In the face of widespread food insecurity and extreme population growth in sub-Saharan Africa, sustainable, resilient systems are needed to increase agricultural yields that can remediate degraded soils. Agroforestry is one of the suggested strategies and shows some promise. Piliostigma reticulatum and Guiera senegalensis are two indigenous woody species that currently grow interspersed in farmers’ fields in Senegal and throughout the Sahel. Shrub management consists of coppicing and burning the above ground biomass before planting. Previous research has shown that decomposition of above ground biomass contributes to soil organic matter and carbon, nitrogen content and phosphorus availability in soils. Hydraulic lift from the long tap roots redistributes moisture to upper soil horizon. In addition, root exudates and root turnover also create ideal habitats for soil microbes. The collective benefits of shrubs on soils results in `resource islands’ across the Sahalien landscape.Optimized intercropping with Piliostigma reticulatum and Guiera senegalensis have been shown to increase yields in groundnut (Arachis hypogaea) and pearl millet (Pennisetum glaucum) (Dossa et al., 2012, 20132; Bright et al., 2017). Research has shown that shrubs can promote microbial diversity and activity beneath the canopy of these shrubs over soil outside the influence of shrubs (Deidhou et al., 2009). However, little is known on whether this shift also occurs in the rooting zone of crops growing adjacent to shrubs. Therefore, the first objective of this research was to characterize the effect G. senegalensis and P. reticulatum have on soil microbial communities in a shrub-millet intercropping systems in farmers’ fields spanning across a rainfall gradient. A fatty acid methyl ester (FAME) extraction and MIDI microbial biomarker library were used to create a community profile on samples from shrub root zones, millet root zones within shrub influence and millet root zones outside of shrub influence in 6 farmers’ fields across the rainfall gradient. Samples were collected in 2013 and 2014. The microbial biomass from the fatty acid analysis ranged from 15.67-23.75 nmol g-1 soil in 2013 and from 15.77-20.87 nmol g-1 soil in 2013. Neither year demonstrated significant influence on microbial biomass or on the distribution of the FAME profile by the shrubs. The regional treatments caused significant changes in the FAME profile in the millet rooting zone treatments. This suggests that while agricultural management interfered with the shrubs ability to create distinctly heterogeneous resource islands, they were still able to maintain a stable soil habitat to sustain the microbial community.The second objective was to determine the potential of diazotrophs to fix atmospheric N within the canopy soil of G. senegalensis and P. reticulatum along synthetic fertilizer rate and precipitation gradients. An incubation study using the stable isotope, 15N in the form of dinitrogen gas (15N2) was used to investigate the activity of nitrogen fixing bacteria in shrub rhizosphere soil, soil treated with shrub litter cover and soil untreated by shrub from two long term research sites in the central and southern regions of Senegal. Samples were collected from both fertilized and not fertilized plots in the rainy and dry season of 2014.The optimized incubations estimate that BNF is capable of providing from 0.25 with fertilizer to 61.6 kg N ha-1 without at Nioro and 0.03 without fertilizer to 1.1 kg N ha-1 with fertilizer at Keur Matar. The high levels of estimated BNF during the dry season, ~50-100 kg N ha-1 at both sites, may have been obscured by the artificial moisture content. These estimations suggest that BNF would contribute significantly to the 9 kg N ha-1 and 22 kg N ha-1 recommended N fertilization for groundnut and Millet. Keur Matar soil diazotrophs were less significantly affected by nitrogen fertilization while both showed significantly more BNF activity during in the incubation of dry season soils. A study using field conditions would provide a more accurate estimation.