Quantifying the spatial variability of groundwater recharge is crucial for water resource management under changing land use conditions. Hydrological model performance is highly influenced particularly for peak discharge simulation for the wetland-dominated watersheds. This study aimed at improving the performance of a numerical model at a wetland-dominated watershed by explicitly accounting for wetland processes. The Soil and Water Assessment Tool (SWAT) was used to simulate the spatial variability of groundwater recharge for the two Land Use and Land Cover (LULC) maps classified in 1996 and 2017. An acceptable model performance during the calibration and validation periods is found by incorporating wetland parameters. The trend of LULC change showed a significant reduction for Wetland and Forest going along with an increase of agricultural land at the rate of 9.56 ​km²/yr. Accounting for the LULC changes in the model, the simulated median groundwater recharge to the aquifer has reduced from 506.25 ​mm to 483.34 ​mm per year. During these periods, the extent of Wetland extremely converted towards Grassland, resulting in a reduction of median groundwater recharge from 508.6 ​mm to 495.3 ​mm per year in wetland-dominated sub-watersheds. In addition, this study demonstrates spatial variability in groundwater recharge with differences in catchment properties such as elevation and slope. Therefore, this study provides spatially quantified values for groundwater recharge under the LULC conditions, which are useful in the management and development of groundwater recharge regions.