Nutrient pollution has become an important issue to solve in stormwater runoff due to the fast population growth and urbanization that impacts water quality and triggers harmful algal blooms. There is an acute need to link the dissolved organic nitrogen (DON) decomposition with the coupled nitrification and denitrification pathways to realize the pattern shifts in the nitrogen cycle. This paper presented a lab-scale cascade upflow biofiltration system for comparison of nitrate and phosphate removal from stormwater matrices through two specialty adsorbents at three influent conditions. The two specialty adsorbents are denoted as biochar iron and perlite integrated green environmental media (BIPGEM) and zero-valent iron and perlite-based green environmental media (ZIPGEM). An initial condition with stormwater runoff, a second condition with spiked nitrate, and a third condition with spiked nitrate and phosphate were used in this study. To differentiate nitrifier and denitrifier population dynamics associated with the decomposition of DON, integrative analysis of quantitative polymerase chain reaction (qPCR) and 21 tesla Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) were performed in association with nitrate removal efficiencies for both media with or without the presence of phosphate. While the qPCR may detect one gene for a single microbe or pathogen and realize the microbial population dynamics in the bioreactors, the 21 T FT-ICR MS can separate and assign elemental compositions to identify organic compounds of DON. Results indicated that ZIPGEM obtained a higher potential for nutrient removal than BIPGEM when the influent was spiked with nitrate and phosphate simultaneously. The sustainable, scalable, and adaptable upflow bioreactors operated in sequence (in a cascade mode) can be expanded flexibly on an as-needed basis to meet the local water quality standards showing process reliability, resilience, and sustainability simultaneously.