『Abstract
　A month-long comprehensive field experiment during the beginning of the flood season of a small mountainous river in southern Taiwan was carried out in 2004. The major goal of the study was to investigate the effect of hydrodynamic sorting related to the salt-water intrusion on the spatial variability of suspended and riverbed sediments and their coupling of different sizes. The experiment included the deployment of an instrumented tetrapod near the river mouth with an upward-looking ADCP and two CTDs mounted at 50 and 100 cm above the bed (cmab), respectively. On three different days along the river, turbidity, salinity and temperature of the water column were profiled; and water samples were taken from the surface and near the bed at different stations. Additionally, one sediment sample was also taken from the riverbed. Suspended sediment concentrations (SSC) were analyzed for five different sizes, i.e.＞500, 250-500, 63-250, 10-63 and 1.2-10μm by filtration. For each riverbed sample the size-composition was analyzed for the subsample that contained lithogenic and nonlithogenic components; and the subsample having major nonlithogenic components (organic matter, carbonates, and biogenic opal) was removed. The grain-size frequency distributions of the riverbed samples were analyzed using a laser particle analyzer. The results were grouped into the following size classes: ＞473, 249-473, 62-249, 10-62, and ＜10μm for comparison with those of the suspended sediments near the riverbed. Some suspended sediment samples were analyzed for POC and PON. Some riverbed samples were analyzed for TOC. Statistical methods of linear regression and Empirical Orthogonal (Eigen) Function (EOF) were used in the data analysis.
　Two-layered estuarine circulation pattern was observed at the tetrapod site. The tidally-driven salt-water intrusion is the major factor influencing the hydrodynamics of the Gaoping River, which in turn, affect the longitudinal and vertical distribution of the suspended sediments and the longitudinal distribution of the riverbed sediments. During the flood, the intrusion front of the salt-wedge creates a dynamic barrier. Upriver from this barrier, the riverbed substrate is coarser, composed of sediments mostly coarser than 249μm. Within the salt-wedge the riverbed substrate is finer, consisting of mostly mud (＜62μm). The barrier creates a trap on the riverbed immediately seaward from the intrusion front, retaining higher percentages of clay-sized sediments and TOC.
　The barrier also creates partition in the terrestrial and marine sources of organic matter in the suspended and riverbed sediments. Within the salt-wedge the major contributor of riverbed TOC is the clay-sized marine sediment transport upriver by the intruding seawater. The terrestrial POC is a minor contributor to the riverbed TOC.
　The riverbed and suspended sediments are coupled. Most size-classes in corresponding suspended and riverbed sediments have a reciprocal relationship (negative feedback) through resuspension and deposition on the water-bed interface. The only size-class of 62-473μm on the riverbed and 63-250μm in the suspension are co-varying (positive feedback). This size class contains largely a transient floc population that is formed and disintegrated in-situ both on the riverbed and in the suspension in the course of a tidal cycle.

Keywords: salt-wedge; dynamic barrier; hydrodynamic sorting; particle size; flocculation; carbon-particle relationship』

1. Introduction
2. Study area and background
3. Field experiments
　3.1. Tetrapod deployment
　3.2. Hydrographic profiling and water and substrate sampling along the river
4. Laboratory analyses
　4.1. analyses of water samples
　4.2. Analyses of riverbed sediment samples
　　4.2.1. Grain-size analysis of subsamples having all components
　　4.2.2. Grain-size analysis of subsamples having only lithogenic components
　　4.2.3. Total organic carbon (TOC) analysis
5. Results
　5.1. Time series measurements on the tetrapod
　　5.1.1. ADCP measurements
　　5.1.2. Sea surface fluctuation and salinity variations with respect to the river discharge and tide
　5.2. Spatial observations along the river
　　5.2.1. Longitudinal variability of water-born constituents
　　5.2.2. Analysis results of riverbed samples
6. Statistical analyses
　6.1. Empirical Orthogonal (eigen) Analysis (EOF)
　　6.1.1. The water column subset
　　6.1.2. The interface subset
　6.2. Linear regression analysis
7. Discussion
　7.1. Grain-size partition related to the salt-wedge
　7.2. Terrestrial versus marine source partition related to the salt-wedge
　7.3. A transient floc population related to the salt-wedge
　7.4. Hydrodynamic sorting and source-to-sink implication related to the salt-wedge
8. Conclusions
Acknowledgements
References