Phosphorus (P) enrichment frequently occurs in the soil used in greenhouse vegetable production (GVP). Minimizing the application of P fertilizer represents a crucial approach to mitigating the accumulation of P in the soil and enhancing its utilization efficiency. However, the changes in bacterial communities and the turnover mechanism of soil P fractions related to soil P cycling after P fertilizer reduction are still unclear. To unravel these complexities, we devised three experimental treatments: conventional nitrogen (N), P, and potassium (K) fertilizer (N1P1K1); conventional N and K fertilizer without P (N1P0K1); and no fertilizer (N0P0K0). These experiments were conducted to elucidate the effects of P reduction on cucumber plant growth, soil P fractions, and the phoD-harboring bacterial community in the P-rich greenhouse soil. The results showed that there were no significant differences between the N1P1K1 and N1P0K1 treatments in terms of plant growth, yield, and P uptake, and the values for the N0P0K0 treatment were significantly lower than those for the N1P1K1 treatment. In a state of P depletion (N0P0K0, N1P0K1), the main P sources were Resin-Pi, NaHCO3-Pi, NaHCO3-Po, and NaOH-Pi. The contents of NaOH-Po and CHCl-Po in the N1P0K1 treatment increased significantly. Without P fertilizer, alkaline phosphatase (ALP) activity, phoD gene abundance, and bacterial community diversity were significantly increased. The abundance of Ensifer in the N0P0K0 and N1P0K1 treatments was 8 and 10.58 times that in the N1P1K1 treatment, respectively. Additionally, total phosphorus (TP) and available nitrogen (AN) were key factors affecting changes in the phoD bacterial community, while Shinella, Ensifer and Bradyrhizobium were the main factors driving the change in soil P fractions, and NaHCO3-Pi and NaOH-Pi were key factors affecting crop yield. Therefore, reducing the application of P fertilizer will increases the diversity of phoD-gene-harboring bacterial communities and promote organic P mineralization, thus maintaining the optimal crop yield.