Recent studies have documented increases in biological nitrogen fixation (BNF) and grain yield in common beans (Phaseolus vulgaris) following biochar application to the soil. However, the mechanism for this response has remained unknown. It is also not known whether this result varies by biochar feedstock or production conditions. A greenhouse experiment was done to isolate the individual components of the biochar and to test the effects on BNF across a variety of contrasting feedstocks and production temperatures. Biochar was manufactured from sugarcane bagasse, rice hulls, maize stover, maize cobs, eucalyptus wood, Delonix regia wood, and prunings of the tea bush (Camilia sinensis) at pyrolysis temperatures of 350 and 550˚C. Each feedstock and temperature combination was treated with steam to remove both the minerals and the volatile matter, HCl (adjusted back to the original pH) to remove only the minerals, and acetone to remove only the volatile matter and compared to unaltered biochar. The biochars were added to a highly degraded oxisol from western Kenya at the equivalent rate of 15 t ha^-1. With biochar, biomass production increased by 293%, nodulation increased by 780%, nodule biomass increased 49 fold, and BNF increased 30 fold. There were significant differences in all metrics between feedstocks and selective significant differences between production temperatures. While plant P and Ca uptake were correlated to BNF, there were no strong correlations to any measured biochar nutrient or biochar pH. Volatile matter was negatively correlated to nodulation and BNF. The data suggests biochar improves plant nutrition and BNF through mechanisms other than direct nutrient addition while high volatile content is inhibitory to nodulation and BNF. It is possible biochar enhanced mycorrhizal colonization, which then improved plant nutrition, but this has yet to be verified.