PERSISTENCE AND EFFECTIVENESS OF RHIZOBIA NODULATING PROMISCUOUS SOYBEANS IN MOIST SAVANNA ZONES OF NIGERIA

The symbiotic performance of promiscuous soybeans depends upon the pop ulation size, effectiveness and survival of indigenous or introduced r hizobia in the field, A pot experiment was conducted using soils colle cted from 13 farmers' fields located in two agroecological zones (nort hen and southern Guinea savanna) in the moist savanna of Nigeria to de termine the relationships between growth response to previous rhizobia l inoculation and the indigenous rhizobial populations. At each farmer 's field, soil was collected in plots which were planted the previous year (1993) to: (1) maize, (2) soybean cv. Bossier, (3) soybean cv, TG X 1456-2E and (4) soybean cv, TGX 1660-19F, The soybean cultivars TGX 1660-19F and TGX 1456-2E were either uninoculated or inoculated with a n enriched population of a mixture of local rhirobial strains. Four te st plants: soybean cultivars TGX 1660-19F (slightly promiscuous), TGX 1456-2B (highly promiscuous), Bossier (non-promiscuous) and cowpea (ty pically promiscuous) were planted in pots containing soils from the ab ove field treatments. Previous inoculation increased shoot dry matter production by an average of 32% over the uninoculated controls in 94 o f the 312 (30%) legume inoculations and farmers' field combinations wh ile the indigenous rhizobia were more effective than the introduced on es in 103 combinations (33%) with an average of 20% increase compared to previous inoculation treatments. Previous inoculation increased bio mass yield of both promiscuous and non-promiscuous soybean varieties. The response to previous inoculation treatments was farmers' fields de pendent and inversely related to the numbers of rhizobia in the soil. Soil rhizobial population ranged from 0 to > 400 cells g(-1) soil and response to inoculation often occurred when numbers of indigenous rhiz obia were fewer than 10 cells g(-1) soil, Numbers of indigenous rhizob ia were generally lower or below detection limit in soils previously c ropped to maize. These results indicate a relationship between rhizobi a cell counts and promiscuous soybean responses, which may be used to indicate under which conditions inoculation will be beneficial to a fa rmer.