Integrating Soil Biological and Chemical Indices to Predict Net Nitrogen Mineralization across California Agricultural Systems

Predicting the capacity of soil to supply crop N (i.e., inorganic N) has proved difficult due to the myriad of interacting soil biophysical factors. We examined soil chemical and biological indices commonly found in commercial soil testing laboratories-total soil C and N, water-extractable organic C and N, K2SO4-extractable C and N, mineralizable C on rewetting, and permanganate- oxidizable C-to estimate net N mineralization across a variety of soils from cover-cropped and non-cover-cropped fields in California (47 sites, n = 157). Results show that both biological and chemical indices are variable in their ability to predict N mineralization, with better relationships being shown for cover-cropped fields. Total C and N contents were the best chemical indicators for both management systems, describing up to 21.8% of the variance. The flush of CO2 measured after capillary rewetting was more variable and had weaker relationships with net N mineralization than respiration measured at 50% water holding capacity. These relationships were only significant in fields that had recently had cover crops and were moderately strong (r = 0.45-0.55), with the strength of the relationship decreasing as the measured interval of the CO2 flush increased. Models were developed for each management system using variables consistently chosen by a partial least squares regression as drivers of N mineralization. These models improved the explained variance in cover-cropped fields but not in the non-cover-cropped fields. The effectiveness of individual predictor variables in these models was inconsistent across climates, illustrating the need for regional and management-specific tests for predicting N mineralization.