THE SOIL FERTILITY EFFECTS OF LEGUMINOUS LEY PASTURES IN NORTHEAST THAILAND .3. ESTIMATES OF SOIL-NITROGEN AVAILABILITY

Paper III of this series reports on effects of ley treatments on the r esults of procedures for assessing soil nitrogen availability and the relationship between the estimates obtained for various availability p rocedures and test crop yield. Nitrogen availability procedures invest igated included the nitrogen mineralisation potential (N(o)) and the r apid procedures waterlogged incubation (WL), extraction with hot 2 M K Cl and extraction with a phosphate-borate buffer (PBB). Values for N(o ) on a composite soil sample were 27 mg N kg-1 using the standard nutr ient solution and 53 mg N kg-1 using a modified nutrient solution (k v alues were also significantly different). Values of N(o) using the mod ified nutrient solution for 0-5 cm soil after the legume treatments sh owed strong treatment effects with values ranging from 44 to 99 mg N k g-1 for the fertilized plots and lower values for the unfertilized plo ts. After two test crops values were reduced but in the fertilized leg ume plots were still higher than for cassava control. An attempt to us e N(o) to predict field nitrogen release was not successful. Values fo r WL (0-5 cm soil fertilized plots) at the end of the ley phase were l ow with significant treatment effects (range 7.6 to 21.8 mg N kg-1). F or the KCl method values were very low (range 2.1 to 5.7 mg N kg-1) ag ain with significant treatment effects. The PBB method gave no signifi cant treatment effects. Yields of the first test crop (roselle) were c orrelated best with WL (R2 = 0.81) followed by organic C (0.73), N(o) (0.69), total N (0.68), KCl (0.51) and PBB (0.33) suggesting WL as the preferred rapid availability index.