Potentials and limitations of mycorrhizal inoculation illustrated by experiments with field-grown cassava

Field inoculation trials with cassava (Manihot esculenta Crantz) were conducted in Quilichao (typic Dystropept soil) and Carimagua (Haplustox soil). In Quilichao, with a large and effective native VA-mycorrhizal (VAM) population, inoculation withGlomus manihotis did not increase cassava yields significantly, neither when different sources and levels of inoculum material were used, nor with different cassava cultivars, or after stabilizing soil temperature through mulching. Field inoculation did result in a decrease of the coefficient of variation with respect to yield. The high dependency of cassava on an effective VAM association was indicated by a marked decrease in yield after eradication of native VAM by soil sterilization. In Carimagua, with a lower native VAM population, mycorrhizal inoculation withG. manihotis increased yields significantly at intermediate levels of 100 kg/ha of applied P, using either inoculum of cassava orPanicum maximum roots or inoculum of a soil-root mixture of maize or tropical kudzu. Higher or lower levels of P decreased the effect of inoculation on yield. There were no significant differences among P sources, ranging from highly soluble triple superphosphate to low solubility rock phosphates. Inoculation with different VAM isolates had a variable effect on cassava yields, and showed that there may be an interaction between P fertilizer level and isolate efficiency. It is concluded that there may be a potential to increase yields or decrease the fertilizer P requirements of cassava through field inoculation with effective VAM isolates, in the vast areas of acid infertile Oxisols and Ultisols with low native VAM fungal populations, represented by Carimagua.