SPATIAL AND TEMPORAL VARIABILITY OF AMMONIUM IN FLOODED RICE FIELDS

Monitoring and modeling the dynamics of N in hooded rice fields requir e reliable methods for both sampling and analysis. Destructive soil sa mpling and centrifugation and in situ sampling of soil solution using microporous polymer tubes (rhizon soil solution samplers; RSSS) were u sed to study the dynamics and variability of NH4+ in two field experim ents. Both methods gave similar results in unfertilized treatments, wh ereas significant differences were measured in treatments with N appli cation. The RSSS were suited for measuring point dynamics or microvari ation of nutrients in the root zone, whereas soil-core sampling was a better method to obtain quantitative N data integrated across a partic ular depth range. We observed a highly significant linear relationship between the logs of soil exchangeable and solution NH4+ only until pa nicle initiation (PI). Microvariability accounted for a large proporti on of the within-field variation and was significantly reduced by bull ring of soil cores. Coefficients of variation were higher in the ferti lized field compared with an unfertilized held and were changing throu ghout the growth period of rice, with the highest variability between PT and flowering. Sampling to obtain representative mean values at the plot or field scale can be optimized with regard to temporal changes in the spatial distribution patterns of NH4+. Before transplanting (TP ) and at the end of the growth period, less samples are needed and can be taken as composite samples of the whole puddled soil layer. Betwee n TP and flowering, more samples are needed and should be taken in sev eral depth ranges.