Soil tests of biological activity as predictors of rice response to nitrogen

Abstract

Carbon and N indicators of soil health are related to soil N dynamics. Previous research has examined these relationships as bases for N fertilizer recommendations with several crops, though not with flooded rice ( Oryza sativa L.). This study determined relationships between two soil tests of biological activity (STBA) (CO ₂ ‐burst and alkali‐hydrolyzable nitrogen [AHN]) and yields of delayed flood rice in Louisiana. The data were from 21 N trials across five sites (silt loam to clay soils). Yield without N ( Y ₀ ), Y ₀ relative to maximum yield ( Y ₀ / Y _M ), and maximum yield increase (Δ Y _M  =  Y _M − Y ₀ ) were significantly correlated with both STBAs in medium‐texture soils. Correlations were weaker in heavy‐texture soils ( Y ₀ / Y _M and Δ Y _M with CO ₂ ‐burst, and Y ₀ / Y _M with AHN). However, rates of N needed to achieve modeled 95% Y _M were not significantly related to either STBA for either textural group. Calibration failed because slopes of the response curves varied substantially among trials. Alternatively, net yield increase (Δ Y  =  Y − Y ₀ ) was well described by models using CO ₂ ‐burst to predict Δ Y _M ( R ²  = 0.85 and 071, medium and heavy soils), with small mean absolute error (∼6% and 9% of average Δ Y _M , medium and heavy soils). The Δ Y approach describes a decreasing yield increase with N rate that further decreases with CO ₂ ‐burst value. Thus, N rate might be based on modeled Δ Y _M rather than a critical soil test value. Further research is needed on this approach and to explain the variability in response to N.