Biochar has been widely used to ameliorate soil quality and increase crop productivity through enhancement of nutrient availability and microbial community. The Karst yellow soil in China is characterized by severe soil degradation owing to intensive nutrient leaching. However, the biochar addition effects on the changes of Karst yellow soil are unclear, and the adequate number of biochar dosages to explain optimum of plant growth in this soil area remains poorly understood. In this study, pot experiments were conducted to examine the effects of biochar addition (1%, 3%, 5%, 7%, and 9% by weight; 0% as a control) on bacterial abundance and community structure via high-throughput sequencing coupled with bioinformatics methods applied to Karst yellow soil with planting ryegrass (Lolium perenne L.) and daylily (Hemerocallis fulva). After adding biochar for 188 days, significantly increased pH, soil organic matter, total nutrient contents, and bacterial abundance, but decreased available nitrogen, were observed. Changed bacterial community structures were found in biochar treatments compared with those without biochar. In both soils of planted ryegrass and daylily, the optimum soil bacterial abundance was found in 7% biochar dosage, but the lowest values were in the controls (0%). Taxonomic analysis identified that Micrococcaceae (24.53%), Oxalobacteraceae (11.87%), and Nocardioidaceae (7.89%) were the dominant family in the soil of ryegrass growth, and Micrococcaceae (16.20%), Xanthomonadaceae (6.94%), and Nocardioidaceae (6.41%) were the dominant family in soil of daylily growth. Canonical correspondence analysis showed that the alterations of soil bacterial abundance and community were highly interrelated with soil chemical properties. The results provided a better understanding of the mechanisms underlying the plant-soil microbe interactions and their responses to biochar dosages in low fertility soil regions.