Changes in soil test phosphorus are not predictable using fertilizer input and crop removal

Volatile fertilizer prices have led to many producers and researchers exploring different immobile nutrient stratagems to increase the efficiency of fertilizer applications, leading to greater economic return. While many nutrient recommendations are based upon a soil test for that one season, for immobile nutrients such as phosphorus (P), applications in one season are likely to provide residual P that can affect soil test values in the following seasons. Our objective was to develop a model that allows us to predict the change in soil test P at different depths using soil test values, fertilizer inputs, and crop removal in no‐till winter wheat ( Triticum aestivum L.) in Oklahoma. We observed that from pre‐plant application in the fall to harvest in the subsequent summer, only the Mehlich 3 extractable P (M3P) in the 0‐ to 5‐cm depth significantly increased. We also noted that P removal was statistically impacted by P fertilizer application but differed by location. Finally, we found that modeling the change in soil test values was possible using fertilizer P inputs, P removal from crop, and residual pre‐plant M3P, but had a poor correlation ( r 2 = 0.27), and therefore, was not accurate enough for prediction purposes. Our results support previous works in suggesting P cycling dynamics are complex and differ from site to site based upon soil texture, genesis, and management, and broad recommendations based on input and removal from a field would be inadequate for accuracy and precision.