The increasing use of untreated or partially treated drainage water in arid agricultural regions raises critical concerns regarding water safety and antimicrobial resistance. This study presents an integrated optimization and performance evaluation of a horizontal subsurface flow constructed wetland (HSSFCW) planted with Typha latifolia for treating raw mixed-source drainage water containing mixed organic, microbial, and metal pollutants. Central Composite Design (CCD) within the framework of Response Surface Methodology (RSM) was employed to optimize hydraulic retention time (HRT; 18-78 h) and plant density (4-20 stems), enabling precise multi-contaminant removal. Under optimized conditions, the system achieved a 93% reduction in total chemical oxygen demand (TCOD), >95% removal of chromium (Cr) and lead (Pb), and a 98.5% (1.82 log(10)) decrease in fecal coliforms, along with up to a 70% reduction in the relative inhibition index (RII, %), indicating a substantial decrease in community-level antimicrobial resistance potential. The integration of kinetic modeling and principal component analysis (PCA) revealed key mechanistic interactions governing pollutant attenuation and resistance dynamics. This study provides a predictive, design-oriented framework for nature-based treatment that simultaneously addresses conventional contaminants and antimicrobial resistance, offering a scalable, energy-efficient, and low-cost platform for safe water reuse in arid and semi-arid regions.