/AnMtgsAbsts2009.53407 Spatial and Temporal Dynamics of Water Fluxes in Paddy Rice Systems.

Tuesday, November 3, 2009: 11:00 AM
Convention Center, Room 411, Fourth Floor

Manon Janssen1, Bernd Lennartz1 and Lin Lin2, (1)Institute for Land Use, Univ. Rostock, Rostock, Germany
(2)Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China
Abstract:
In contrast to upland cropping systems, rice paddies are characterized by a flooded and water saturated zone above an unsaturated soil horizon. The vulnerability of wet rice fields to water and solute losses is mainly a function of the impermeability of the plough pan and the earthen banks surrounding the site. This study was initiated to investigate the impact of the cultivation history on water fluxes in paddy systems and to identify relevant mechanisms to be explicitly considered in water flux computations for management purposes.

Water and dye infiltration techniques have been used to quantify water fluxes and visualize flow patterns. The pan structure of a relatively young paddy soil (i.e., 20 years of rice cultivation) differed from that of an older paddy soil (i.e., 100 years) with respect to thickness, bulk density, and macroporosity. Even after 100 years of continuous wet rice cultivation (puddling) a further decrease in the saturated hydraulic conductivity could be observed. Preferential flow processes, are a major cause for water losses in the field itself but also in the bunds. Bio-pores play a major role in water routing, but single shrinking cracks may also be operational weeks after re-wetting. Accordingly, drying and wetting cycles predominantly modify soil properties, also because bio-activity is strongly influenced by the water regime.

       The probability density function (pdf) of the saturated hydraulic conductivity of the plough pan was bimodal in most cases. A simulation model was developed to depict water fluxes in paddies accounting for the spatial properties of the plough pan in a Monte Carlo type approach. The temporal dynamic of the system is explicitly considered by a weighing factor which modifies the distribution of the two peaks in the pdf (preferential and matrix flow). The weighing factor is a function of time elapsed since last ponding.