Three experiments were conducted in an acid infertile Oxisol from the Llanos Orientales of Colombia. It was shown that greater increases in nitrogen yield in the tops (N yield) due to N fertilization ofPuerariaphaseoloides (CIAT germplasm accession no. 9900), Stylosanthes capitata no. 1019, Centrosema macrocarpum no. 5065 and Desmodium ovalifolium no. 350 occurred in undisturbed soil cores than in pots of disturbed soil. Inoculation significantly increased N yield ofC. macrocarpum in soil cores, but not in pots.
In screening trials where a range of Rhizobium strains was used to inoculate D. ovalifolium, P. phaseoloides and C. macrocarpum grown in undisturbed soil cores, strains CIAT 2335, 2434 and 1780, respectively, caused the greatest increases in N yield. Inoculation caused greater increases of nodule numbers relative to the uninoculated control in C. macrocarpum than in D. ot,alfolium or P. phaseoloides.
When each legume was inoculated with the most effective strain by different methods, and grown in soil cores, it was found that granulated inoculant (0.5 g/seed) was more effective than seeds pelleted with 50 g inoculant/kg for the small-seeded D. ovalifolium, but was similar in effectiveness to pelleted inoculant for the larger-seeded P. phaseoloides. With the relatively large-seeded C. macrocarpum, granulated inoculant was less effective than pelleted inoculant. No clear differences between different types of pellets were observed.
The results show that undisturbed cores of acid infertile Oxisol can be used to screen for Rhizobium strains tolerant to these adverse soil conditions, although selected strains should then be subjected to further screening in the field, to determine whether they would make suitable commercial inoculants. It may not be possible to obtain maximum responses to inoculation by all legumes, unless the inoculation methods used here are improved.