SSC (mg/l) D 600 400 200 02/28 03/03 03/07 03/11 03/15 03/20 03/24 03/28 Fig Time series of (A) waterlevel (h), (B) onshore directed wave orbital velocity U1/3 (m s¡1) at z D 0.09 m, (C) mean current velocity magnitude Ur (m s¡1) at z D 0.09 m (positive is seaward, negative means landward) and (D) suspended sediment concentration SSC (mg l¡1) at z D 0.1 m measured at tripod A was present (Fig 2A) The tripod in the river (A) was continuously submerged with a minimum water level of 0.5 m The mangrove area was Xooded 9–10 h per day during spring tide (Figs 3A and 4A), with a maximum water level of 0.9 m at tripod C During neap tide, this area was Xooded 0–5 h a day with a maximum water level of 0.2 m at tripod C Tidal currents were predominantly diurnal with a small semi-diurnal oscillation during ebb tide Maximum spring tidal currents measured at tripod A were 0.5 m s¡1 seaward directed and 0.25 m s¡1 landward and maximum neap tidal currents are 0.3 m s¡1 seaward directed and 0.15 m s¡1 landward (Fig 2C) Tidal currents hardly existed in the mangrove area (Figs 3C and 4C) Measured tidal currents did generally not exceed 0.03 m s¡1 The currents, measured at 0.09 m above the bed, could occasionally reach 0.1 m s¡1 at tripod B, during the Wnal stages of the ebb tide Wave motions were expressed in terms of the average value of the highest one-third peak onshore wave orbital velocities (U1/3) per burst U1/3 values at tripod A reached 0.1 m s¡1 during low tide (Fig 2B) Maximum U1/3 values at tripod B (Fig 3B) occurred between March and 11 March 2000 and were between 0.15 and 0.2 m s¡1 Maximum U1/3 values at tripod C occurred in the same period and were 0.08 m s¡1 (Fig 4B) Sediment concentrations at tripod A in the river were generally 30 mg l¡1 with peaks of 600 mg l¡1 (Fig 2D) These peaks can hardly be correlated to local measurements, since 570 P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 A 1.5 h (m) 0.5 U1/3on (m/s) B 0.2 0.15 0.1 0.05 C 0.5 Ur (m/s) 0.4 0.3 0.2 0.1 D 600 SSC (mg/l) 400 200 02/28 03/03 03/07 03/11 03/15 03/20 03/24 03/28 Fig Time series of (A) waterlevel (h), (B) onshore directed wave orbital velocity U1/3 (m s¡1) at z D 0.09 m, (C) mean current velocity magnitude Ur (m s¡1) at z D 0.09 m and (D) suspended sediment concentration SSC (mg l¡1) at z D 0.1 m measured at tripod B the peaks are most probably caused by advected clouds of sediment Minimum concentrations at tripod B were also between 30 and 50 mg l¡1 (Fig 3D) Peaks in the sediment concentration on the mudXat-mangrove transition exceeded 300 mg l¡1 and predominantly occurred in the period between and 11 March The related wave orbital velocities were over 0.12 m s¡1 (Fig 3B) The overall trend of the sediment concentration at tripod C was similar to that at tripod B (Fig 4D) The maximum wave orbital velocities of almost 0.2 m s¡1 at tripod B (Fig 3B), resembled a wave bed-shear stress of 0.2 N m¡2 Whitehouse et al (2000) showed that the critical bed shear stress for erosion of muddy beds lies between 0.1 and N m¡2 for beds with a bulk density between 1000 and 1200 kg m¡3 Bulk densities were not measured for the study area, however, values between 1000 and 1200 kg m¡3 are most common for estuarine mud (Whitehouse et al., 2000) The maximum orbital velocity at tripod C was 0.08 m s¡1 (Fig 4B) and resembled a bed-shear stress of 0.049 N m¡2 This value was too small to erode the sediment from the bed 4.2 Actual sedimentation rates The average amount of dry weight of the sediment collected by the sediment traps was 337 g in the dry season campaign and 1513 g in the wet season campaign The amount of sedimentation on a trap was hardly inXuenced P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 A 571 h (m) 0.8 0.6 0.4 0.2 B 0.2 on (m/s) 0.15 U 0.05 1/3 0.1 C 0.5 Ur (m/s) 0.4 0.3 0.2 0.1 D 600 SSC (mg/l) 400 200 02/28 03/03 03/07 03/11 03/15 03/20 03/24 03/28 Fig Time series of (A) waterlevel (h), (B) onshore directed wave orbital velocity U1/3 (m s¡1) at z D 0.09 m, (C) mean current velocity magnitude Ur (m s¡1) at z D 0.09 m and (D) suspended sediment concentration SSC (mg l¡1) at z D 0.1 m measured at tripod C by the diVerences in bed levels of the traps This was also observed by Cahoon and Lynch (1997) In the dry season, the average sedimentation rate in the dense mangrove vegetation, based on trap measurements, was 1.94 g cm¡2 yr¡1 The values ranged from 0.37 g cm¡2 yr¡1 in the back of the mangroves to 4.06 g cm¡2 yr¡1 in the front zone of the dense mangrove vegetation (Fig 5) Similar higher accumulation rates in the exposed front zone of the mangrove vegetation as compared to the rearside was found by e.g., Bird (1986), Lynch et al (1989) and Smoak and Patchineelam (1999) Sedimentation rates in the pioneer zone were much higher and ranged from 4.57 g cm¡2 yr¡1 to 7.38 g cm¡2 yr¡1 (Fig 5) Erosion took place at the riverside, which was shown by a 1–2 cm elevation of the traps above their surroundings In the wet season, the sedimentation rate in the dense mangrove area ranged from 0.86 g cm¡2 yr¡1 in the rear to 5.97 g cm¡2 yr¡1 in the front zone The sedimentation rates varied from 22.52 g cm¡2 yr¡1 to 57.89 g cm¡2 yr¡1 in the pioneer zone The highest sedimentation rate was measured on the bare riverbank: 102.3 g cm¡2 yr¡1 (Fig 5) A marker (pure kaolinite) was dispersed over the surface in two locations under the dense mangrove vegetation during the measuring campaign in March Thin section analysis of the top 10 cm of the sediment demonstrated an average sedimentation on this marker layer of 0.47 mm 572 P Van Santen et al / Journal of Asian Earth Sciences 29 (2007) 566–575 A B Fig Actual sedimentation rates in g cm¡2 yr¡1 per trap measured during (A) three weeks in March and (B) two weeks in July (8.1 mm yr¡1), respectively 0.84 mm (14.6 mm yr¡1) in weeks The sediment collected by the sediment traps in the dry season campaign mainly consisted of silt (between and 50 m) The average amount of silt was 71.5% of which 66% was Wner than 20 m and 34% coarser than 20 m Some 14% of the total amount of sediment on the traps consisted of clay (50 μ m 20–50 μ m 2–20 μ m