FERTILIZER INPUTS, NUTRIENT BALANCE, AND SOIL NUTRIENT-SUPPLYING POWER IN INTENSIVE, IRRIGATED RICE SYSTEMS .1. POTASSIUM UPTAKE AND K-BALANCE

Research in many countries indicates a negative K balance in intensive , irrigated rice systems but comparative studies across different envi ronments are few. Using a uniform sampling methodology, we measured K uptake, K use efficiency, and K balance in six different fertilizer tr eatments of long-term fertility experiments with rice at 11 sites in f ive Asian countries. Depending on the absolute yield level, K uptake r equirements of rice ranged from 17 to 30 kg K per ton of grain. For yi elds greater than 8 t ha(-1), total K uptake exceeded 200 kg ha(-1). T he K balance at most experimental sites was negative, with an average net removal of 34-63 kg K season(-1). There was significant depletion of soil K reserves at many sites. Based on these data, we estimated th at the amount of K cycled annually from the soil into rice plants is 7 -10 million t in irrigated rice systems of Asia. About 1 million t of this total amount is removed with the harvested grain. Present recomme ndations for K addition in most intensive irrigated rice domains are i nsufficient to replace K removal. However, response to K can only be e xpected on soils with deficient supply capacity and where other nutrie nts, particularly N and P, are not limiting. Efficient K management fo r rice must therefore be based on the K input/output balance, the achi evable yield target, and the effective K-supplying power of the soil.