Agriculture is associated with increased nutrient inputs to the landscape, which can drive the eutrophication of critical downstream systems, such as the Chesapeake Bay. To address these concerns, restoration techniques such as afforestation, the planting of trees on land that was not recently forested, can be utilized to decrease nutrient loss in impacted watersheds. Despite efforts to remediate the myriad of impacts from agriculture on our freshwater systems, we still lack an understanding of the long-term recovery trajectory for the restoration of croplands. Here, we used monthly composite water quality data from the greater Rhode River watershed (near Edgewater, Maryland), collected as a part of the Smithsonian Environmental Research Center Weir Network. We compared 44 years (1978-2023) of total and dissolved nitrogen (N) and phosphorus (P) from a continuously forested watershed and an agricultural watershed that was replanted with trees in 2013 as part of a tree diversity experiment, BiodiversiTREE. Preliminary results suggest that dissolved N concentrations decreased after fertilizer application ceased and afforestation occurred primarily due to decreases in nitrate (p=0.05). In contrast, TN and TP concentrations increased post-afforestation, likely resulting from increased organic matter. Interestingly, soluble reactive phosphorus (SRP) concentrations increased across the entire study period, with a greater increase post-afforestation (p<0.01). The observed increase in SRP underscores the potential influence of legacy P, which may be mobilized for many years after N and P inputs are reduced, delaying water quality recovery. Comparatively, nutrient concentrations remained constant through time at the fully forested watershed. Our results highlight the importance of utilizing restoration techniques to reduce nutrient export in impacted systems, while underscoring that managers must anticipate multi-decadal recovery periods driven by legacy nutrient mobilization.