Abstract

            In the Mekong Delta, intensive triple rice cropping is depleting soil K. Analytical methods for characterization of various soil K pools need to be evaluated for prediction of soil K deficiency risk using a simulation model. Possible methods include solution K extracted by 0.01M CaCl₂, exchangeable K by 1M NH₄OAc pH7, non-exchangeable K by 1M boiling HNO₃, 0.2M Na TetraPhenylBorate (5 min), and total K by concentrated HClO₄-HF. Two greenhouse experiments were set up, each for 3 rice crops, to study changes in soil K pools: an exhaustion experiment with 19 soils and 5 weeks per crop, and an experiment with 4 soils without or with fertilizer K, and 3 months per crop. Soil K fractions including K(CaCl₂), K(NH₄OAc), K(HNO₃), and K(NaTPB) were measured at the beginning and at the end of each crop in the first experiment and after 3 crops in the second experiment. K(HClO₄-HF) was measured only at the beginning of the experiment.

            In the exhaustion experiment, the order of the amounts of extracted K was 0.01M CaCl₂ < 1M NH₄OAc pH7< 1M boiling HNO₃ < 0.2M Na-TPB (5 min). R² values for the relationships between K fractions and K uptake by one and three crops  were 0.73 and 0.61, respectively for K(CaCl₂), 0.88 and 0.89 for K(NH₄OAc), 0.51 and 0.57 for K(HNO₃), 0.58 and 0.71 for K(NaTPB) if all soils, and 0.79 and 0.90 if only clay soils were taken into account. K(NH₄OAc), which includes both soluble K and exchangeable K, was used as the first pool in a three-pool soil K model. The increased correlation between crop K uptake and K(NaTPB) with successive crops, and the higher R² values for clay soils suggested that K(NaTPB) could be used to describe soil K with intermediate availability to the crop. We developed a 3-pool soil K model (Fig. 1), including (1) labile K (LK) extracted by NH₄OAc, (2) intermediate K (IK) defined as the difference between K(NaTPB) and K(NH₄OAc), and (3) stable K (SK) defined as the difference between K(HClO₄-HF) and K(NaTPB). The flows between these pools in the two experiments were calculated based on the net effects of forward and backward reactions.

Figure 1: The three-pool soil K model. The flows between soil pools are denoted by three letters, the first is always F (flux), and the next two letters stand for the pool from where the flux departs and where the flux arrives, respectively. L stands for labile, I for intermediate,  S for stable K, and U for crop K uptake.

In the second experiment IK was larger, whereas in the first experiment IK was smaller at the end than in the beginning. So, in the second experiment the calculated flows of K leaving the intermediate pool were smaller than the flows entering the intermediate pool, which is unlikely. The reason for the decrease in IK in the first experiment is that under successive exhaustion conditions, the soil had insufficient time to recover. In contrast, in the second experiment the extraction by the crop was less intensive and the soil had more time to recover. Buffering of interlayer K extracted with the short-term (5-min) 0.2M NaTPB method resulted in no measurable increases or decreases of the IK pool in clay soils. In conclusion, K(NH₄OAc) can be used as the labile K pool in the soil K model, but net K(NaTPB) failed to act as the intermediate K pool. A longer extraction period with NaTPB may allow quantifying an additional pool of interlayer K .