|Cabrera, Miguel - UGA|
Submitted to: Ecological Modelling
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: February 1, 2001
Publication Date: N/A
Interpretive Summary: Too much or too little nitrogen (N) reduces cotton yields. Estimating N availability in no-tillage systems is difficult because crop residues tie-up and release N as they decompose. Two versions of a computer model known as CERES were used to simulate the available N from decomposing soil organic matter in cotton-crimson clover and cotton-rye cropping systems. The version modified to allow user input of soluble carbohydrate, cellulose, and lignin pool sizes provided better simulations of field data than did the original version of the model. The original model tended to over predict N availability which was most likely related to overestimation of soil organic matter decomposition and release of N and not due to overestimation of crop residue decomposition and N release. The modified model was a better tool for estimating N needs for cotton in conservation tillage cover crop systems than the original model.
Technical Abstract: Knowing the amount of nitrogen available from crop residues and soil is important for determining fertilizer N needs of a crop. Nitrogen availability in conservation tillage systems is more difficult to assess because of uncertain interactions of surface residues with N mineralization processes. Knowing residual N availability in cotton systems is critical because both over- and under fertilization can reduce lint yields. The CERES plant growth models use a moderately complex N submodel that could be useful as a tool to estimate N needs for cotton. Simulations of in situ net N mineralization under two cotton (Gossypium hirsutum L.) cover crop systems with the original N submodel (CERES-N) and a version modified to allow user input of soluble carbohydrate, cellulose, and lignin pool sizes (CERES-NP) were compared to field data. Both model versions indicated in situ net N mineralization following crimson clover (Trifolium incarnatum L.) was greater than following rye (Secale cereale L.) which agreed with measured results from1997 and 1998. Simulations of in situ N mineralization were better with CERES-NP than for CERES-N and were improved for both versions when using decomposition parameters determined from data of a previous field study. When simulations were compared to data for biomass and N loss from bagged residues, the original model tended to over predict in situ net N mineralization due to overestimation of soil organic matter N mineralization. CERES-NP appeared to be a better tool for estimating N needs for cotton in conservation tillage cover crop systems than the original CERES-N.